Liftand Escalator Symposium Proceedings 2016

6TH SYMPOSIUM ON
LIFT & ESCALATOR

TECHNOLOGIES
Volume 6
September 2016
ISSN 2052-7225 (Print)
ISSN 2052-7233 (Online)
www.liftsymposium.org
Legal notices
The rights of publication or translation are reserved.
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any
form or by any means without prior permission of the copyright holders. Requests for republication
should be made via www.cibseliftsgroup.org
© 2016 The University of Northampton, The CIBSE Lifts Group, LEIA and the authors.
No representation is made with regard to the accuracy of the information contained in these
proceedings. No legal responsibility or liability is accepted by in relation to errors or omissions.
Any commercial products included within this publication are included for the purpose of
illustration only and their inclusion does not constitute endorsement or recommendation.
Organizing Committee
Dr Jonathan Adams, The University of Northampton

Eur Ing David Cooper, The CIBSE Lifts Group (Lift Academy)/LECS (UK) Ltd
Mrs Elizabeth Evans, The CIBSE Lifts Group
Professor Stefan Kaczmarczyk, The University of Northampton (Co-Chair)
Mr Nick Mellor, LEIA/ The University of Northampton
Dr Richard Peters, The CIBSE Lifts Group (Co-Chair)
Scientific Committee
Professor Lutfi Al-Sharif, The University of Jordan, Jordan

Dr Eur-Ing Gina Barney, Gina Barney Associates, U.K.
Dr Rosa Basagoiti, University of Mondragón, Spain
Dr K. Ferhat Celik, Blain Hydraulics GmbH, Germany
Theresa Christy, Lerch Bates Inc, USA
Dr Lee Gray, UNC Charlotte, USA
Dr Jack Hale, University of Newcastle, U.K.
Professor Ignacio Herrera, University of Extremadura, Spain
Mr Philip Hofer, Schindler Elevator Ltd, Switzerland
Professor Erdem Imrak, Istanbul Technical University, Turkey
Dr Ana-Maria Lorente Lafuente, Germany
Dr Bruce A Powell, The Bruce Powell Company, USA
Adam J Scott, Sweco, U.K.
Kevin Seaborne, London Underground, U.K.
Dr Marja-Liisa Siikonen, Kone, Finland
Dr Eur-Ing Rory Smith, The University of Northampton, U.K.
Dr Albert So, The Asian Institute of Built Environment, Hong Kong
Dr Seiji Watanbe, Mitsubishi Electric Corp., Japan
Professor Weidong Zhu, University of Maryland Baltimore County, USA
Professor Xabier Arrasate, Mondragon University (MU) in the Basque Country, Spain
FOREWORD
It is with great pleasure that we present the proceedings of the 6th Symposium on Lift and Escalator
Technologies, 21-22 September 2016, organised jointly by The Lift Engineering Section of the
School of Science and Technology, The CIBSE Lift Group and LEIA.
The Lift Engineering programme offered at The University of Northampton includes postgraduate
courses at MSc/ MPhil/ PhD levels that involve a study of the advanced principles and philosophy
underlying lift and escalator technologies. The programme aims to provide a detailed, academic
study of engineering and related management issues for persons employed in lift making and allied
industries.
The CIBSE Lifts Group is a specialist forum for members who have an interest in vertical
transportation. The group meets regularly to promote technical standards, training and education,
publications and various aspects of the vertical transportation industry. The CIBSE Lifts Group
directs the development of CIBSE Guide D: Transportation systems in buildings, the de facto
reference on vertical transportation.
LEIA is the UK trade association and advisory body for the lift and escalator industry with a
membership covering some 95% of the lift and escalator industry. LEIA members supply
passenger and goods/service lifts, stairlifts, homelifts, lifting platforms, escalators, passenger
conveyors and a range of component parts for such products. LEIA members undertake the
maintenance and modernisation of more than 250,000 products falling within the scope of the
Association. LEIA provides advice on health, safety and standards matters, promotes education and
training especially through its distinctive distance learning programme so is proud to be a co-
organiser of the Symposium.
The Symposium brings together experts from the field of vertical transportation, offering an
opportunity for speakers to present peer reviewed papers on the subject of their research. Speakers
include industry experts, academics and post graduate students.
The papers are listed alphabetically by first author details. The requirement was to prepare an
extended abstract, but full papers were accepted from the invited speakers where they preferred to
offer them. The submissions are reproduced as they were submitted, with minor changes in
formatting, and correction of obvious language errors where there was no risk of changing meaning.
Professor Stefan Kaczmarczyk, and Dr Richard Peters

Co-Chairs and Proceeding Editors
CONTENTS
A Study into the Influence of the Car Geometry on the Aerodynamic Transient 1
Effects Arising in a High Rise Lift Installation
Hayder Al-Jelawy1, Stefan Kaczmarczyk1, Seyed Mirhadizadeh1, Dhirgham
AlKhafaji2
1
The University of Northampton, UK.
2
The University of Babylon, Iraq.
Towards a Systematic Methodology for the Design of Elevator Traffic Systems in 2

High Rise Office Buildings
Lutfi Al-Sharif, Ghazi Al Sukkar, Asma' Hakouz, Nama' A. Al-Shamayleh
The University of Jordan, Jordan
Lifts in Health: Health Technical Memorandum 08-02 Revisited 3

Gina Barney
Gina Barney Associates, UK
Integration of Lift Systems into the Internet of Things and the Need for an Open 4
Standard Information Model
Jonathan Beebe
Jonathan Beebe Ltd, UK
Your Lift Journey – How Long Will You Wait? 5

Caroline Bird1, Richard Peters1, Elizabeth Evans1, Stefan Gerstenmeyer2,
1
Peters Research Ltd, UK
2
thyssenkrupp Elevator Innovation GmbH, Germany
Evaluating a Holistic Energy Benchmarking Parameter of Lift Systems by using 6

Computer Simulation
Ricky Chan, Albert So, Stefan Kaczmarczyk
The University of Northampton, UK
The Report of Thorough Examination as a Management Tool for Maintenance 7

Dominic Dawson
Zurich Engineering, UK
Multicar Dispatching 8
Stefan Gerstenmeyer1,3, Richard Peters2,3
1
2
3
The 1935 Code of Practice for the Installation of Lifts and Escalators 9
Lee E. Gray
University of North Carolina, USA
London Underground Escalator Passenger Safety Strategy Improving Passenger 10

Safety on Escalators
Phil Harley, Kevin Seaborne
London Underground, UK
Pilot for Standing on Both Sides of Escalators 11

Celia Harrison, Neera Kukadia, Paul Stoneman, Grant Dyer
London Underground, UK
An Overview of India, Travelling Tall 12

Anandi Khandekar
TAK Consulting Pvt. Ltd, India
How Current Technology Trends are Empowering Us All to Drive Innovation 13

Ben Langham
London Underground AP JNP Escalator Services, UK
Changed Requirements In The International Lift Market Ask For New Pulley 14
Types With Better Tension Equalization Features
Jawk Meijer, Lothar Sieber
Schwartz GmbH Technical Plastics, Germany
Modelling of a Rope-Free Passenger Transportation System for Active Cabin 15

Vibration Damping
Jonas Missler1, Thomas Ehrl2, Benedikt Meier2, Stefan Kaczmarczyk3, Oliver
Sawodny1
1
University Stuttgart, Germany
2
3
University of Northampton, UK
Dynamic Lift Control for Improvements in Energy Efficiency 16

Vicente Pacheco
IMEM LIFTS, Spain
Fire Lifts, Escalators & Moving Walks Management System (FEMS) in an 17

Airport
Giovanni Pappalardo
ANACAM Lift Association, Italy
Global Dispatcher Interface 18
Richard Peters
Map-Based Active Compensation of Lateral Vibrations in Elevators 19

Raúl Monge, Javier Saumell and José Manuel Rodríguez
Instituto Tecnológico de Aragón, Spain
The Role of Economic Factors in Traffic Planning and Selection of Lift 20

Equipment
Rory S. Smith
Exploring the Concept of Using Lifts to Assist the Evacuation of Very Tall 21
Buildings
Peter Sumner
WSP Parsons Brinckerhoff, UK
Understanding the Requirements of the New EN-81 Standards with Respect to 22

Speed Monitoring, Speed Reducing and Prevention or Stop Devices
Sam Tanno
UK
Installing and Calibrating Loop Amplifiers to EN81-70 so that Test Certificates 23

can be Produced for Audio Frequency Induction Loop Systems (AFILS) in
Accordance with BS EN 60118-4
John Trett
C.E. Electronics Ltd, UK
A Study into the Influence of the Car Geometry on the Aerodynamic
Transient Effects Arising in a High Rise Lift Installation
Hayder Al-Jelawy1, Stefan Kaczmarczyk1, Seyed Mirhadizadeh1, Dhirgham
AlKhafaji2
School of Science and Technology, Department of Engineering and Technology, The University of
1
Northampton, UK.
2
College of Engineering, The University of Babylon, Iraq.
Keywords: Computational fluid dynamics, car geometry, turbulent flow, computer simulation,
aerodynamic performance
Abstract: One of the main goals in designing a high-speed lift system is developing a more
aerodynamically efficient car geometry that guarantees good ride comfort and reduces energy
consumption. In this study, a three-dimensional computational fluid dynamics (CFD) model has
been developed to analyse an unsteady turbulent air flow around two cars moving in a lift shaft. The
paper is focused on transient aerodynamic effects arising when two cars pass each other in the same
shaft at the same speed. The scenarios considered in the paper involve cars having three different
geometries. Aerodynamic forces such as the drag force that occur due to the vertical opposite
motions of the cars have been investigated. Attention is paid to the airflow velocity and pressure
distribution around the car structures. The flow pattern in the boundary layer around each car has
been calculated explicitly to examine the flow separation in the wake region. The results presented
in the paper would be useful to guide lift designers to understand and mitigate the aerodynamic
effects arising in the lift shaft.
1 INTRODUCTION
The fast development of super-high-speed elevators has been facing significant challenges related to
aerodynamic problems, such as high air resistance to the car movement, vibration of the lift car,
excessive pressure fluctuation and noise generated inside the car as it is moving along the shaft.
These problems occur because of the high-speed air flow around the moving car which could be
increasing around the sharp edges. The ride quality is very important for the passengers’ safety and
comfort. Accordingly, it is essential to understand the aerodynamic forces and mitigate their effects.
In their work, Matsukara, Y. et al. [1] and Teshima, N. et al. [2], studied two types of noise. They
stated that the mechanical noise is much smaller than the aerodynamic noise for high-speed lifts. In
order to reduce the aerodynamic noise, Matsukara used a streamlined cover at both sides of the lift
car (top and bottom). On the other hand, Toshima studied the impact of removing the apron which
has to be installed at the bottom of each lift car due to legal requirements. Eventually, the noise was
reduced in a range of (4.1 – 4.3 dB (A)) in Matsukara’s work. Teshima was able to reduce the
aerodynamic noise by producing a guide plate for the apron. They carried out two experiments in
wind tunnels where the cars are stationary facing dynamic air.
Bai, H. et al. [3], have drawn attention to the fact that using wind tunnels is considered to be a total
deviation from the real situation where the car and the air are dynamic. Thus, four different shapes
of moving cars inside a cylindrical hoistway were studied. According to the consideration of
average pressure difference, they considered the proper car shapes to be parabolic, spherical,
conical and cylindrical respectively. This work did not take into account the fact that real cars and
hoistways are rectangular in shapes, as they established that both the car and the hoistway are
cylindrical in shape.
A numerical simulation has been done by Shi, L. et al. [4]. Their work was focused on a two-
dimensional model of unsteady turbulent boundary layer flow around a lift car passing a counter-
1-2 6th Symposium on Lift & Escalator Technologies
weight in the same shaft with different velocities and horizontal gaps. They found a severe increase
of aerodynamic forces when the car passes the counter-weight.
Based on the 2-Dimensional work of Wu, R. et al. [5], the Coriolis force is much smaller than the
lateral aerodynamic buffeting force when two conveyances pass each other.
In 2015, Wu, R., et al. [6], simulated a 3-Dimensional work to compare the lateral aerodynamic
buffeting force and the clearance size of two kinds of rope-guided conveyances (mine lift and mine
cages). Their study shows that the aerodynamic buffeting effect is directly proportional to the
clearance size between the conveyances.
Mirhadizadeh, S. et al. [7], developed a computational software platform for high-speed lift systems
by using MSC Dytran solver. Their model predicted the aerodynamic interactions in high-rise high-
speed lift systems by utilizing CFD and Multibody Dynamics techniques.
According to the previous studies, the aerodynamic performance of high-speed lifts has become
very important as lifts are getting faster in order to achieve the best design. Due to the aerodynamic
influence, a three-dimensional aerodynamic model is developed in order to have a good
understanding of lift cars passing each other in one shaft with different geometries at the same
speed and with the same horizontal clearance between them.
2 GOVERNING EQUATIONS
A three-dimensional incompressible flow has been considered in order to have a better
understanding of the flow. The transport flow variables are governed by two basic physical
principles which are the conservation of mass and momentum. These variables are the pressure (𝑝)
and the flow velocity (𝑢𝑖 ). Mathematical statements of the fluid physical principles are called
Navier-Stokes (N-S) equations and shown in equations (Eq. 1, Eq. 2).
𝜕𝜌 𝜕(𝜌𝑢𝑖 )
+ = 0. 𝑖 = 1,2,3. (1)
𝜕𝑡 𝜕𝑥𝑖
𝜕(𝜌𝑢𝑖 ) 𝜕(𝑢𝑖 𝜌𝑢𝑗 ) 𝜕𝑝 𝜕𝜏𝑖𝑗

+ + 𝜕𝑥 − − 𝐹𝑖 = 0 𝑖, 𝑗 = 1,2,3. (2)
𝜕𝑡 𝜕𝑥𝑗 𝑖 𝜕𝑥𝑗
Where 𝜌 is the mean mass density, 𝐹𝑖 represents the body forces, 𝑝 is the pressure and 𝜏𝑖𝑗 is the
shear stress in the fluid.
3 THE MODEL DESIGN AND COMPUTER SIMULATION

Three simplified CAD designs of a lift car are shown in Figure 1. It shows the geometries of lift
cars. The lift’s design depends on the shape of the host building, according to Bai et al. [1], most of
the lift cars and hoistways are rectangular in shape. Therefore, this study considers the rectangular
shape to be investigated with different top and bottom aerodynamic shroud shapes of the lift cars.
A Study into the Influence of the Car Geometry on the Aerodynamic Transient Effects Arising in a High 1-3
Rise Lift Installation
1m
1m
1m
1m
Figure 1 Schematic layout of three different shapes of the shroud (a) flat; (b) triangular; (c)
hemi-cylindrical
The computer simulation has been implemented in MSC Dytran commercial software system [8].
The system’s fluid solver based on the Finite Volume Method (FVM) is used to generate Eulerian
mesh which is then used to model the dynamic motion of the air around the lift car.
The cars are considered to be rigid bodies with masses 2000 kg each. The air is an ideal gas with
properties as follows: density is 1.2041 𝑘𝑔 𝑚−3, specific heat ratio is 1.401 at 20 °𝐶 and the gas
constant is 287 𝐽 𝑘𝑔−1 𝐾 −1 . In order to have a simple interpolation between the grids, the Cartesian
square grids have been applied in order to reduce the computer time. The interface between the car
and the air grids moves at the same speed. In the present simulation, accurate flow simulations have
been done by taking into account the grid resolution. The unsteadiness of the flow has been
resolved by setting the integration time step at 1 × 10−4 𝑠.
The total number of grid points is approximately 127,000. Lifts are located 12m vertically apart
from each other in order to reduce the computational cost. Figure 2 shows the cross-sectional area
of the hoistway/cars layout.
Figure 2 Cross-sectional area of the simplified cars/hoistway model (top view), all dimensions
in mm
4 RESULTS AND DISCUSSION

To clarify the results, several time stations for the vertical opposite motion of lifts are illustrated in
figure 3. The steady-state solution is applied as an initial condition at 𝑡 = 0.0 when the lifts start
moving. The lifts will be aligned side by side at 𝑡 = 0.3, and the crossing event will be finished
at 𝑡 = 0.6. Each car moves at the speed of 20 𝑚⁄𝑠 in vertical opposite directions.
1 2
1 2
12 m 1 2
2 1
2 1
Figure 3 The time sequence of the relative position of each lift

Figure 4 illustrates the air field and its velocity profile around each car. Also, it shows the pressure
distribution over the cars’ bodies. Both cars are moving inside the hoistway at the same speed (20
m/s) so that Mach number is 0.06. The hoistway height is 90 m, and the lateral (horizontal) distance
between the two lifts is 0.5 m (see figure 2). The Reynolds number (𝑅𝑒) based on the car width
scale is 2.66 × 106 which is calculated as follows:
𝜌𝑉𝐿
𝑅𝑒 = 𝜇
(3)
where:
𝜌: the air density (1.2041 𝑘𝑔𝑚−3) (at 20 °𝐶)
𝑉: the lift car velocity (20 𝑚𝑠 −1 )
𝐿: the lift car width (2 𝑚)
𝜇: the dynamic viscosity of air (1.81 × 10−5 𝑘𝑔𝑚−1 𝑠 −1)
1 2 1 2 1 2
Figure 4 Side view of the flow field around the lifts and the pressure distribution at 𝒕 = 𝟎. 𝟑.
The pressure load fluctuations in the gap between the two cars may cause noise and vibrations. In
the simulation the gamma low equation of state has been applied in order to estimate the initial
value of the pressure as 103.1 kPa [9]. Figure 5 shows the overall pressure acting upon the side
walls of the lift cars. The pressure would be building up and reaching its highest values during the
crossing event. In the scenario considered in this simulation study the event starts when the cars are
12 m apart from each other and ends after the cars have passed each other and are again separated
by the same distance. Consider, for instance, the time instant when the two lifts are alongside each
other (𝑡 = 0.3). For lift 1 with triangular shrouds, the maximum air pressure acting on the side wall
is then 97.7 kPa, and the maximum air pressure acting on the side walls of the lift with hemi-
cylindrical and flat shrouds is 98.2 kPa and 99.1 kPa, respectively.
(a)
(b)
Figure 5 Pressure fluctuation over the side walls of each lift (a) at lift 1 side wall facing lift 2.
(b) at lift 2 side wall facing lift 1.
One of the main forces acting on the lift body is the drag force. The time histories of the
aerodynamic drag force are shown in Figure 6. The drag forces have been determined through a
simulation test with the gravity effects removed. It is clear that the drag force fluctuation acting on
the lift with triangular shrouds is less than the other lifts.
(a)
(b)
Figure 6 the time history of the aerodynamic drag forces (a) at the top of lifts 1 at the bottom
of lifts 2; (b) at the bottom of lifts 2
The resultant overall forces that act on the coupling surface due to the fluid effects are shown in
Figure 7. The fluctuation shows that the highest forces occur when the two cars pass each other. It is
clear that the resultant force acting on the triangular lifts approaches approximately 23 kN. On the
other hand, the highest forces affecting the other shapes are 46.8 kN and 44.9 kN for the flat and
hemi-cylindrical lifts respectively.
Figure 7 The overall forces acting on the coupled surfaces of the lift cars
The drag force will also have an effect on the flow in the wake region of each lift. In this kind of
engineering problem, predicting and modelling the shear layer separation is essential because it has
a significant impact on the opposite moving bodies due to the vortex shedding in its wake region.
The air flow patterns and velocities are illustrated in Figure 8. The diagrams presented in this figure
show that the turbulence behind the ‘flat’ lifts is much higher than the turbulence corresponding to
the other shapes. Furthermore, the flow behind the ‘triangular’ lift 1 tends to reattach at the wall
side rather than to the side of the moving lift 2. Thus, lift 2 will experience less turbulence during
the crossing event.
1 1 1
2 2 2
(a)
2 2 2
1 1 1
(b)
Figure 8 The air flow patterns and its velocity in the hoistway while the lifts pass each other
(a) at 𝒕 = 𝟎. 𝟐 (b) at 𝒕 = 𝟎. 𝟒
5 CONCLUSION
The very fast development of the construction of high-rise buildings raises an essential need for the
design of high-speed lifts. The aerodynamic performance of these lifts has been discussed in the
paper. Attention has been paid to the scenario in which two lifts are passing each other in the same
hoistway. The flow field, pressure distribution, velocity, drag forces and the flow patterns have been
studied. It was revealed that the lift car geometry design plays a significant role in the aerodynamic
performance of the lift itself. From the aerodynamics point of view, the results also indicate that the
triangular shape of the lift’s top and bottom shroud would be the best design in comparison with the
flat and the hemi-cylindrical shapes.
REFERENCES
[1] Matsukura, Y., E. Watanabe, Y. Sugiyama and O. Kanamori, (1992). New mechanical
techniques for super high-speed elevators. Elevator Technology. 4, Proceedings of ELEVCON’92.
Amsterdam, the Netherlands, pp. 174-181.
[2] Teshima, N., Miyasako, K. & Matsuda, H., (1992). Experimental and numerical studies on
ultra-highspeed elevators. Elevator Technology, 4, pp.276–285.
[3] Bai, H., Shen, G. and So, A., (2005). Experimental-based study of the aerodynamics of
super-high-speed elevators. Building Services Engineering Research and Technology, 26(2),
pp.129–143.
[4] Shi, L., Liu, Y., Jin, Z., Cao, Z., (2007). Numerical Simulation of Unsteady Turbulent Flow
Induced by Two-Dimensional Elevator Car and Counter Weight System. Journal of
Hydrodynamics, Ser.B, 19(6), pp.720–725.
[5] Wu, R., Zhu, Z., Chen, G., Cao, G. and Li, W., (2014). 1263. Simulation of the lateral
oscillation of rope-guided conveyance based on fluid-structure interaction. Journal of
Vibroengineering, 16(3).
[6] Wu, R., Zhu, Z. & Cao, G., (2015). Computational Fluid Dynamics Modeling of Rope-
Guided Conveyances in Two Typical Kinds of Shaft Layouts. PloS one, 10(2), p.e0118268.
[7] Mirhadizadeh, S., Kaczmarczyk, S., Tongue, N., Al-Jelawy, H., Feldhusen, P., Delk, W.,
Anderson, K., Dudde, F., (2015). Modelling and Computer Simulation of Aerodynamic Interactions
in High-Rise Lift Systems. In The 5th Symposium on Lift and Escalator Technologies.
Northampton, UK, pp. 117–123.
[8] MSC. Software Corporation, (2013). Dytran Explicit Solution for Transient Structural
Dynamics and Fluid-Structure Interaction. Product Datasheet-DytranTM.
[9] Powell, K.G., Roe, P.L., Linde, T.J., Gombosi, T.I. and De Zeeuw, D.L., (1999). A solution-
adaptive upwind scheme for ideal magnetohydrodynamics. Journal of Computational Physics,
154(2), pp.284-309.
BIOGRAPHY
(1) Mr. Hayder Al-Jelawy
Hayder has a master’s degree in Mechanical Engineering from the University of Technology in
Iraq. His expertise is in the area of applied mechanics and computational fluid dynamics. Currently,
Mr. Hayder is a PhD. Student at the University of Northampton. He is also an associate member of
the Institution of Mechanical Engineers and in the Institute of Physics.
(2) Professor Stefan Kaczmarczyk

Stefan Kaczmarczyk has a master’s degree in Mechanical Engineering and he obtained his
doctorate in Engineering Dynamics. He is Professor of Applied Mechanics and Postgraduate
Programme Leader for Lift Engineering at the University of Northampton. His expertise is in the
area of applied dynamics and vibration with particular applications to vertical transportation and
material handling systems. He has been involved in collaborative research with a number of
national and international partners and has an extensive track record in consulting and research in
vertical transportation and lift engineering. Professor Kaczmarczyk has published over 90 journal
and international conference papers in this field. He is a Chartered Engineer, being a Fellow of the
Institution of Mechanical Engineers, and he has been serving on the Applied Mechanics Group
Committee of the Institute of Physics.
Towards a Systematic Methodology for the Design of Elevator
Traffic Systems in High Rise Office Buildings
Lutfi Al-Sharif1, Ghazi Al Sukkar2, Asma' Hakouz2, Nama' A. Al-Shamayleh1

1
Mechatronics Engineering Department, The University of Jordan, Amman 11942, Jordan
2
Electrical Engineering Department, The University of Jordan, Amman 11942, Jordan
Abstract. Elevator traffic system design has been traditionally based on rules of thumb and the
designer’s judgement and expertise. This is especially true for high rise buildings. This paper
attempts to develop a systematic methodology for the design of high rise buildings, by the use of
rational rules.
In order to ensure clarity and consistency, the paper defines the terms sector, zone and stack.
The systematic methodology is built around the use of rational rules. Rational rules differ from
rules of thumb in a number of ways, and these are discussed in the paper. Six rational rules are
presented and used in the design of elevator systems in high rise buildings. The rules are triggered
by the checking of a number of design parameters such as the waiting time and the travelling time,
as well as the core area used up and the number of elevators in the group. A simulator for incoming
traffic and a single entrance is used in order to obtain the parameters for a design and then to trigger
the rational rules.
Note: An earlier version of the paper has been sent out to a number of industry experts for their
comments. Nine industry experts have provided detailed written commentary on the content of this
paper. Their responses, as well as a detailed section of case studies, have been included in the final
copy of this paper which has been sent to a peer reviewed journal for review and potential
publication.
Nomenclature
AWT average waiting time in seconds
AQT average queue length in persons
ATT average travelling time in seconds
1 INTRODUCTION
Elevator traffic system design has been traditionally based on rules of thumb and the designer’s
judgement and expertise. This is especially true for high rise buildings. These are two examples of
the general rules of thumb that are used:
• The number of floors in a single zone should not exceed 18 to 20 floors.

• Sky lobbies should be introduced for buildings exceeding 50 to 60 floors.
Other examples of simple rules of thumb that are used in the conventional design of elevator traffic
systems can be found in [1]. There are three problems with the use of rules of thumb:
• They do not explicitly provide an explanation for the rationale underlying the rule to others,
despite the fact that the designer who uses them does.
• Following on from the previous point, if the assumptions on which the rule was based
change, the rule cannot be changed accordingly.
• Rules of thumb cannot be used to develop a systematic design methodology.
The aim of this document is to develop a set of rules that can guide the designers throughout the
elevator traffic system design process for high rise buildings. Solely for clarification in this piece of
work, a high rise building is defined as any building that has more floors than those that can be
accommodated in a single zone (thus requiring multiple zones or a sky lobby, both of which are
defined in section 2 of this document). The rules presented later in this paper present a
methodology for deciding how many floors can be accommodated in one zone.
The rules will be based on rational reasoning, whereby the rationale on which the rule is based will
be clearly stated. This ensures that where the underlying assumptions change, the rule changes
accordingly. In addition, the rules are fully transparent showing the threshold values of the
different parameters for the different rules. The designer can thus change these thresholds as he/she
sees fit.
It will be assumed that the designer starts with a calculation that will provide a starting point for the
simulation. The design process followed in this paper has been based on the methodology found in
[2] and [3]. The round trip time calculation using equations has been based on the equations found
in [4], [5] and [7]. Under certain situations, it is necessary to use the Monte Carlo simulation
method to evaluate the round trip time [6] and the average travelling time [8]. The design process
then moves to simulation in order to fine tune the design.
It will also be assumed that the designer possesses the required skills to carry out the design of a
single zone elevator traffic system. The methodology for designing a single zone elevator traffic
system is considered to be beyond the scope of this document. In effect, the design of a single zone
elevator traffic system is the basic building block that will be re-used in all high rise building
designs.
Section 2 introduces the terminology that is used to describe how the building is split into different
units, such as sectors, zones and stacks. A clear terminology in this regard is essential for
understanding the rest of this paper. Section 3 provides an overview of the work to date in the area
of high rise vertical transportation system design. Section 4 emphasises a basic principle in using
the rules that will be later introduced in this paper (namely that the rules are there to guide and aid
the designer rather than present a final solution). Section 5 discusses the impact that destination
group control will have on the design of vertical transportation system for high rise buildings.
Section 6 introduces the concept of normalisation in the context of elevator traffic systems. The
core of the paper is in section 7 which presents the six rules. Conclusions are drawn in section 8.
2 SECTORS, ZONES AND STACKS

No unanimous agreement exists in the industry on the exact definitions of the terms sector, zone
and stack. In order to avoid any ambiguity, these terms are defined as shown below within this
piece of work.
Sector: A group of floors (usually, but not necessarily, contiguous) that are grouped together
in the controller software and are served by one or more specific elevator(s). The allocation
of a sector to one or more elevators and the sector’s composition are not necessarily fixed
and can be dynamically changed from one round trip to the next.
Zone: A group of mostly contiguous floors that are served by a number of elevators
operating in one group. The size and composition of a zone is fixed (e.g., location of the
machine room) and cannot be altered. It is usually motivated by the need to reduce the
average travelling time and the need to restrict the number of elevator cars in a group to
eight or fewer. It results in a saving in floor area on the floors above the lower zone(s).
Towards a Systematic Methodology for the Design of Elevator Traffic Systems in High Rise Office 2-3
Buildings
Zoning can be also be used as a tool for traffic segregation (e.g., hotel, offices, residential).
A zone can contain a number of sectors.
Stack: A stack is formed when a number of zones are grouped together and served by a
common sky lobby that channels the incoming traffic. The lowest stack is in fact served by
the main entrance and does not require a sky lobby. A stack that is served by a sky lobby
can be thought of as a building that has been placed inside another building. A stack can
contain a number of zones.
An example of a chart that graphically illustrates the use of zones in the design of high rise
buildings can be found in [9].
3 OVERVIEW OF EXISTING WORK ON HIGH RISE BUILDING

TRANSPORTATION DESIGN
This section presents a general overview of the various contributions to high rise vertical
transportation system design.
A good example of the detailed design of a high rise building (2 IFC in Hong Kong) has been
presented in 2004 by To & Yip [10]. It has 88 floors and a population of 15,000. It has seven
zones and two sky lobbies served by double deck elevators.
There are eight rules of thumb listed in To & Yip [10]. Some of these have been reproduced below:
1. The target handling capacity for the local zones shall be more than or equal to 12% and the
target interval less than or equal to 30 seconds.
2. The target handling capacity for the shuttle systems shall be more than or equal to 15% and
the target interval less than or equal to 20 seconds.
3. Top down sky lobby design systems are to be avoided.
4. There shall be no more than 8 elevators in each group.
5. The number of floors in a local zone should be around 15 floors.
6. The rated car capacity should be 1600 kg (20 persons).
7. The rated car capacity of double deck elevators should be 1600/1600 kg or 1800/1800 kg.
8. When used, a shuttle elevator shall serve no more than three local zones.
9. The passenger journey shall comprise no more than one transition between different elevator
systems (e.g., a shuttle and then a local zone).
10. For each elevator that forms part of a local zone it shall serve no more than two floors (i.e.
the ratio of floors to elevators in the local zone should be in the ratio of 2:1).
Jochem Wit [21] presents a number of building design examples on the use of destination group
control to remove the need for zoning a building. Destination group control has been used as a
means of segregating the different modes of traffic in the building.
In [9] it is shown that the two most important parameters that influence the design of a high rise
building are the number of floors above the main entrance and the total population.
An expert system is described in Alexandris [11]. It uses forward and backward chaining inference
mechanisms in order to accept or reject certain solutions. It has a set of if-then rules. An example
of one of the rules is:
If passenger waiting time is more than 50 seconds then reject solution
Another rule uses natural language descriptions:

If “loading is high” AND passenger waiting time is normal AND system cost is reasonable
then accept solution.
He points out that a user would like to query the software as to how the decision was made for a
certain design. He also discusses a rule base in which the user can modify or amend existing rules
or add new ones.
In [12] Barney presents a general overview of vertical transportation systems in tall buildings. The
paper contains clear definitions of low, mid and high rise, tall and very tall and skyscrapers. It also
contains an excellent overview of the different arrangements of high rise design buildings (example:
Petronas Towers).
Browne & Kelly [13] present an overview of the simulation carried out to assess the performance of
the elevator traffic system for two of the buildings in the World Trade Center (destroyed in the
attack of 11th September 2001).
Caporale [13] suggests normalising the average journey time (AJT) (25% of the 5 minutes, or 75
seconds). The reasoning for the five minute suggestion is not clear. The link might be the fact that
five minutes is used as the basis for quantifying the arrival rate in a building (i.e., elevator systems
in buildings are designed based on the expected arrival rate expressed as a percentage of the
building population in the peak five minutes, denoted as AR%).
Fortune [14] states that the key to efficient high rise design is to stack the zones on top of each
other. He also suggests that a two-minute headway should be achieved for the shuttle elevators. He
also lists the seven technical problems that face any high rise design. He then outlines a general
methodology for even going higher by effectively stacking buildings on top of each other (50 to 60
floor high buildings stacked on top of each other).
Howkins [16] classifies buildings further as follows:
• 40-60 floors denoted as tall buildings, of which many exist and can provide information and
feedback.
• 60-80 floors denoted as very tall buildings, of which a good number exist and can provide
information and feedback.
• 80+ floors denoted as super-high-rise buildings, of which not many exist (less than 20).
• 150+ floors denoted as super-high-rise/super-volume buildings, of which none exist at
present.
Howkins [16] then:
• Calculates the actual core area and the lost potential rent from such an area.
• Presents a systematic procedure for designing elevator systems in high rise buildings.
• Suggests that the population density falls for high rise and tall buildings to a density much
lower than 10 m2 per person.
Mitric presents in [17] and [18] the concept of a total useful area in the building and presents a set
of curves that peak at a certain arrangement.
Powell uses the term banking (meaning zoning) and uses dynamic programming to decide on the
optimum arrangement. [19].
Buildings
4 GUIDANCE TO THE DESIGNER RATHER THAN AUTOMATED DESIGN

It is not the intention of this piece of work to embed these rules into automated software that will
complete the system design independently. The aim of this piece of work is to provide a set of rules
that will be used to guide the designer throughout the design process. Judgement will be required at
each stage by the human designer in accepting, rejecting or modifying the suggestions by the
software.
For example, the designer could use a combined calculation and simulation software
package that provides the outputs of the design process over a number of stages. At each stage, the
software will issue notifications, warnings and suggestions to the designer. It is up to the designer
to heed the warnings and then accept, reject or modify the suggestions. The design process then
proceeds to the next stage.
5 EFFECT OF DESTINATION GROUP CONTROL

With the increased popularity of destination control systems, and as they become more of a standard
feature in elevator group control systems, it is acknowledged that some of these rules will need to
be amended. As an example, one of the rules presented later in this document uses the value of the
average travelling time as a trigger for the introduction of multiple zones. However, in cases where
all the design parameters are acceptable except for the value of the average travelling time, the use
of destination group control could address this problem. In the long term, the use of destination
group control could have a significant impact on the use of zones as well as their numbers. The use
of destination control systems without resorting to zoning could lead to loss of floor space, but can
be used as a future proofing insurance policy against changes in the building population or its use.
As can be seen in the discussion above on destination group control, the advantage of clearly stating
the rationale underlying a rule makes the rules robust to changes in technology and current
acceptable performance parameters (as opposed to the use of rules of thumb).
Jochem Wit [21] provides an interesting case study in which he uses a destination control system in
a high rise building to segregate the different modes of traffic heading to different parts of the
building (e.g., hotel, office, residential) with overall savings arising from the sharing the capacity of
the elevators in the group.
6 PARAMETERS AND NORMALISATION

As will be seen in the next section, the antecedents (the first statement in the if-then rule) of most of
the rules are based on checking one of the parameters of the elevator traffic design (e.g., number of
elevators in the group) or one of the performance parameters (e.g., average waiting time). In the
rules presented, the authors have assumed acceptable thresholds for these parameters (e.g., 90 for
the average travelling time and 30 seconds for the average waiting time). These are subjective
decisions, and others could use different parameters (e.g., the maximum value of the average
waiting time) and different values for such parameters if desired. This is an advantage rather than a
disadvantage and makes the use of the rules more flexible.
For the sake of completeness, the definitions of the average waiting time and the average travelling
time assumed in this document are shown below:
• Passenger waiting time: The time from the arrival of the passenger in the lobby until he/she
starts boarding the elevator (i.e., it does not include his/her boarding time). It is
acknowledged that these differ from the ones in [22] which have been proposed by a number
of industry experts.
• Passenger travelling time: The time from the start of the passenger boarding the car until
he/she completes alighting at his/her destination. It is acknowledged that these differ from
the ones in [22] which have been agreed by the industry.
• The average of each of the two parameters above is the average of the waiting time or
travelling time of all the passengers in the simulation workspace, respectively.
Normalisation is a powerful tool that allows the generalisation of the rules across different buildings
and different scenarios. One of the parameters that will be normalised in the rules introduced in the
next section is the average queue length. It is meaningless to quote this as an absolute number and
it makes more sense to normalise it by dividing it by the rated car capacity. The normalised average
queue length represents the number of car loads waiting in the lobby on average, and is effectively
a measure of the system performance.
Caporale suggest the normalisation of the average waiting time as a percentage of the five minutes
design period (300 s).
7 THE RULES
This section presents the six main rules. Each rule also has some sub-rules. The rules that are used
are crisp. The problem generally with crisp rules is that they have a clearly defined threshold,
something that does not well reflect the way human experts think. It is hoped that these rules will
be further developed in the future to fuzzy rules based on fuzzy logic.
Some of these rules are invoked at the calculation stage, while others are invoked at the simulation
stage. Both the antecedent statement and the consequent statement are shown inside curly brackets.
Where more than one antecedent is present, their relative strength is indicated inside square
brackets.
7.1 Rule 1
This rule is effectively a trigger for zoning and appears at the calculation stage.
If
{the number of elevators is more than 8 for conventional group control (or 12 for
destination control) [stronger antecedent]
and
the car capacity is more than 26 persons/2000 kg [weaker antecedent]}
then
{zone the building (or increase the number of zones if already zoned)}
It is also possible in some cases to use sectoring instead of zoning, offering more flexibility for the
future, but leading to a loss of floor area.
The rationale for limiting the elevator car capacity is the fact that larger cars become very
inefficient when passengers are boarding and alighting during a stop.
The rationale for limiting the maximum number of elevators in a group is to provide sufficient time
for the passengers to get to the desired elevator through the crowded lobby in good time. A better
(and more rational) expression of this rule would be to use the passenger-to-elevator-lobby-
travelling-time, but little information is available currently on this detail.
Buildings
7.2 Rule 2
This rule addresses the problem of excessive average travelling time (assuming all other parameters
are acceptable). It is invoked during the simulation phase.
If
{the average travelling time is more than 90 s}
then
{zone the building (or increase the number of zones if already zoned)}
The reason for the limit of 90 s is based on passenger behaviour and tolerance to journey length. In
general passengers are around twice as tolerant to travelling time as they are to waiting time.
The rule above assumes that conventional non-sectored group control is used in the simulation. It is
also possible in some cases to use sectoring instead of zoning, offering more flexibility for the
future, but causing loss in the floor area. This future proofs the elevator system in the building
against future changes. The term sectoring is used here in its widest meaning, whereby destination
group control is considered an advanced mode of sectoring.
7.3 Rule 3
This rule provides guidance to the user on the split of the building population between the various
zones. It is invoked during the simulation stage.
When zoning, divide the population into the following percentages:
Two zones: lower zone, 57%, upper zone 43%.
Three zones: lower zone 43%, middle zone: 30%, upper zone 27%.
Four zones: 1st zone 29%, 2nd zone 27%, 3rd zone 22%, 4th zone 22%.
The rationale for this rule is to try to equalise the number of elevators in each group serving each
zone. Preference is given to the following if possible:
• Equal elevators in each zone (for symmetry).

• An even number of elevators in each zone.
The calculation stage will assign appropriate speeds to the elevators in different zones (as for
example where the HARint Plane methodology is used [2]). This is usually based on the rational
requirement of travelling between terminal floors in less than 20, 25 or 30 seconds (accounting for
acceleration, deceleration and jerk). The origins of the three suggested values are explained in the
HARint Plane methodology paper [2]. Testing the three values could result in the reduction of the
number of elevators, or optimising the speed depending on the results.
7.4 Rule 4
This rule is added as an extra check to ensure the adequacy of the car capacity that results from the
calculation stage and is invoked at the simulation stage. The rationale for this rule is that the car
loading sometimes needs to be increased under simulation from the value stipulated in the
calculation stage. This is due to the random effects of queuing theory.
If
{all the following parameters are acceptable (number of elevators, car capacity,
average travelling time)
and
the average waiting time is more than 30 s
and
the average queue length is more than the car capacity (or the normalised average
queue length is more than 1)}
then
{increase the car capacity}
The antecedent of this rule uses the average waiting time and the average queue length, which in
turn are heavily dependent on the value of the workspace (i.e., the period over which passengers are
generated for the purposes of simulation). A typical value for the workspace is 900 s (15 minutes).
More details about the effect of the workspace can be found in [23].
7.5 Rule 5
This rule can be used to invoke the use of double-deckers at the calculation stage.
If
{car capacity is very large (much larger than 26 persons, e.g., 48 persons)
and
the number of elevator in the group is acceptable}
then
{use double deckers}
The rationale for this rule is the saving in floor area which results from the use of a smaller number
of double deck elevators compared to a larger number of single deck elevators.
7.6 Rule 6
Zoned systems discussed earlier can be referred to as direct from ground (DFG) systems, in which
passengers have the luxury of being able to travel to their destination in one trip. The main rational
driver for introducing sky lobbies is the loss of floor area, in addition to the physical limitation
imposed by the fact that steel ropes place a limit on the maximum possible travel. Hence this rule
uses the loss in floor area used by the elevator shafts, lobbies and machine rooms as a trigger for the
use of sky lobbies.
The antecedent for this rule is the ratio of the net area to the area used by the elevators (shaft, lobby,
machine room). When this ratio exceeds a certain value, then the building ceases to be feasible and
sky lobbies must be introduced.
If
{the ratio of the net area to the elevator area exceeds 4 to 1 respectively}
then
{introduce sky lobbies}
This rule ensures that the building efficiency (net area to gross area) does not deteriorate. It could
be based on a simple 10 m2 net area per person and ISO 4190-1 areas for elevator shafts and
machine rooms; but could be based on whatever information is available to the designer (e.g.,
population per floor). It has also been found in practice that it is difficult to introduce more than
four zones in a building while still satisfying this area rule.
Buildings
A fictitious building with progressively increasing numbers of floors (10, 20, 40, 50 and 60 floors)
has been used to illustrate the ratio used in the antecedent in the rule. It can be seen that the
threshold ratio of 4:1 is exceeded somewhere between 40 floor and 60 floors. The results are shown
in Table 1.
Table 1 Areas taken by the elevators in five fictitious buildings using direct from ground
(DFG) arrangements.
Number of floors above Net area: Area taken up by elevators

the main entrance (shaft, lobby and machine room):
10 13.7:1
20 8.8:1
40 4.4:1
50 3.77:1
60 3.18:1
Ear comfort due to change in pressure at high speeds and long distances could also be used as a
secondary trigger for sky lobbies. It has been suggested that when the travel distance is more than
300 m and the speed is around 8 m/s a sky lobby is recommended in order to allow passengers to
rest and adapt to the change in pressure.
8 CONCLUSIONS
This document has presented six rules that can be used to guide the elevator traffic system designer
throughout the design process. It has been assumed that the design process proceeds in two stages:
calculation and simulation.
It is not the intention to use the rules for automated design software, but instead to guide the
designer through the design process by issuing notifications, warnings and suggestions.
The rules have been based on rational reasoning; with the obvious advantage that where the
underlying assumptions change, the rules can be easily adapted to suit.
The rules guide the designer as to when to zone the building in order to reduce the average
travelling time and in order to keep the number of elevators in the group below a pre-defined
number. The rules also use the net areas to the elevator areas ratio as the trigger for the introduction
of sky lobbies.
It is worth noting that all of the rules presented here are crisp rules. Crisp rules suffer from the
problem of making a sudden change once a parameter has exceeded a threshold value. The use of
fuzzy logic and fuzzy rules would be more appropriate and would better reflect the nature of human
judgement in the design process.
REFERENCES
[1] Al-Sharif L. Introductory Elevator Traffic System Design (METE VI). Lift Report 2015;
41(3): 32-44. May/Jun 2015.
[2] Al-Sharif L, Abu Alqumsan A M and Abdel Aal O F. Automated optimal design
methodology of elevator systems using rules and graphical methods (the HARint plane). Building
Services Engineering Research & Technology 2013; 34(3): 275-293.
[3] Al-Sharif L, Abdel Aal O F, Abu Alqumsan A M and Abuzayyad M A. The HARint Space:
A Methodology for Compliant Elevator Traffic Designs. Building Services Engineering Research &
Technology, 2015; 36(1): 34–50.
[4] Al-Sharif L and Abu Alqumsan A M. Stepwise derivation and verification of a universal
elevator round trip time formula for general traffic conditions. Building Services Engineering
Research and Technology 2015; 36(3): 311-330.
[5] Al-Sharif L. The effect of multiple entrances on the elevator round trip time under up-peak
traffic. Mathematical and Computer Modelling 2010; 52(3-4): 545-555.
[6] Al-Sharif L, Dahyat H and Al-Kurdi L. The use of Monte Carlo Simulation in the
calculation of the elevator round trip time under up-peak conditions. Building Services Engineering
Research and Technology 2012; 33(3): 319–338.
[7] Al-Sharif L, Abu Alqumsan A M and Khaleel R. Derivation of a Universal Elevator Round
Trip Time Formula under Incoming Traffic with Stepwise Verification. Building Services
Engineering Research and Technology 2014; 35(2): 198–213.
[8] Al-Sharif L, Abdel Aal O F and Abu Alqumsan A M. Evaluating the Elevator Passenger
Average Travelling Time under Incoming Traffic Conditions using Analytical Formulae and the
Monte Carlo Method. Elevator World 2013; 61(6): 110-123.
[9] Al-Sharif L and Seeley C. The effect of the building population and the number of floors on
the vertical transportation design of low and medium rise buildings. Building Services Engineering
Research and Technology 2010; 31(3): 207-220.
[10] To A and Yip P K. The challenge of vertical transportation system for two IFC. First
Interntional Conference on Building Electrical Technology (BETNET) 2004; 1: 152-157. The
Institute of Engineering and Technology. Hong Kong 11th to 13th October 2004.
[11] Alexandris N. LIFTES – An expert system for lift system design. Elevator Technology
1988; 2: 1-9.
[12] Barney G. Vertical Transprotation in Tall Buildings. Elevator World 2003; 51(5): 66-71.
[13] Browne J J and Kelly J J. Simulation of Elevator System for World’s Tallest Building.
Transportation Science 1968; 2(1): 35-56.
[14] Caporale B. Editor’s Overview, Calling all Cars and Traffic Analysts. Elevator World 2004;
52(9): 4. September 2004.
[15] Fortune J. Mega High Rise Elevators. Elevator World 1995; 43(7): 63-69. July 1995.
[16] Howkins R. Elevator Core Areas, A comparison with Existing Structures and those of the
Future. Elevator World 1998; (11):58-64. November 1998.
Buildings
[17] Mitric S. Elevator Systems for Tall Buildings, Part I, Single Mode Elevator Systems.
Transportation Science 1975; 9(1): 54-73. February 1975.
[18] Mitric S. Elevator Systems for Tall Buildings, Part II, Mixed Mode Elevator Systems.
Transportation Science 1975; 9(1): 74-85. February 1975.
[19] Powell B. Elevator Banking for High Rise Buildings. Transportation Science 1975; 9(3):
200-210.
[20] Powell B. Optimal Elevator Banking under Heavy Up-Traffic. Transportation Science 1971;
5(2): 109-121.
[21] Wit J. Sharing Elevator Capacity: Exploring the Unused Potential of Stacked Mixed-Use
High Rise Buildings. Elevator Technology 2014; 20: 262-272.
[22] Barney G. Towards agreed traffic design definitions. Elevator World 2005; 53(2); 108.
February 2005.
[23] Al-Sharif L, Abu Alqumsan A M, Ghanem W, Tayeh I and Jarrar A. Modelling of Elevator
Traffic Systems Using Queuing Theory. 4th Symposium on Lift and Escalator Technologies 2014.
Vol 4, pp 9-18. University of Northampton, September 2014. doi: 10.13140/RG.2.1.1392.0165.
Lifts in Health: Health Technical Memorandum 08-02 Revisited
Gina Barney1
Gina Barney Associates, PO Box 7, Sedbergh, LA10 5LU
1
Key words: lifts, healthcare, hospitals
Abstract: The Department of Health (DH) is responsible for the health and adult social care matters
in England, along with a few elements of the same matters which are not otherwise devolved to the
Scottish Government, Welsh Government or Northern Ireland Executive. It oversees the English
National Health Service (NHS). The NHS employs more than 1.6 million people, putting it in the
top five of the world’s largest workforces together with the US Department of Defence,
McDonalds, Walmart and the Chinese People’s Liberation Army. The NHS in England is the
biggest part of the system, catering to a population of 53.9 million and employing more than 1.3
million people. The DH publishes Health Technical Memoranda (HTM) and Health Building Notes
(HBN). HTM 08-02 Lifts provides guidance and recommendations for lifts to be provided in all
healthcare buildings from the simplest rural practice with one lift to high rise facilities with many
lifts. Lifts were originally covered in HTM 2024: 1995. This was replaced by HTM 08-02 in
February 2010, which was written by the author and peer reviewed by an expert panel. It is held to
be authoritative in the UK healthcare field. Since 2010 many changes have occurred in regulations,
standards and the state of the art. The author has updated HTM 08-021 and presents her work in
this paper. She also describes the structure of the HTMs and HBNs published by the DH.
1 HOW THE DEPARTMENT OF HEALTH MANAGES ITS ESTATE

The Department of Health (DH) has a duty of care as a healthcare provider to ensure appropriate
governance, effective management, the application of best practice engineering standards and
policy during the whole building lifecycle of all healthcare buildings.
Figure 1: Whole building lifecycle
1
Published June 2016
Figure 2: Health specific documents
Health Technical Memoranda (HTM) underpin the DH's duty of care as they give comprehensive
advice and guidance on the design, installation and operation of specialised building and
engineering technology used in the delivery of healthcare. The HTMs are supported by Health
Building Notes (HBN), which give best practice guidance on the design and planning of new
healthcare buildings and on the adaptation/extension of existing facilities. Annex A gives more
information on HTMs and HBNs.
Table 1: Health Technical Memoranda
HTM00 Policies and principles

(applicable to all Health Technical Memoranda)
HTM01 Decontamination
HTM02 Medical gases
HTM03 Ventilation systems
HTM04 Water systems
HTM05 Fire safety
HTM06 Electrical services
HTM07 Environment and sustainability
HTM08 Specialist services
Lifts in Health: Health Technical Memorandum 08-02 Revisited 3-3
2 THE HEALTH TECHNICAL MEMORANDA

The Health Technical Memoranda are structured into a suite of eight core subjects and one
overriding HTM as shown in Figure 2 and Table 1.
Some subject areas are developed into topics shown as -01, -02, etc. and subdivided into Parts A, B,
etc. For example, HTM 06-02, Part A will represent: Electrical services – Safety – Low voltage.
HTM 08 covers specialist services as the examples shown in Table 2.
Table 2: Examples of specialist services
HTM 08-01 [Specialist services] Acoustics

HTM 08-06 [Specialist services] Pathology laboratory gas systems
Lifts are a specialist service and designated HTM 08-02: Lifts, 2016 Edition.
3 THE HISTORY OF HTM 08-02

The DH published HTM 2024 in 1995. It thus predated the Lifts Regulations 1997 and by 2008
was totally out of date with changes to legislation and the state of the art. It also comprised four
parts with a great deal of repetitive text and some parts were prescriptive and were not performance
based. Realising this, the DH commissioned this author to revise HTM 2024 into the new HTM
suite with the designation HTM 08-02.
HTM 08-02 was to comprise one part, be up to date, cover new technologies and be performance
based. As part of the suite of HTMs it was to link with them and any supporting HBNs. The
project started in September 2008 and was sent to the Central Office of Information (the then
government publishing house) in September 2009 and was published in February 2010. The new
HTM was thoroughly peer reviewed by both lift industry members from all areas of activity and
NHS representatives.
For some time this author has pointed out to the DH that the 2010 edition needed urgent revision.
Eventually the DH found some funds to engage an independent publishing house 2 to carry out a
partial revision of the 2010 edition concentrating mainly on the legislative and state of the art
changes. This author was engaged to do this.
The scope of HTM 08-02: 2016 is slightly changed from the 2010 edition and is:
1.1 This Health Technical Memorandum covers new lifts installed in healthcare buildings.
However, the recommendations in this Health Technical Memorandum can also be used as
guidance for the upgrading of the safety and performance of existing lifts.
1.2 It is assumed that equipment with the latest in lift safety technology is provided and that the
drive systems are either electric traction or electric hydraulic.
1.3 This Health Technical Memorandum does not specifically cover manually-operated lifts,
lifting platforms or stair lifts, escalators or moving walks, where specialist advice should be
sought (see also Appendices H and J). Escalators, moving walks, lifting platforms and stair
lifts come under the Supply of Machinery (Safety) Regulations 2008. Some guidance is given in
the provision of escalators in Appendix H. It is not anticipated that healthcare buildings will
2
The Government closed the COI in 2010.
contain architectural barriers requiring the provision of lifting platforms, stair lifts or platform
stair lifts. However some guidance is given in the provision of lifting platforms and platform
(wheelchair) stair lifts in Appendix J.
1.4 Neither does this HTM cover the movement of dangerous materials and gases in lifts. See
Health Technical Memorandum 02-01 – ‘Medical gas pipeline systems’ for guidance.
4 THE PROBLEM OF REVISING HTM 08-02 IN EARLY 2016

At May 2016:
 The Lifts Regulations had not been published to supersede those of 1997.
 The date of withdrawal of the BS EN 81-1/2 is set for 31 August 2017.
 The date of withdrawal of the supporting harmonised EN 81 family is set for 31 August
2018.
 There are projects in progress that must finish by 31 August 2017 under BS EN 81-1/2.
 There are new projects which need to meet BS EN 81-20/50 and not all equipment meets the
new standards.
At first this author tried to consider both main scenarios of EN 81-1/2 and EN 81-20/50. At peer
review3 this was suggested as the wrong approach. Thus at the final proofing stage it was decided
to meet the future not the past. The important note in the box below was incorporated into the
revised HTM 08-02.
Important note to users of this edition of HTM 08-02

Since the publication of HTM 08-02 in 2010, a number of major changes to European
Directives, UK Acts and Regulations and in particular to the BS EN 81 suite of standards
have occurred.
During the development of BS EN 81-20 and BS EN 81-50 as replacements to the long-
standing BS EN 81-1 and BS EN81-2, it was realised that the scale of changes being
required of manufacturers in their product ranges would necessitate a period of grace
where new designs could be developed, tested and certified. The final date of withdrawal
of BS EN 81-1 and BS EN 81-2 was set to be three years after the publication of BS EN
81-20 and BS EN 81- 50 – that is, on 31 August 2017. This means that for three years,
manufacturers may use either of these standards to build their products, but on 1 September
2017 the older standards will be fully withdrawn.
A consequence of the publication of BS EN 81-20:2014 and BS EN 81-50:2014 is that the
supporting harmonised standards and supporting unharmonised standards also need
revision. The timescale for the completion of this task is to be 31 August 2018.
Thus a state of flux exists in compliance to the Lifts Regulations. Users of this HTM must
be apprised of the critical dates and the consequences of not complying with the relevant
standards at the time a new lift is placed in service. Serious consequences can result from
overlooking this state of flux, especially when projects overrun the critical dates.
3
This author is deeply grateful to her industry colleagues listed in the Acknowledgements for their constructive
comments.
The major changes since the 2010 edition of HTM 08-02 are:
 This edition of HTM 08-02 reflects changes to the legal and standards requirements and
their effect on the presumption of conformity to the Lifts Regulations applicable when a lift
is put into service.
 References are made to the new BS EN 81-20/50 standards in place of the older BS EN 81-
1/2 standards.
 Restructuring of Chapter 2 to Statutory requirements and regulatory environment and
Chapter 3 to Professional roles and responsibilities.
 Inclusion of the latest BREEAM credit system.
 Deletion of the Appendix concerning energy-efficient designs and reference made to the BS
EN ISO 25745 series of standards.
 Revision of references.
A number of formatting and other editorial corrections were also made.
5 THE CONTENTS OF HTM 08-02: 2016

The contents of HTM 08-02: 2016 are similar to the 2010 edition and are shown in Table 3.
Table 3: Table of contents
PREFACE
Executive summary
Acknowledgements
Glossary of terms
1.0 Introduction
SECTION 1: MANAGEMENT POLICY
2.0 Statutory requirements and regulatory environment
3.0 Professional roles and responsibilities
4.0 Lift provision
SECTION 2: DESIGN CONSIDERATIONS
5.0 Lift planning
6.0 Lift equipment
SECTION 3: COMMISSIONING VALIDATION, CHECKS AND TESTING
7.0 Commissioning, validation and checks
8.0 Testing
SECTION 4: OPERATIONAL MANAGEMENT
9.0 Management of lift installations
10.0 Maintenance
11.0 Modernisation and upgrading
Appendix A – Relevant statutory regulations
Appendix B – Project stages according to BS 5655-6:2011
Appendix C – Supplement to BS 8486-1:2007 +A1:2011
Appendix D – Supplement to BS 8486-2:2007 + A1:2011
Appendix E – Typical instructions for the safe release of passengers trapped in electric traction lifts
Appendix F – Typical instructions for the safe release of passengers trapped in a hydraulic lift
Appendix G – Typical instructions for the safe release of passengers trapped in a machine-room-
less electric traction lift
Appendix H – Guidance in the provision of escalators
Appendix J – Guidance in the provision of lifting platforms and platform (wheelchair)
stair lifts
References
6 PARTICULAR CONSIDERATIONS IN THE HEALTHCARE ENVIRONMENT

6.1 Equipment considerations
Healthcare premises vary from a rural practice to a large high rise hospital. HTM 08-02 attempts to
cover this range.
It therefore suggests a lifting platform might be used in a rural two storey health centre where
access to the upper floor is occasionally required by mobility impaired persons (patients and staff).
An example of this is the Sedbergh Health Centre.
Alternatively, in a large busy hospital which has only three floors, the installation of escalators
(backed up by easily located lifts) might be a good solution. An example is the Whittington
Hospital, London.
6.2 Design considerations
There are specific requirements particular to the health care environment. Here are some examples,
without comment.
Example 1: Lighting
#46.131 The car should be illuminated to, at least, 100 Lux at floor level and on all control panels
using a method of illumination that will not cause sensory discomfort to those patients lying on a
trolley or bed (see also paragraphs 6.15–6.19).
Example 2: Medical assistance

#6.124 A trapdoor may need to be provided in the roof of the car:
• in the event of a trolley/stretcher/bed lift breaking down between floors; and
• where the floor-to-floor distance is too great to provide medical assistance (not rescue)
from a landing.
#6.125 The need for such a trapdoor should only be provided after a rigorous risk assessment. The
trapdoor should be held locked by a manual bolt accessible from the lift car roof, be interlocked
electrically with the lift machine and comply with BS EN 81-20.
Example 3: Special access controls

#6.146 Some healthcare buildings may require special lift control features to restrict access in
secure areas. It is normally sufficient to restrict access to the lift lobby. However, where high
security is required, for example in mental health wards, it may be necessary to provide special
facilities on the control panel to prevent unauthorised use of the lift. This can be achieved by
replacing landing-call pushes with key switches or swipe card-reader switches.
Example 4 Emergency bed service (code blue control)

#6.151 An emergency bed service (EBS) facility should be available in any lift that serves a theatre
area and is also available for general use. The facility should also be provided in emergency care
areas where the entrance level is above or below the reception.
4
The symbol "#" indicates the HTM 08-02: 2016 clause.
7 PROFESSIONAL ROLES AND RESPONSIBILITIES

The DH has a tightly defined hierarchy for estates management represented by Figure 3.
In both HTM 08-02: 2010 and 2016 there are number of roles specific to lifts.
Designated Persons (Lifts)

3.4 The Designated Person (Lifts) is an individual appointed by a healthcare organisation (a board
member or a person with responsibilities to the board) who has overall authority and responsibility
for lifts and their safe operation. They have a duty to prepare and issue a general policy statement
in relation to lifts and their safe operation, including the organisation and arrangements for
carrying out that policy. The policy should include reference to mandatory examinations, record-
keeping, emergency procedures and training of personnel.
Figure 3: Professional roles
Note: The Designated Person reports to the healthcare organisation's Board of Directors.
Authorising Engineer (Lifts)

3.8 The Authorising Engineer (Lifts) is a chartered engineer with appropriate experience, whose
appointment is the responsibility of the Designated Person (Lifts). The person appointed should
possess the necessary degree of independence from local management to take action within this
guidance including the implementation, administration and monitoring of the safety arrangements
defined in BS 7255.
Authorised Person (Lifts)

3.10 The Authorised Person (Lifts) is nominated by the Authorising Engineer (Lifts) and has the key
operational responsibility for the specialist service. The person will be qualified and sufficiently
experienced and skilled to fully operate the specialist service. The person nominated should be able
to demonstrate a thorough familiarisation with the system by having attended appropriate
professional courses. An important element of this role is the maintenance of records, quality of
service and maintenance of system safety (integrity).
Competent Person (Lifts)

3.14 A Competent Person5 (Lifts) is a person, suitably trained and qualified by knowledge and
practical experience, and provided with the necessary instructions to enable the required work to
be carried out safely (from BS 7255).
Lift Steward
3.16 A Lift Steward is a person nominated by the Authorised Person (Lifts) to undertake simple
daily monitoring of lifts in order to check their correct operation. See paragraphs 10.7– 10.13.
Lift Warden
3.17 Appointed by management, a Lift Warden will help to evacuate occupants during emergencies
by using an evacuation lift. There are three types of lift warden:
• Lift Warden (Floor);
• Lift Warden (Control); and
• Lift Warden (Car).
3.18 Training in the use of equipment will be by the Authorised Person (Lifts) and by the site Fire
Safety Adviser in relation to the emergency evacuation duties.
Lift Release Warden

3.19 A Lift Release Warden is a person, suitably trained and qualified by knowledge and practical
experience, and provided with the necessary instructions to enable the safe release of passengers
from lifts (see also paragraphs 9.15–9.34). They should be recommended by the Authorised Person
(Lifts), be formally appointed by management, and should undergo refresher training annually.
Note: If, under the terms of the maintenance contract, the release of trapped passengers is always
to be carried out by the lift maintenance contractor, rather than by in house staff, this post may not
be required.
8 CONCLUDING REMARKS
This revised edition of the Health Technical Memorandum – Lifts gives comprehensive advice and
guidance on the planning, design, installation, commissioning, testing, maintenance and operation
of new lifts (vertical transportation) in healthcare buildings. It also provides supporting information
that can be used in specifications for manufacturers, procurement contracts and the briefing of
design teams.
Although this Health Technical Memorandum is applicable to new installations, it can be used for
the upgrading and modernisation of existing installations, and is of use at various stages during the
inception, design, commissioning, testing and maintenance of lift services. It is intended to be read
by directors of estates and facilities, buildings services engineers, electrical and mechanical
engineers, facilities managers, architects, premises designers, consulting engineers, equipment
suppliers, equipment examiners, testers and maintainers. It can be used by bodies, organisations and
5
Not to be confused with the Competent Person under LOLER
individuals, who carry out the various duties indicated in this HTM for example when carried out
by outside contractors or under a Public Finance Initiative (PFI) contract.
ACKNOWLEDGEMENTS
To the commentators of the draft: Michael Bottomley, Roger Howkins, Ian Jones, Nick Mellor,
Vince Sharpe and many other industry members with whom the author has had discussions.
BIBLIOGRAPHY
Health Technical Memoranda are available from the UK Government’s website at:
https://www.gov.uk/government/collections/health-technical-memorandum-disinfectionand-sterilization
Health Building Notes are available from the same site at:
https://www.gov.uk/government/collections/health-building-notes-core-elements
ANNEX A: About Health Technical Memoranda and Health Building Notes

Health Technical Memoranda (HTMs) give comprehensive advice and guidance on the design,
installation and operation of specialised building and engineering technology used in the delivery of
healthcare. The focus of Health Technical Memorandum guidance remains on healthcare-specific
elements of standards, policies and up-to-date established best practice. They are applicable to new
and existing sites, and are for use at various stages during the whole building lifecycle.
Health Building Notes (HBNs) give best practice guidance on the design and planning of new health
care buildings and on the adaptation extension of existing facilities. They provide information to
support the briefing and design processes for individual projects in the NHS building programme.
All Health Technical Memoranda should be read in conjunction with the relevant parts of the
Health Building Note series.
BIOGRAPHICAL DETAILS
Dr Gina Barney is well known to the world-wide lift industry, owing to her many activities in the
field. Currently she is Principal of Gina Barney Associates, English Editor of Elevatori, Member of
the Chartered Institution of Building Services Engineers (CIBSE) Lifts Group Committee, Member
of the British Standards Institution (BSI) Lift Committees, UK expert to two International Standards
Organisation TC178/WG6 Traffic design and WG10 Energy efficiency of lifts and escalators.
Dr Barney has had a wide ranging career starting in the electronics industry, which eventually led to
the award of a doctorate on four quadrant thyristor power control of DC motors in 1965. After
many years in universities at Birmingham, UMIST and Manchester as lecturer, senior lecturer and
Director of Computer Networking, Dr Barney took early retirement in 1990 to concentrate on
consultancy.
Her first contact with the lift industry was in 1968, when she researched Ward-Leonard lift control
systems. Since then she has been active as a researcher, consultant, lecturer in the traffic design,
traffic control and circulation areas. These “soft” subjects have been complimented by “hard”
subjects of lift surveys, audits, contract supervision, safe release training, etc.
Dr Barney is the author of over 100 papers and is the author, co-author or editor of over 20 books
(not all on lifts). Her main activities currently are technical writing (she is a member of the Society
of Authors) with respect to standards and publications and various training courses. She is also a
Member of the Academy of Experts.
Dr Barney has the degrees of BSc, MSc and PhD and the professional qualifications of CEng, FIEE
and Eur.Ing. She was recently elected to an Honorary Fellowship of CIBSE for exceptional service
to the Institution and is a Freeman of the City of London and was recently admitted to be a
Liveryman of the Worshipful Company of Engineers.
Integration of Lift Systems into the Internet of Things and the Need
for an Open Standard Information Model
Jonathan Beebe
Jonathan Beebe Ltd., 2 Heap Bridge, Bury, BL9 7HR UK, [email protected]
Keywords: Internet of Things, smart lift/elevator services, open standards, information model,
semantic interoperability.
Abstract. The Internet of Things (IoT) is currently the subject of hype and is still in the process of
consolidation from a number of visions of its purpose and the benefits it will bring. This paper starts
with a review of the development and current status of, and motivations for, the IoT and continues
with a discussion of the potential for integrating lift systems into it. The conclusion is that a top-
level semantic layer for the IoT architecture is key to the successful delivery of so-called smart
building and smart urban services - particularly when machines talk to machines without human
intervention. It is at the semantic level that raw data is transformed into valuable and meaningful
information, and it is the semantic level that can unlock the imaginative potential to engineer a
smart urban environment in which lift systems play an important role. The new services will
inevitably require the exchange of information across disciplines, between different corporate as
well as private third-party agents and will highlight the importance of agreed standards upon which
systems from different suppliers can interoperate. The paper concludes with an overview of an open
standard information model for representing the semantics of lift (and escalator) operation which
could support this requirement.
1 INTRODUCTION
For the Internet of Things (IoT) to have generated as much excited discussion as it has already
received it must encompass more than simply a network of interconnected devices. Indeed, the term
brings to mind a vision which parallels the Internet as it currently exists, offering ubiquitous access
at any time to an enormous number and range of devices sharing information and cooperating in an
open flexible and ever-changing manner. The reality as it evolves may turn out to be less impressive
than the current hype would suggest, but it is apparent that success will depend on the development
of standards for interaction at all levels in the hierarchy of application software.
Following an overview of the IoT, this paper discusses some opportunities for integrating lift
systems to enable a mode of operation that is "smarter" and more integrated with the urban
environment. Innovative business models will be required, in addition to new technology, for this
level of integration to be achieved. The paper concludes with a discussion of standards for
information exchange within the domain of lift system operation.
2 DEFINITIONS AND OVERVIEW OF THE INTERNET OF THINGS

In common with many rapidly evolving computer technologies the definition of the Internet of
Things has a variety of interpretations which varies according to the vision of its developers.
Also, in common with other significant innovations in computation, the IoT is the combination of
several developments that have matured at the same auspicious moment. It is difficult to say which
is the cause and which the effect, but it is certainly true that the massive uptake of such innovations
is the result of rapid reduction in cost of ownership and that the cost of supplying products and
services is immensely reduced by its massive uptake.
The IoT is the subject of hype that can be characterised by the Gartner "Hype Cycle" graph [1] (see
Figure 1) and, as Haller[2] wrote, there are conflicting views of its correct definition. In August
2015 Gartner[3] published a "Hype Cycle" report on the IoT, marking several key areas of
application development as "On the rise" whereas key platform technologies are already at the peak,
with some early application areas already sliding into the "trough of disillusionment" possibly due
to the immaturity of those areas that are currently being "hyped".
Figure 1 Gartner Hype Cycle [1]

The Wikipedia definition of IoT[4] leads us to the "Overview of the Internet of Things" published
by International Telecommunication Union[5]
" A global infrastructure for the information society, enabling advanced services by
interconnecting (physical and virtual) things based on existing and evolving interoperable
information and communication technologies"
Other definitions[6] stress the ubiquitous nature of the Internet, its constant availability coupled
with the scope and variety of "things": anything, anywhere at any time. A more refined view is
offered by Uckelmann et al[7],
" … accurate and appropriate information may be accessed in the right quantity and
condition, at the right time and place at the right price. The Internet of Things is not
synonymous with ubiquitous/pervasive computing, the Internet Protocol (IP),
communication technology, embedded devices, its applications, the Internet of People or the
Intranet/Extranet of Things, yet it combines aspects and technologies of all of these
approaches."
Although the IoT is supported by a large number of technical frameworks and infrastructure with
clear and standard definitions, the above demonstrates the importance of the conceptual nature of
the IoT.
It is nothing new for pieces of equipment (e.g. the "machines" of Machine to Machine[8] or M2M)
to exchange status and control information, or that elements of an installation (e.g. the "things" of
IoT) are networked and can be addressed individually from multiple remote locations. So the value
of the IoT lies in its power in " enabling advanced services "[5].
The IoT results from the convergence of a number of technologies that have reached a level of
maturity. In this respect there is a parallel with the evolution of the "smart phone", which is the
convergence of mobile phone, portable computer, internet access, camera and GPS receiver, thereby
enabling a multitude of novel applications that require all these facilities to be combined in a
pocket-sized user device. So the camera is also a bar-code scanner, the screen is a map, etc. It is not
enough that the technology has been developed to a point where it can be mass-produced and so
becomes affordable. It is also necessary that a soft, conceptual, virtual evolution has also taken
Integration of Lift Systems into the Internet of Things and the Need for an Open Standard Information 4-3
Model
place whereby engineers have imagined and then implemented new capabilities that are only
possible by virtue of the interoperation of the new technologies.
Another parallel can be drawn with Wikipedia - we had encyclopaedia before, but the Internet has
given open access to readers who are now able to view and provide comments on the collected
knowledge of subject-domain editors in virtually every field of human understanding and that
knowledge and the comments on it are continually being reviewed by those editors. It is available
everywhere (in theory at least) and immediately. This universal accessibility and collective
contribution is another characteristic synonymous with the Internet.
From the several overviews of the IoT, Jayvardhana et al[9] discuss "A Vision, Architectural
Elements and Future Directions", Mazhelis et al[10] provide a commercial perspective, and the
TIVIT report[5] provides an Overview, Technical requirements and the IoT reference model.
Helpfully Atzori et al[11] summarize three "main visions" of the IoT:
• "Things-oriented" vision - focuses on the things’ identity and functionality
• "Internet-oriented" vision - emphasizes the role of the network infrastructure
• "Semantics-oriented" vision - focuses on systematic approaches towards representing,
organizing and storing, searching and exchanging the things-generated information
and it is the third, Semantics oriented, vision that this paper focuses on in relation to the integration
of lifts into the IoT because that is where the greatest value that is specific to the domain of lift
operation can be added and because doing so is key to fulfilling the requirement of the IoT in
"enabling advanced services". Although significant developments and technical challenges have
been overcome in the first two visions it is the semantic vision that is possibly the hardest to realise
but which offers the greatest potential for innovation.
Similarly, the IEEE[6](p7) paper, presents a model of "Technological and social aspects related to
IoT" and a three-tier architecture[6](p11) with an "Applications" top tier independent of the
supporting "Networking and Data Communications" tier and below that the "Sensing" tier. Again it
is the "Applications" tier where there is most scope to add value that is specific to the operation of
lift systems.
Ragget [12] of the W3C organisation looks at the challenges and the risk of fragmentation of the
IoT and proposes a "Web of Things"
• Things standing for physical and abstract entities
• Applications decoupled from underlying protocols
• Shared semantics and rich metadata
The "Applications" here will be specific to a business domain or discipline and the "Shared
Semantics" need to be described by a formal definition, such as a UML domain model[13].
Another feature of the current "human" Internet, which is taken for granted, is search facilities. In
support of this, services in the IoT must be discoverable so that devices which have never
communicated with them before can use these services and vice-versa that services can discover
devices. Datta et al[14] provide a categorization of the current landscape for the discovery of
resources and propose a framework.
3 LIFT SYSTEMS AND THE INTERNET OF THINGS

3.1 Current initiatives
There are some well publicised commercial initiatives, in the context of lifts and escalators, which
use Internet technology although they do not fulfil the IoT vision of semantic interoperability and
information sharing (this is by no means a comprehensive list):
3.1.1 Thyssenkrupp MAX
The Thyssenkrupp website[15] announces:
"… MAX analyzes real-time data from elevators around the world and provides our
maintenance control hub with a vast level of detail. This allows us to assess the health of
connected elevators and their components."
The resulting data is processed by software applications which "learn" to interpret the information
in terms of maintenance requirements rather than simply presenting it to a human operator in its
original form.
3.1.2 KONE
Kone also recently announced[16] a collaboration with IBM to gather and analyse trend information
from a global distributed network in an initiative to provide proactive maintenance. Significantly,
the announcement mentions[17] provision of an Application Programming Interface (API) for use
by application developers.
3.1.3 Otis Elevator Company Gen2 lift
In March 2016 Otis Elevator announced[18] services for Gen2 lifts:
• eCall™ mobile application developed exclusively by Otis, through which residents or
visitors to buildings can call and direct the elevator at a distance from their smartphones.
• eView™ offers building managers the opportunity to provide customized information
directly to passengers.
• Connected reporting and access in real-time equipment to performance data.
3.1.4 flexyPage displays

flexyPage[19] displays located in cars or on landings provide integrated status information on lift
status accompanied by information (graphics and text) relevant to the building as a whole, or
possibly the destination floors selected.
3.2 Some possibilities for the future
The above examples are early entries into a complex application environment that inevitably will be
complimented by others and which will evolve to provide greater sophistication, scope and
integration with other services that support the concept of smart buildings. The following are
outline suggestions for possible applications that exploit fully the visions of IoT, particularly that of
semantic interoperability allied with controlled accessibility for the public, crossing discipline
boundaries previously marked by "vertical silos"[20]. However, it must be recognised that
initiatives for such cross-fertilisation can only be developed in an open market:
3.2.1 Monitoring and Generating Twitter feeds
I would like to thank Beth Allan for tweeting:
" Beth Allan @adolwyn 1 Nov 2015
I should know better than to try & use the elevator in my building before 5pm on a weekend
at the beginning of a month. #movingday #longwait".
This information could be used very constructively if monitored by intelligent applications. Whilst
Model
not controlling the lifts directly, such applications would influence high-level policies of smart lift
control systems or smart buildings.
An important aspect of this example is the ability of a member of the public freely to provide
information which can lead to modification of the operation of the services that they use thus
fulfilling both the community participation and semantic-interoperability roles of the IoT.
To make such an application a reality would require the combination of very diverse business
capabilities and would no longer be practicable or commercially viable for a lift manufacturer or a
social media company to develop and maintain in isolation. Perhaps new business models will be
required, based on the sharing and trading of information.
3.2.2 Smart lifts cooperating with the smart buildings they serve.
Smart buildings interacting with smart cities[20] need to share information about the operation of
their lifts and also feed the lifts with external information such as calendars, weather conditions,
public transport availability and delays, to build a picture of likely passenger demand. However,
any such applications must remove all personal data from communications.
Hotels and commercial buildings might provide personalised "intelligent" and "pro-active" lift
service through apps running on personal mobile devices - suggesting destination floors based on
interests and affiliations of passengers, or their access permissions to different floors. At the
TEDxAmsterdam 2015 event, architect Ron Bakker[21] said of The Edge building in Amsterdam:
" It’s Monday morning. You enter your office building. The elevator is waiting for you and
knows on which floor you need to get off." ... "It might sound like fiction, but if you work at
the Edge this is your reality. "
This is in fact the Internet-enabled version of the vision offered by Barney and dos Santos[22] back
in 1985! It is a particularly useful example, because it emphasises the need for open interaction with
external information services that was not yet practicable in the vision of reality for "The lift for the
Year 2000" but now proffered by the visions of the IoT.
3.2.3 Integrated Energy Management
New business models might be established through lifts that predict estimates of their energy
requirements for the next 24 hours to the building, or directly to energy suppliers. Or conversely, it
might be the building that produces estimates based on a catalogue of energy usage reports acquired
from the lifts, given a knowledge of the factors likely to affect lift traffic demand profiles. In return
for such predictions, which would contribute to more efficient scheduling of energy provision, the
supplier could offer a cheaper tariff. Again, semantic interoperability of the IoT is a pre-requisite.
4 GENERIC STANDARDS
The IoT relies on many standards and standard services e.g. Wi-Fi, IPv6 addressing (required to
addresses every individual connected thing uniquely), http and web-service protocols, time-of-day
services, load-balancing, cloud computing, etc. Critical to the widespread adoption of IoT is a
standardised and reliable security infrastructure. As ETSI says on its website[23]
"Smart objects produce large volumes of data. This data needs to be managed, processed,
transferred and stored securely.
The use of standards
• ensures interoperable and cost-effective solutions
• opens up opportunities in new areas
• allows the market to reach its full potential
The more things are connected, the greater the security risk. So security standards are also
needed to protect the individuals, businesses and governments which will use the IoT."
5 DOMAIN- SPECIFIC STANDARDS

In order to implement applications for the IoT that are specific to a business domain we must be
able to understand the domain-specific information contained in any data that is generated and
transmitted via the network. This is a more significant requirement if there is a machine (rather
than a human who can make inferences and interpretations) in both the generating and the
consuming roles, which is often the case in the IoT. A lexicon or dictionary of the relevant terms
and the rules for their use is required to support such applications. A formal definition of the
information terms of the business domain and the relationships between those terms is described by
an ontology[24] or schema[25][26]. Schemas and ontologies are usually machine-readable and at
the same time perform the role of documentation to explain the terms to human readers.
There are already schemas and ontologies, which describe the information in a number of specific
domains, and libraries are published and supported by some standards organisations such as the
Smart Appliances Reference ontology (SAREF published by TNO[27]) for devices commonly used
in buildings eg a light switch[28] (see Figure 2 ) or an HVAC unit, though significantly this library
does not include a lift (or elevator).
Figure 2- SAREF: Extract from ontology for - Light Switch

5.1 Standards specific to the Lift Systems Domain
Some standards specific to lift systems do exist, and continue to be enhanced, which describe the
component parts - BIM (proposed)[29] - and the current state of the basic elements - CANopen-
Lift[30] - of a lift. However, there is a real need for an agreed and adopted formal schema to
describe the day-to-day operational characteristics of a system of lifts and this is a key requirement
for successful integration of lifts into the IoT.
Model
5.2 Standard Elevator Information Schema

The Standard Elevator Information Schema (SEIS)[31] was developed, by the author of this paper,
in 2003 from earlier work to model the information and operation of a lift system, using standard
software development tools and methods. While it certainly describes the data normally associated
with remote monitoring (car and landing call registration and cancellation, car movement, floor-
position, and door state) the SEIS is much more comprehensive, including elements of processed
information such as demand profiles, journey plans and energy consumption. This level of detail
enables the development of sophisticated applications and services, against a common standard,
which therefore do not have to concern themselves with, for example, the intricacies of individual
car control. The current state of development of IoT applications for lifts makes it thoroughly
appropriate that such a standard should now be widely adopted throughout the Building Services
industry.
SEIS is a formal specification of operational information of Lifts (and a similar schema exists for
Escalators). It describes the information entities, their interrelationships and the possible values that
may be assigned to each of their individual attributes. It is not a protocol - it doesn't describe a
sequence of messages.
The schema defines many complex data types which display a hierarchical structure. At the highest
level of granularity there are:
• Static data
• Dynamic data
• Events
• Exceptions
All dynamic data and events are time stamped so that network latency is not a problem and periods
of disconnection or non-availability of consumer services can be overcome simply by concatenating
information in a buffer until the relevant service is ready to receive it.
Another significant point of note is that the classes of information described in SEIS do not have
definitions of the "Services" which they might offer because it is felt to be too early to standardise
them. This work should be postponed for a subsequent revision of the standard after the current
version has become well used and understood.
SEIS is described in a formal language - an XML Schema Document (XSD) [26][32][33] - which
means that in addition to providing legible documentation for a human readership, the schema can
be referenced automatically by the devices (the "things") that are processing the information which
it describes. Thus it can be used by an application on a receiving device automatically to validate
incoming information that the sending device claims is conformant to the standard. The SEIS can
also be used by many commercially available software development tools automatically to generate
the software code for a complete class library to manipulate data conforming to the schema. For
example, the tool with which the schema itself was developed can generate such code in a variety of
languages (C++, C#, VB6, VB.Net or Java) but many other options are available from a range of
tools. This level of automated code generation can reduce significantly the cost of developing and
subsequently maintaining the applications that will underpin "smart" lifts connected to the IoT
whilst leaving application developers free to choose their preferred implementation language.
The scope of SEIS is deliberately constrained within the domain of lift operation in order to
minimise the risk of duplication (or contention) with other standards, for example:
• geospatial information properties,

• access authorisation
• security (for example see IET Cyber Security Consortium report[34])
5.2.1 Building Information Model

Currently, it appears to be possible to integrate SEIS with the requirements of BIM for lifts as
defined in the CIBSE Elevator Product Description Template[29]. There is no benefit in re-
inventing or competing with this information model nor any intention to, but certainly it could be
very beneficial to align and integrate these two standards. However, an open BIM library for lifts is
needed at least as a base for manufacturers to produce their own specialisations. Resources to
accomplish this can be found on the OPENBIM[35] and buildingSMART[36] websites and tools
such as DigiPara[37] could be used to deliver the resulting library to lift system designers.
5.2.2 CANopen-Lift
CANopen-Lift concentrates on monitoring and controlling the state of individual standard
components of a lift and so, in a similar manner to BIM, represents a very complimentary model to
SEIS. To this end, it could be overlaid with a schema (XSD).
6 CONCLUSIONS
The IoT is a much "hyped" and rapidly developing environment offering enormous scope for new
joined-up applications and services in a parallel of those offered so far by mobile technologies and
the Internet in its current form. It embodies a vision of integrated applications sharing information
with an understanding of its semantics.
Key to the success of the IoT is the availability of standards and in particular domain-specific
standards that will enable applications from different suppliers to interact in novel ways.
The SEIS is an enabler for new applications and services in the lift systems domain to be integrated
within the IoT. It offers a standard model to developers of services and applications for the IoT
platform, regardless of the underlying networking and physical technologies used. It embodies a
level of "expertise" which therefore doesn't have to be developed and, more significantly, to be
maintained by individual suppliers.
REFERENCES
[1] Gartner, Gartner Hype Cycle graph, Jeremykemp at English Wikipedia [GFDL
(http://www.gnu.org/copyleft/fdl.html) or CC BY-SA 3.0
(http://creativecommons.org/licenses/by-sa/3.0)], via Wikimedia Commons;
https://en.wikipedia.org/wiki/Hype_cycle (retrieved 11/04/2016).
[2] S. Haller, SAP Research Centre, Zurich The Things in the Internet of Things, http://www.iot-
a.eu/public/news/resources/TheThingsintheInternetofThings_SH.pdf (retrieved 17/04/2016);
presented at Internet of Things 2010 Conference, Tokyo, Japan.
[3] Gartner, Gartner's 2015 Hype Cycle for Emerging Technologies Identifies the Computing
Innovations That Organizations Should Monitor, Gartner.com, Stanford, USA, 18-08-
2015http://www.gartner.com/newsroom/id/3114217 (retrieved 18/04/2016).
[4] Wikipedia, Internet of Things, https://en.wikipedia.org/wiki/Internet_of_Things (retrieved
12/04/2016).
Model
[5] International Telecommunication Union (Telecommunication Standardization Bureau), p1 sec

3.2.2, Overview of the Internet of Things, 06/2012,
http://www.itu.int/ITU-T/recommendations/rec.aspx?rec=y.2060 (retrieved 12/04/2016).
[6] IEEE, Towards a definition of the Internet of Things, pp7,11,17
http://iot.ieee.org/images/files/pdf/IEEE_IoT_Towards_Definition_Internet_of_Things_Issue1_
14MAY15.pdf (retrieved 12/04/2016).
[7] Uckelmann, Dieter, Mark Harrison, and Florian Michahelles. "Architecting the Internet of
Things." Springer (2011) preview available at http://link.springer.com/book/10.1007/978-3-642-
19157-2
[8] Wikipedia, "Machine to Machine", https://en.wikipedia.org/wiki/Machine_to_machine,
(retrieved 11/04/2016).
[9] Jayavardhana Gubbi, Rajkumar Buyya, Slaven Marusic, Marimuthu Palaniswami, "Internet of
Things (IoT): A Vision, Architectural Elements, and Future Directions", Department of
Electrical and Electronic Engineering, The University of Melbourne, Vic -3010, Australia;
http://arxiv.org/ftp/arxiv/papers/1207/1207.0203.pdf (retrieved 13/04/2016)
[10] O.Mazhelis et al, Internet-of-Things Market, Value Networks, and Business Models: State of
the Art Report, TIVIT Internet of Things Programme, Jyväskylä University Printing House,
Jyväskylä, Finland, 2013 ISBN 978-951-39-5251-8,
http://www.internetofthings.fi/extras/IoTSOTAReport2013.pdf (retrieved 19/04/2016).
[11] Luigi Atzori, Antonio Iera, Giacomo Morabito, The Internet of Things: A survey, 1389-1286
Computer Networks, Volume 54, Issue 15, 28 October 2010/$ Elsevier B.V.
doi:10.1016/j.comnet.2010.05.010, http://dx.doi.org/10.1016/j.comnet.2010.05.010 (retrieved
11/04/2016).
[12] Dave Raggett, W3C, Using semantics and rich metadata to bridge IoT silos — W3C’s work
on the Web of Things, Presentation to ETSI M2M Workshop, Dec 2015,
https://docbox.etsi.org/Workshop/2015/201512_M2MWORKSHOP/S03_STANDARDSandAL
LIANCES/W3C_RAGGETT.pdf (retrieved 12/04/2016)
[13] James Rumbaugh, Ivar Jacobson, Grady Booch, The unified modelling language reference
manual,. p. cm. The Addison-Wesley object technology series; 1999; ISBN 0-201-30998-X
[14] Soumya Kanti Datta, et al, " Resource Discovery in Internet of Things: Current Trends and
Future Standardization Aspects", Mobile Communications Department, Eurecom Biot, France
http://www.eurecom.fr/en/publication/4716/download/cm-publi-4716.pdf (retrieved
22/04/2016).
[15] Thyssenkrupp, "MAX, the elevator industry’s first real-time, cloud-based predictive
maintenance solution", corporate website, https://max.thyssenkrupp-elevator.com/en/ (retrieved
22/04/2016).
[16] KONE Corporate Website, "KONE joins forces with IBM for IoT cloud services and
advanced analytics technologies", 19/02/2016, http://www.kone.com/en/press/press/kone-joins-
forces-with-ibm-for-iot-cloud-services-and-advanced-analytics-technologies-2016-02-19.aspx
(retrieved 10/05/2016).
[17] Internet of Business website, "Kone partners with IBM IoT for smarter buildings",
04/03/2016, Vinelake, http://www.internetofbusiness.co.uk/insight/2016/03/04/kone-partners-
with-ibm-iot-for-smarter-buildings/(retrieved 10/05/2016).
[18] Otis Elevator Corporate Website, http://www.otisworldwide.com/b2a-
newsrec.html?item=20160314&linkId=22259429 (retrieved 06/05/2016).
[19] Flexipage Corporate Website, http://flexypage.de/en/displays-hotels (retrieved 22/04/2016).

[20] CEN-CENELEC-ETSI, "Smart and Sustainable Cities and Communities' Co-ordination
Group(SSCC-CG) Final Report", http://www.etsi.org/images/files/SSCC-CG_Final_Report-
recommendations_Jan_2015.pdf (retrieved 22/04/2016).
[21] TEDx, " TEDxAms speaks to Ron Bakker, architect of The Edge, the world’s most
sustainable office building ", TEDx Amsterdam, 27-11-2015,
http://tedx.amsterdam/2015/07/ron-bakker-on-the-edge/ (retrieved 22/04/2016).
[22] G. C. Barney, S.M. dos Santos, “Elevator Traffic Analysis Design and Control”, Peter
Peregrinus Ltd., London UK, Revised second edition 1985, ISBN 0-86341-042-1,1985.
[23] ETSI website, "Internet of Things", http://www.etsi.org/technologies-
clusters/technologies/internet-of-things (retrieved 22/04/2016).
[24] Wikipedia, "Ontology", https://en.wikipedia.org/wiki/Ontology_(information_science
(retrieved 13/04/2016)
[25] Wikipedia, "Database schema", https://en.wikipedia.org/wiki/Database_schema, (retrieved
13/04/2016)
[26] W3C (World Wide Web Consortium), "What is an XML Schema",
http://www.w3.org/standards/xml/schema, (retrieved 16/04/2016)
[27] TNO, "Smart Appliances Reference ontology " Netherlands Organisation for Applied
Scientific, Research, https://www.tno.nl/en/about-tno/organisation/. (retrieved 10/04/2016)
[28] TNO, "Smart Appliances REFerence (SAREF) ontology",
http://ontology.tno.nl/saref/saref_LightSwitch.html (retrieved 10/04/2016).
[29] CIBSE, "CIBSE Elevator Product Description Template", Status: proposed,
http://www.cibse.org/knowledge/pdts/elevator-product-data-template-(pdt) (unavailable for
download 10/04/2016).
[30] CANopen, "Application profile for lift controls", CAN in Automation (CiA) e. V. DE-90429
Nuremberg, Germany, http://www.can-
cia.org/index.php?eID=tx_nawsecuredl&u=2971&g=11&t=1461410101&hash=d59d39f38981e
11de66f92c54ba86ea8dcffec17&file=fileadmin/resources/documents/groups/417_4v02000002.
pdf (retrieved 22/04/2016).
[31] Jonathan Beebe, "Standard Elevator Information Schema", http://www.std4lift.info/
(retrieved 17/04/2016).
[32] W3C, "W3C XML Schema Definition Language (XSD) 1.1 Part 1: Structures",
http://www.w3.org/TR/2012/REC-xmlschema11-1-20120405/, (retrieved 16/04/2016)
[33] W3C, "W3C XML Schema Definition Language (XSD) 1.1 Part 2: Datatypes",
http://www.w3.org/TR/2012/REC-xmlschema11-2-20120405/, (retrieved 16/04/2016)
[34] IET, "CyberSecurityAndBIM-bim insight", http://www.theiet.org/sectors/built-
environment/design/documents/bim-cyber-security.cfm?type=pdf
[35] OPEN BIM The Open Toolbox for BIM, http://www.openbim.org/ (retrieved 17/04/2016).
[36] BuildingSMART International home of openBIM, http://www.buildingsmart.org/.
[37] DigiPara, Corporate website, www.digipara.com/ (retrieved 17/04/2016).
Model
Jonathan Beebe graduated in Electronics and was awarded a Ph.D for his research into the use of
computers in the management of lift systems. Subsequently, he was employed to design and
implement software for dispatcher algorithms, single car controllers and also for remote
performance monitoring of lift systems. He has continued this work throughout his career as a part-
time consultant along with an active interest in research. Dr Beebe also spent 25 years in full-time
employment as a software Design Authority developing large applications for government
departments, banks and other financial institutions using Internet technologies. Much of this work
involved modelling the processes and information of clients' businesses as well as designing
methods and tools for the software development process itself.
Your Lift Journey – How Long Will You Wait?
Caroline Bird1, Richard Peters1, Elizabeth Evans1, Stefan Gerstenmeyer2,
Peters Research Ltd
1
Bridge House, Station Approach, Great Missenden, Bucks, HP16 9AZ, UK

[email protected]
thyssenkrupp Elevator Innovation GmbH,

2
Bernhäuser Str. 45, 73765 Neuhausen, Germany

[email protected]
Keywords: quality of service, waiting time, lift journey, elevator, lift, survey, questionnaire,
dispatching algorithm
Abstract. When passengers start their lift journey they initially wait for their call to be answered.
Whilst travelling to their destination their trip is often interrupted by intermediate stops which are the
result of other passengers’ calls. Dispatching algorithms optimize the handling capacity and quality
of service of lift groups. The main criteria for quality of service is currently average passenger waiting
time. Travel, overall journey time and number of stops are additional criteria. But which is the most
important for passengers when they think about their trip? How can dispatching algorithms be
improved and tailored to meet passengers’ expectations? An online questionnaire has been conducted
asking people how they feel while using lifts and to help identify what passengers want and expect.
The questions and results from the survey are presented and it is shown how the results can be applied
to existing dispatching algorithms.
1 INTRODUCTION
The main criterion used to assess quality of service in the lift industry is average passenger waiting
time. Many research papers show that long periods of waiting for a product or service lead to
increased customer dissatisfaction [1, 2]. However, there has been very little empirical research into
the experience of waiting for and travelling in lifts.
When thinking about waiting in the context of lift-usage it important not to think of it in isolation, but
in conjunction with the overall lift journey. Most situations in which waiting is required tend to have
a period of waiting for a product or service followed by the end goal, being the receipt of that product
or service. Lift usage is relatively unique in that is comprises a waiting stage, an in-between period
of travelling in the lift, and then the end goal of reaching your destination. One of Maister’s key
principles [3] is that occupied time feels shorter than unoccupied time. In the lift context, this suggests
that travelling time feels shorter than waiting time [4].
Dissatisfaction associated with waiting for a lift could also be due to the anxiety experienced. There
are few indicators that show how long the passenger will wait before the lift arrives. Once the lift
arrives, anxiety reduces as the passenger is on the way to his or her destination [5].
An important consideration in the psychology of waiting is that “waits must be appropriate” both in
cause and duration [6]. Customers are happy to wait for an appropriate amount of time, taking into
account the situational factors. There is a limit to the amount of time passengers will wait and travel
before they become impatient, which is dependent on individual factors [7]. Waiting is an inevitable
part of lift-usage, and customers are prepared to accept these inevitable waits, for a reasonable period
of time. An additional benefit to waiting is that customers often deem a product or service to have an
increased value if they have had to wait for it [8]. However, if a wait extends beyond what is deemed
reasonable in the eyes of the customer, customer dissatisfaction will greatly increase.
There is a very wide range of factors that could be studied in regards to the psychology of waiting in
the context of lifts. This paper focuses specifically on how passengers’ waiting time preference is
affected by overall journey times and intermediate stops.
The definition of waiting time and other stages of the passenger journey are defined by CIBSE Guide
D [9]. However, for interaction with the general public, less precise definitions are required as there
is no possibility of communicating the full engineering definitions. Consequently, for this research
and throughout this paper simplified terms are used. Wait Time (WT) refers to the time from when
the passenger enters the lobby until their lift has arrived. Travel Time (TT) refers to any time from
when the passenger enters the lift, until they arrive at their floor, including any Intermediate Stop
(IS). An IS occurs when the lift is stopped for other passengers to enter or exit the lift. Journey Time
(JT) is the sum of WT and TT.
Smith and Gerstenmeyer [4] posed the scenario in which a dispatcher had two journey options, both
had identical overall JTs, but differing WTs and TTs. They concluded that the journey with the shorter
WT mostly would be the preferred option. Smith and Gerstenmeyer questioned whether participants
would be willing to make trade-offs between WT and TT. For example, would passengers prefer a
reduced overall JT at the expense of an increased WT?
The lift journey is often considered as comprising two parts: waiting and travelling, with travelling
deemed as an occupied, anxiety-free portion of the journey. Barney and Dos Santos [10] observed
that with increasing number of ISs there is an increased level of frustration in passengers. This
suggests that the travel portion of a lift journey ought to be considered in two parts, transit and stops,
accounting for their different psychological impact.
The research used an online survey to reach the maximum number of participants with limited
difficulty on the part of the participant. Animations were used throughout in order to present a realistic
representation of the different parts of the lift journey, whilst removing the bias that would be present
in live footage. As a principle of waiting is that emotions dominate [6], the survey questions were
designed to include emotive language, asking participants to focus on how they would feel
experiencing a particular journey.
This survey presented participants with hypothetical journey scenarios, aiming to answer two main
questions: (i) How do people feel about different parts of the lift journey? (ii) What is the preferred
period of time to wait for a lift, and how does this differ depending on the factors of overall JT and
subsequent ISs?
The research may be applied to improve the design of dispatching algorithms to account for Quality
of Service (QoS) from the passengers’ prospective.
2 METHODS
2.1 Participants
The use of an online survey ensured a quick data collection process. Data was collected anonymously
from 278 participants, no personal details were collected. Participants were an opportunity sample
who either received a link to the survey directly by email, or were made aware of the survey on social
media.
2.2 Materials and Procedure
The survey consisted of five questions in total, with the option to exit the survey after the first
question. This was included to allow participants that were no longer interested in completing the
survey to exit easily, whilst still allowing the retention of their data from the first question.
Your Lift Journey – How Long Will You Wait? 5-3
Question 1. The first question had four parts to it. Each part included a basic animation depicting a
particular aspect of a lift journey (waiting, travelling, first IS, second IS), along with four multiple
choice response options, see Figure 1. Participants were asked how they felt at each part of the lift
journey. Response options were four animated faces, from green smiley face, yellow semi-smiley
face, orange semi-sad face and red sad face, scored 1-4 respectively. The aim of this question was to
ascertain an easily comparable rating system for how happy participants are to experience each aspect
of the lift journey, relative to other aspects.
How do you
feel waiting for
your lift?
Figure 1 Presentation of Question 1 part 1 with multiple choice response options.

Question 2. This question aimed to assess participants’ preferred WT. Participants were provided
with three theoretical journeys. All journeys had the same overall JT, but with different WTs and
TTs, see Table 1. The options were presented in the format given in Figure 2.
Table 1 Question 2 hypothetical journey options
WT (s) TT (s) JT (s)
15 45 60
30 30 60
2 58 60
Figure 2 Format of hypothetical journeys in Questions 2 and 3

Question 3. Participants were provided with a different version of Question 3, depending on their
answer to Question 2. Participants were presented with two theoretical journeys in the same format
as the previous question. The first journey was identical to their answer to Question 2, the second
hypothetical journey consisted of a longer WT, but a shorter overall JT, see Table 2. We knew from
Question 2 that participants were happy to wait for their chosen length of time. By including this
manipulation, it allowed an analysis of whether participants are willing to increase their WT in order
to achieve a shorter JT.
Table 2 Hypothetical journey options in Question 3
Question 2 answer Question 3 possible answers
WT (s) TT (s) JT (s) WT (s) TT (s) JT (s)
15 45 60
15 45 60
30 20 50
30 30 60
30 30 60
40 10 50
2 58 60
2 58 60
15 35 50
Question 4. This question considered the effect of ISs on preferred WTs. Participants were given
four journey options with varying WTs and number of ISs, see Table 3.
Table 3 Hypothetical journey options in Question 4, with times presented in seconds

WT (s) TT (s)/ISs JT (s)
40 20/0 60
30 30/1 60
15 45/2 60
2 58/4 60
Figure 3 Format of hypothetical journeys in Question 4

Question 5. This final question was another assessment of WT versus TT using a Likert scale with
considering ‘relative lengths’ rather than providing participants with specific timings for a theoretical
journey. See Figure 4 for a presentation of the possible answers. The results of this question provide
a specific ratio of preferred WT to TT.
Figure 4 Format of possible answers to Question 5
3 ANALYSIS AND RESULTS

3.1 Question 1
The independent variable was ‘part of the journey’, with four levels: waiting, travelling, first IS, and
second IS. The dependent variables were the smiley face ratings scored 1-4 (happy-sad). A Friedman
test was used due to the repeated measures design and non-parametric ordinal data. This was
supplemented with Bonferroni-corrected Wilcoxon Signed-rank tests.
The Friedman test found that there was a significant difference in the how people felt about the
different parts of the journey, χ2(3) = 453.11, p < 0.001, see Figure 5. Travelling in the lift received
the lowest rating, followed by waiting, first IS and second IS.
Post-hoc Wilcoxon Signed Rank tests (α=.008) showed a significant difference between all
conditions, see Table 4.
Median score 2
0
Waiting Travelling 1st Stop 2nd Stop
Figure 5 Median happiness score for each part of the journey
Table 4 Post-hoc Wilcoxon Signed Rank tests
Waiting Travelling 1st Stop
Travelling T = 69.00, p <0.001
1st Stop T = 71.38, p = 0.975 T = 82.00, p < 0.001
2nd Stop T = 84.75, p < 0.001 T = 52.00, p < 0.001 T = 79.63, p < 0.001
3.2 Question 2
There were three possible answers to this question, 15 second WT, 30 second WT and 2 second WT.
Therefore, a one-sample t-test testing for a significant difference in the journey chosen was used.
There was a significant difference in the journey chosen (t(256) = 39.69, p < 0.001, M = 1.69), with
significantly fewer people choosing a 2 second WT (13%), and the same number of people opting for
both a 15 second WT (44%) and 30 second WT (44%).
3.3 Question 3
There were three different versions of Question 3 depending on the response participants gave to
Question 2. One sample t-tests were used for each of the versions.
Question 3a. Participants could opt to stay with a 15 second WT but overall 60 second JT, or change
their preferred lift journey to 30 second WT but overall 50 second JT. Significantly more participants
opted to stay with a 15 second WT (67.57%) than change to a 30 second WT (32.43%; t(110) = 29.67,
p < 0.001).
Question 3b. Participants could stay with a 30 second WT and 60 second JT, or change their preferred
lift journey to 40 second TT and 50 second JT. Significantly more participants stayed with the same
JT with a 30 second WT (52.34%) than changed to a 40 second WT (47.66%; t(110) = 31.02, p <
0.001).
Question 3c. Participants were presented with the option of staying with a 2 second WT and 60
second JT or choosing a longer WT of 15 seconds, but shorter 50 second JT. Significantly more
participants opted to change to a 15 second WT (53.33%) than stay with a 2 second WT (46.67%;
t(30) = 17.04, p < 0.001).
80
Number of participants 70
60
50
40 Same JT
30 Shorter JT
20
10
0
15 sec WT 30 sec WT 2 sec WT
Question 2 Answer
Figure 6 Number of participants who chose each journey dependent on WT preference in
Question 2
3.4 Question 4
In order to ascertain the relative impact of number of ISs on WT, answers from Question 2 became
the independent variable in this analysis. This therefore contained three levels, 15 second WT, 30
second WT and 2 second WT. The dependent variable was the number of stops chosen from 0 stops
to 4 stops, scored 1 to 4 respectively. Due to the independent measures design and non-parametric
ordinal data a Kruskal Willis test was used.
The Kruskal Willis test found a significant difference in preference for ISs in participants journeys,
dependent on their preferred WTs as ascertained in Question 2: χ2(2) = 20.20, p < 0.001, see Figure
7. Participants who chose in Question 2 to have either a 15 second or 30 second WT opted for a
journey with a 30 second WT but only 1 IS. Participants who chose a 2 second WT in Question 2
opted for a journey with a 15 second WT and 2 ISs.
3.5
3
Median Score
2.5
2
1.5
1
0.5
0
15 sec WT 30 sec WT 2 sec WT
Question 2 answer
Figure 7 Median score of ISs preference dependent on WT preference in Question 2
3.5 Question 5
This question consisted of a Likert scale with 9 response options ranging from: less time waiting than
travelling (scored 1), through to equal time waiting and travelling (scored 5), through to more time
waiting than travelling (scored 9). Participants’ data provided a mean score of 3.96.
4 DISCUSSION
4.1 Q1: feelings at different parts of the journey?
Participants gave travelling the highest happiness rating, followed by waiting, which was one step
down on the scale, but still indicated a positive feeling. This supports previous research that travelling
is more desirable than waiting due to the reduction in anxiety once the passenger is in the lift [4, 5].
It also supports Norman’s [6] principle, that if waits are appropriate and reasonable passengers are
happy to experience them.
Participants were less happy experiencing ISs than waiting, indicating a negative feeling for both ISs.
The negative feeling increased from the first IS, to the second IS. These findings are able to provide
empirical support for the observations made by Barney and Dos Santos [10] that passenger’s
frustration level increased with the number of ISs they experienced. In dispatching decisions,
Maister’s principle that “occupied time feels shorter than unoccupied time” should not be interpreted
to assume that the delay associated with ISs is occupied time.
4.2 Q2: Preferred WT?

Participants were initially presented with three different hypothetical journeys. A small group of 13%
of the participants opted for a journey with only a 2 second WT, with the rest of the participants were
split evenly between choosing the 15 second and the 30 second WT. The most interesting point of
note from these results is the varied responses, with no particular preference shown for one WT over
another. This supports the idea that “waits ought to be appropriate” [6], but suggests that different
people having different interpretations of “appropriate”.
4.3 Q3: Wait longer for shorter arrival?
The aim of this question was to assess whether the benefit of a shorter overall JT was good enough
to warrant an increase in the more negatively perceived WT. The survey found that the two groups
of participants who chose previously to have a WT of either 15 seconds or 30 seconds, opted to stay
with their original choice, rather than chose the alternative shorter journey. In contrast, of the small
group of participants who originally chose a 2 second WT, a significant, but small majority of
participants chose instead to have a shorter overall JT at the expense of an increase WT. For
participants who had already chosen a substantial WT of either 15 or 30 seconds, the benefit of a
reduced JT was not enough to encourage them to endure an increased, negative waiting period.
However, for participants who had originally chosen only a 2 second WT, an increase in that WT still
provided those participants with an “appropriate” WT, with the added benefit of a reduced JT.
4.4 Q4: Intermediate stops?
Participants who initially chose (prior to the follow-up JT question) to have a WT of 30 seconds went
on to choose an IS journey with 1 stop, and a 30 second WT. Participants who had chosen a WT of
15 seconds in Question 2 now increased this to a WT of 30 seconds, and only 1 IS, as opposed to 2
ISs if they remained with a 15 second WT. A reduction in overall JT was not incentive enough to
encourage this group of participants to increase their WT, whereas a reduction in the number of ISs
was. This suggests that ISs provide a more negative experience than waiting.
Participants who originally chose a WT of 2 seconds chose to increase their WT to 15 seconds in
order to reduce the number of ISs from 4 to 2. This affirms than passengers’ preferences are not best
served by simply considering WT.
4.5 Q5: Waiting time versus travelling time

The structure of this question allowed for proportional data to be collected on WT preferences, as
opposed to the fixed timings seen in the previous questions. Previous work suggested a rule of thumb
that WTs could be a quarter of the total journey [11]. The results from this question suggest that
passengers would prefer nearer a third of overall JT waiting. This is also reflected in the results from
the other questions in which 44% of participants preferred to spend a quarter of the time waiting, and
44% preferred to spend half of the time waiting.
4.6 Limitations
Whilst steps were taken to ensure the validity of results, all surveys have limitations.
The first limitation is in regards to the ecological validity of the survey. Despite the use of animation
to create a virtual lift journey experience for the participants, and the use of emotive language to
encourage the correct frame of mind for the participants, answering questions on the survey is not the
same as experiencing an actual lift journey.
Levinson [12] carried out a study on waiting and travelling in the context of driving. Half of the
participants were given a set of scenarios via a computer survey format, and the other half experienced
the scenarios in a driving simulator. Levinson found significantly different results depending on the
condition the participants were in. Other research has also found that time is often perceived as
passing differently to the actual passage of time [2] and that the extent to which perceived time differs
from reality is conditional on whether the participant is waiting or travelling [13].
The second limitation of this study is the applicability of the results to dispatch operations with
customer facing dispatch systems in their current form. In this study participants were presented with
all lift journey options and then allowed time to make the decision about which journey they would
prefer to take. At present, this is not an option afforded to lift passengers. Therefore, whilst a
passenger may prefer a journey with a longer WT, because it is a direct journey with no ISs, the
passenger would not know this information at the time of waiting for their lift. For that passenger
they simply experience an increased WT, without explanation, and therefore would be experiencing
negative emotions and a poorer quality of service [5]. This is seen in a study by Van Houton [14] in
which a lift was programmed to have an increased door delay extending both the waiting and
travelling times of the lift journey. They found that when the door delay was increased, significantly
more participants opted to take the stairs instead of experience the unexplained increased wait.
Providing passengers with information regarding their WT could not only reduce confusion, but also
increase the perceived quality of the service [6].
5 QUALITY OF SERVICE DISPATCHING

The principle of considering passengers’ perception is already built into some dispatch algorithms. A
combination of Quality of Service (QoS) criteria can be considered by weighting individual
components to create a pain index [11]. This provides the basis for QoS dispatching, optimizing
dispatching choices based on our best understanding of passenger preferences.
Waits ought to be appropriate [6]. However, the survey responses affirm that WT alone should not
be used as a proxy for QoS.
This research provides empirical data on which to base the weighting of WT and TT. TT includes
periods of unoccupied time during ISs. To improve perceived performance with QoS dispatching,
the passenger delay associated with ISs needs to be accounted for separately.
Whilst this paper has studied participant’s feelings towards ISs, similar studies and assumptions could
also be made regarding other lift journey delays. For example:
1. False stops which occur when passengers endure an IS without a passenger joining or
alighting the lift.
2. In systems with more than one cabin or lift per shaft, there are periods when passengers
experience delay while other unseen passengers are joining or alighting another cabin or lift
[4].
3. In cases where the best dispatching compromise is to load a passenger while travelling in the
opposite direction to their final destination, resulting in a reverse journey [15].
QoS dispatching will benefit from the best possible communication with passengers. An unexplained
pause in lift operation or an unexpected reverse journey leads to confusion and mistrust.
Current dispatchers do not reflect the wide range of preferences reflected in the survey results. It is
conceivable that personal preferences could be collected with a smart phone app and accounted for
in dispatching decisions. The app could also support supplementary communication with the
passenger.
6 CONCLUSION
The findings in this research confirm the principle of appropriate WTs [6], but suggest that the
concept of appropriate may vary greatly for different passengers. They also suggest that the travelling
part of a lift journey should not be considered as one section alone, but instead as a section containing
multiple parts, to account of the unoccupied times of stopping interspersed within the travelling. The
reduction in ISs is more motivating than a reduction in overall JT to increase WTs.
The application of the research through Quality of Service (QoS) dispatching has been discussed. In
addition to choosing an appropriate optimisation function, the best possible communication with
passengers has been highlighted as a priority in order to improve passenger perception.
Areas for future research include evaluating how people feel about other lift journey delays such as
false stops, pauses in operation due to another cabin or car sharing the same shaft, and reverse
journeys.
REFERENCES
[1] Zhou, R. and Soman, D. (2008) Consumers' waiting in queues: The role of first-order and
second-order justice. Psychology and Marketing. 25 (3), 262-279.
[2] Houston, M. B., Bettencourt, L. A. and Wenger, S. (1998) The relationship between waiting
in a service queue and evaluations of service quality: A field theory perspective. Psychology and
Marketing. 15 (8), 735-753.
[3] Maister, D. (1985) The Psychology of Waiting Lines [online]. Available from:
http://davidmaister.com/articles/the-psychology-of-waiting-lines/ [Accessed 02/12, 2014].
[4] Smith, R. and Gerstenmeyer, S. (2013) A review of Waiting Time, Journey Time and Quality
of Service. In: Symposium on Lift and Escalator Technologies. Northampton:
[5] Hui, M. K. and Tse, D. K. (1996) What to Tell Consumers in Waits of Different Lengths: An
Integrative Model of Service Evaluation. Journal of Marketing. 60 (2), 81-90.
[6] Norman, D. (2008) The Psychology of Waiting Lines [online]. Available from:
http://www.jnd.org/dn.mss/the_psychology_of_waiting_lines.html [Accessed 05/07, 2014].
[7] Strakosch, G. and Caporale, R. (2010) The Vertical Transportation Handbook, Fourth
Edition. Hoboken; New Jersey: John Wiley & Sons, Inc.
[8] Koo, M. and Fishbach, A. (2010) A Silver Lining of Standing in Line: Queuing Increases
Value of Products. Journal of Marketing Research. 47 (4), 713-724.
[9] CIBSE (2015) CIBSE Guide D: 2015 Transportation systems in buildings. London: The
Chartered Institution of Building Services Engineers.
[10] Barney, G. and Dos Santos, S. (1977) Lift Traffic Analysis Design and Control. London:
Pereginus.
[11] Smith, R. and Peters, R. (2004) Enhancements to the ETD Dispatcher Algorithm. In:
Elevator Technology 14, Proceedings of Elevcon 2004. The International Association of Elevator
Engineers.
[12] Levinson, D., Harder, K., Bloomfield, J., Winiarczyk, K., Levinson, D. D., Engineering, D.
O. C. and Harder, K. (2004) Weighting Waiting Evaluating Perception of In-Vehicle Travel Time
Under Moving and Stopped Conditions. Transportation Research Record Journal of the
Transportation Research. 1898 , 61-68.
[13] Zhang, L., Xie, F. and Levinson, D. (2009) Illusion of Motion. Transportation Research
Record: Journal of the Transportation Research Board. 2135 , 34-42.
[14] Houten, R. V., Nau, P. A. and Merrigan, M. (1981) Reducing elevator energy use: A
comparison of posted feedback and reduced elevator convenience. Journal of Applied Behavior
Analysis. 14 (4), 377-387.
[15] Gerstenmeyer, S. and Peters, R. (2014) Reverse Journeys and Destination Control. In:
Symposium on Lift and Escalator Technologies. Northampton:
Caroline Bird graduated from Royal Holloway, University of London in 2015 with a First Class
Honours in Psychology. During her student years she participated in lift surveys for Peters Research.
She continued part time after graduation to lead the research project presented in this paper.
Richard Peters has a degree in Electrical Engineering and a Doctorate for research in Vertical
Transportation. He is a director of Peters Research Ltd and a Visiting Professor at the University of
Northampton. He has been awarded Fellowship of the Institution of Engineering and Technology,
and of the Chartered Institution of Building Services Engineers. Dr Peters is the author of Elevate,
elevator traffic analysis and simulation software.
Elizabeth Evans is the General Manager of Peters Research Ltd. She is also involved in the technical
aspects of business, project managing commercial and research projects. She is the Treasurer of the
CIBSE Lifts Group and a member of the Lift & Escalator Symposium organizing committee. In 2012
she was awarded the CIBSE Carter bronze medal.
Stefan Gerstenmeyer is the Head of Traffic and Group Control at thyssenkrupp Elevator Innovation
GmbH. He has been involved in R&D projects relating to group and dispatcher functions for lift
controls including multi car lift systems. He is a post graduated research student at the University of
Northampton.
Evaluating a Holistic Energy Benchmarking Parameter of Lift
Systems by using Computer Simulation
Ricky Chan1, Albert So2, Stefan Kaczmarczyk3
School of Science and Technology, The University of Northampton, United Kingdom
123
1
[email protected],
2
[email protected]
3
[email protected],
Keywords: Energy consumption, benchmarking parameter, lift traffic, lift drive, normalization,
computer simulation
Abstract. At present, there are benchmarking parameters to assess the energy performance of lifts ,
e.g. one in Germany adopted by VDI (4707-1/2), one internationally published by ISO (BS EN ISO
25745-2:2015), and the other in Hong Kong adopted by The Hong Kong Special Administrative
Region (HKSAR) Government. These parameters are mainly checking the energy consumed by a
lift drive without considering real time passenger demands and traffic conditions; the one in Hong
Kong pinpointing a fully loaded up-journey under rated speed and the two in Europe pinpointing a
round trip, bottom floor to top floor and return with an empty car, though including energy
consumed by lighting, displays, ventilation etc. A holistic normalization method by Lam et al [1]
was developed a number of years ago by one of the co-authors of this article, which can assess both
drive efficiency and traffic control, termed J/kg-m, which is now adopted by the HKSAR
Government as a good practice, but not specified in the mandatory code. In Europe, the energy unit
of Wh has been used but here, Joule (J), i.e. Ws, is adopted to discriminate the difference between
the two concepts. In this article, this parameter is evaluated under different lift traffic scenarios
using computer simulation techniques, with an aim of arriving at a reasonable figure for
benchmarking an energy efficient lift system with both an efficient drive as well as an efficient
supervisory traffic control.
1 INTRODUCTION
The energy consumption of lift systems, in the past, did not receive much attention because it only
accounts for a relatively small percentage of total energy consumption of a building. In fact, this
statement is only correct when a commercial office building is considered, but not for residential
buildings. According to the statistics of a government department in Hong Kong overseeing energy
efficiency, the total energy consumption of the lift system in a typical office building is less than
11% of the energy consumption of the whole building (Yeung and Lau [2]). According to Lift
Report [3], in Europe, the energy consumption of lifts typically represents 3 to 8% of the total
energy consumption of buildings, depending on the structure and usage of the building, and the type
and number of lifts. The report published in 2010 estimated that there were around 8.5 million lifts
in operation worldwide. In 2016 this figure should be close to eleven to twelve million, with an
estimated growth of around 670,000 per year.
Schroeder [4] developed a generalized formula to calculate the annual energy consumption of lifts
per square metre of building space. Doolaard [5] compared the relative consumption of energy by
hydraulic lifts, AC-2 lifts, and ACVVVF lifts. Al-Sharif [6] discussed several topics related to the
energy consumption of lift systems by comparing the consumption of various types of drives and
outlining the concept of regenerating power back into the supply grid.
In this paper, a review of various issues regarding energy efficiency is made, with particular
reference to the mandatory Building Energy Code (BEC) 2015 [7] in Hong Kong and the guideline
[8], BS EN 25745-2. This paper aims to provide a holistic energy efficiency benchmarking
parameter that covers all types of drives, including but not limited to AC2, ACVV, DCWL, DCTL,
ACVVVF (scalar and vectored), PMSM, linear machines, hydraulic etc.
2 LITERATURE SURVEY
2.1 VDI 4707 and ISO 25745-2

The first energy guideline for lifts and escalators could refer to VDI 4707 initiated in Germany with
guidelines published by the Association of German Engineers (VDI), a draft of which appeared at
the end of 2007. It classifies lift performance into seven categories, “A” the best and “G” the worst.
The classification is based on two measurements, namely “travel” and “stand-by”. A mathematical
procedure is employed to analyze the measurement with reference to usage category, speed, rated
load and travel height to arrive at the classification. The “travel” demand is the total energy
demand of the lift during trips at specified trip cycles and with a defined load while the resultant
specific demand value is given in mWh/m-kg. Four usage categories were defined, namely “low”,
“medium or occasionally”, “high or frequently” and “very high or very frequently”. The actual
procedures of measurement and analysis are detailed in ISO 25745-1 and ISO 25745-2. In our
study, J/kg-m is consistently adopted due to the big difference between the two concepts.
The terminal landings cycling test is defined in Clause 2.18 of ISO 25745-1:2012, stating that the
empty car is continuously cycled between the bottom terminal landing and the top terminal landing,
with the door operations enabled. Section 4 of ISO 25745-2 further defines running energy
measurements in two ways, the one between two terminal landings with two complete door cycles
(termed a reference cycle), or the one between two predetermined landings with two complete door
cycles (termed a short cycle). Based on the running energy of a reference cycle, Erc, and the
running energy of a short cycle, Esc, and their corresponding one-way travel distance, src and ssc,
the average running energy consumption per metre of travel, Erm, can be estimated by equation (1).
1  E rc − E sc 
E rm =   (1)
2  s rc − s sc 
From this Erm, other parameters such as Essc (start/stop energy consumption for each trip), Erav
(running energy of an average cycle) and Erd (daily running energy) etc. can be evaluated.
2.2 The Building Energy Code of Hong Kong
The first code of practice related to energy in Hong Kong is perhaps the Code of Practice for
Overall Thermal Transfer Value (OTTV) in Buildings [9] published by the Hong Kong Government
in April 1995. Then, in 1997, a task force with four sub-committees was established within the
Electrical & Mechanical Services Department (EMSD) of the HKSAR Government to draft codes
of a similar nature but on different building systems, namely Lighting, Air-Conditioning, Electrical
Services and Lifts and Escalators between 1997 and 1999. In 2012, the four codes, and others, were
combined into one document, Code of Practice for Energy Efficiency of Building Services
Installation, called BEC in short [10]. Under the enforcement of the Building Energy Efficiency
Ordinance Cap 610 in the same year, this combined code of practice became mandatory in Hong
Kong. All new and extensively retrofitted buildings need to comply with the code of practice. By
2015, the code was slightly revised with some tightened clauses and published in 2015 [7]. As a
companion to the code, a set of guidelines was also published by the EMSD [8].
Evaluating a Holistic Energy Benchmarking Parameter of Lift Systems by using Computer Simulation 6-3
The item in the BEC that is closely related to ISO 25745 may perhaps be the limit of maximum
allowable electrical power of motor drives. Inside the BEC, tables provide the maximum power of
a motor drive with respect to the rated load and the rated speed of a lift as measured under a fully
loaded rated speed with upward movement. There are separate tables for hydraulic lifts, escalators
and passenger conveyors.
2.3 The proposed Benchmarking Parameter, J/kg-m
So far, it can be observed that all existing international standards or national guidelines mainly
concern the efficiency of the lift along a standard trip, either no-load or full-load. But we should be
aware that most real journeys are neither full loaded nor no loaded. The Hong Kong BEC concerns
the power consumption, not accumulated energy, during a full-loaded rated speed up journey
although under some circumstances regenerative braking is mandatorily required. ISO 25745-2
concerns the whole reference cycle by measuring the accumulated energy during both no-loaded up
and no-loaded down journeys, including acceleration, deceleration and rated speed operation. If
regenerative braking is employed, its performance is also included in Erc. Having said that,
assessment by the two schemes is restricted to the motor drive alone.
One of the authors of this paper, together with other researchers, raised an argument some eleven
years ago by So et al [11] that merely an energy efficient motor drive is not the ultimate solution to
an energy efficient lift system. Efficiency of the drive can only account for the hardware
performance, whereas the main saving should come from supervisory traffic control. In that 2005
paper by So et al [11] it was shown that by using the same motor drive, a significant reduction in
energy consumption could be obtained by using different traffic controllers. One with artificial
intelligence associated with energy saving could achieve a distinctive result. Based on this
argument, a good benchmarking parameter for energy comparison must take care of both the
physical drive performance as well as the soft traffic control algorithms. Therefore, the idea of J/kg-
m was suggested.
The basic concept of J/kg-m is simple. It is the average energy required to convey one unit of mass,
passengers or goods, a distance of one metre, irrespective of direction over a fixed and agreed
period of time. An energy efficient motor drive can of course lower such an average value, but an
energy efficient supervisory control system can lower the value by a more significant amount, the
illustration of which is the main theme of this paper. To evaluate this benchmarking value, three
measurements have to be made:
i) energy consumed, in Ws or J, over the fixed period of time, T, say 2 hours (7200 sec) long;
ii) mass of load, in kg, inside the car, at any time within T;
iii) position of car, in m, along the hoist-way at any time within T; this is to estimate the
distance traveled by the car.
This parameter has been included in the guideline of the BEC published in 2012 by EMSD [8] as a
good practice recommended to lift owners, manufacturers and maintenance contractors. However,
although (i) could be easily measured by an external power meter (actually mandatory in the 2015
BEC) [7], (ii) and (iii) are usually not readily available to the lift owner or user. Thanks to the
publication of the recently approved BACnet objects through ASHRAE [12] for lifts and escalators,
all three can be obtained by the appropriate implementation of the relevant BACnet objects.
Within the period of time from 0 s to T s, say two hours, i.e. 2 * 3600 = 7200 s, there could be N
number of brake-to-brake journeys of one car or several cars belonging to the same bank. The ith
brake-to-brake journey commences at the instant when the brake is released at the departing floor
for the car to accelerate and ends at the instant when the brake is applied again for the car to park at
the destination floor. During this journey, wi kg of load is conveyed and a total distance, di m, is
displaced, where i runs from 1 to N.
Without loss of generality, this definition also applies to a bank of lift cars. A time increment, ∆T,
say 15 minutes, can be defined so that another time period from ∆T to T+∆T can be formulated. The
same process is conducted within this new time period, and goes on and on. At the same time, the
total energy, ET(k), consumed during a particular period, the kth period, of T s has to be recorded. It
is obvious that ET(k) includes not just the consumption of the motor drive but others including
lighting, ventilation and indication etc. Eventually, one J/kg-m (k) value can be found for each kth
time period, either for one car or a bank of cars. A daily or weekly average can finally be obtained.
So, for the k th time period, the following equation (2) is valid. Any brake-to-brake journey across
the two limits of the kth period could also be included in equation (2) as it does not affect the
statistics by much.
ET (k )
J/kg - m ( k ) = N
(2)
∑
i =1
wi (k )d i (k )
3 VALIDATION OF THE BENCHMARKING PARAMETER BY SIMULATION

3.1 Background of the simulation software
As described and defined earlier, the advantage of the benchmarking parameter, J/kg-m, is that it
can tell at a glance whether the energy consumption of the lift is within a well-established and well-
recognized energy efficient range. But how to find out this benchmarking figure as a reliable and
convincible iconic number for reference by the lift industry and professionals needs great effort to
carefully assess the quality of service for existing lift installations, said Richard Peters based on
Strakosch and Caporale [13]. He again said that assessing lift transportation to know the exact
passenger demand was not an easy task.
To know the passenger demand more precisely, we need to quantify the lift traffic, and most
important is to know how to measure it. Before the advent of this software, the previous researchers
might try to clamp the traffic analyzers in the lift control systems to log on the operational data. For
the passenger demand, they could simply do manual traffic surveys. All these methods were
suggested by Richard Peters with Strakosch and Caporale [13]. The latter however are so laborious
and time consuming by waiting in the lift lobby and inside the lift cars to count the numbers of
passengers in their waiting and transit times. Fortunately there was a very good simulation tool [14]
which has been developed since 1989 from its first version, and now been upgraded to its 8th
version.
The features of this software include all functions of previous versions, with a newly added real-
time instantaneous energy consumption, kWh = 3,600,000 J. Through this added energy function,
we can make the evaluation and assessment of the benchmarking parameter, the J/kg-m, feasible.
That means that the instantaneous power consumption, kW, has become part of the data display
during analysis consumed in every single time interval, say five minutes of the lift, either in its up-
peak, inter-floor or down-peak operation throughout the simulation. Figure 1 below indicates the
simulation display with real-time animation of those input lift cars, four of them, going up and
down during the simulation. To make this energy consumption more accurate, all the mandated
inputs on building data, lift data and passenger data must be checked and inputted with care. Then
the energy consumption (kWh) and the instantaneous cumulative power consumption (kW) are
indicated in graphical form as the gauges displayed near the bottom of the slide in Figure 1.
The fundamental and advanced functions of using this software to do the design work, such as to
find out the round trip time, the average transit time, average waiting time and queue lengths, etc.
will not be mentioned in this exercise. In this paper, the main concern is to use different scenarios to
test the benchmarking parameter, J/kg-m. It is obvious that the smaller the value of the J/kg-m , the
more energy efficient the lift (So et al) [11].
Figure 1: The simulation display with the feature on lift power consumption
4 RESULTS AND ANALYSIS

4.1 Raw data converted to spread sheet for analysis
Table 1 below illustrates the output from the simulation software regarding brake-to-brake journeys
with details such as time, position, load etc. Finally when the individual time spent by journeys of
the four lifts is known, and their corresponding traveled distances, the parameter, the J/kg-m at the
prescribed moving time period or window can be made available. Table 2 shows how J/kg-m is
evaluated after the energy consumed and the load-distance products are known.
Table 1 Part of the Raw Data extracted from the simulation results of the four lifts
Readable Time
Lift
Time Floor Load Traveled Time used
no.
In (hr:min:sec) From/To Floor kg*m
(sec) (21 story) (kg) Dist. (m) (s)
(1~4)
=Time/3600
1 27946.1 7:45:46 1 75 From G/F
1 27974.4 7:46:14 16 75 60 To F16 4500 0:00:28
1 27982.3 7:46:22 16 0 From F16
1 28010.6 7:46:50 1 0 60 To G/F 0 0:00:28
1 28142.2 7:49:02 1 0 From G/F
1 28148 7:49:08 2 0 4 To F2 0 0:00:06
Note: The shaded heading in grey was the raw data generated from the software
Table 2 One of the calculation examples of J/kg-m vs. moving time period
Moving Time
Sum of Total
Period or Sum of J
kg*m J/kg-m Average
Window
8:45 258057010.8 5105700 50.542925 -
9:00 257868597.3 5414400 47.62644 43.383857
9:15 252151793.9 5643300 44.681621 43.383857
9:30 246846141.8 5825400 42.37411 43.383857
9:45 243568562.3 5928300 41.085735 43.383857
10:00 229031523.7 5676000 40.350867 43.383857
10:15 232685791.1 5448900 42.70326 43.383857
10:30 232426069.8 5165100 44.999336 43.383857
10:45 224709987.8 5061600 44.395051 43.383857
11:00 222824627 5440500 40.956645 43.383857
11:15 212200154.6 5657700 37.506435 -
This energy model is available to simulate the energy use of the lift in question traveling at up peak,
down peak and inter-floor traffic patterns. Inputs are available when this energy model has been
converted to the analytical data, say, in spread sheet format. For each lift car, the power consumed
during a journey can be defined for different passenger loads, say at its 0, 20, 40, 60, 80 and 100%
in both up and down directions.
4.2 Simulation in lift traffic control and energy consumption

In the simulation work of this paper, different lift operating scenarios were attempted, i.e. using
different sets of the lift configurations; such as the building data, the lift data and finally the
passenger data. Subject to limitation in space of this paper and the intention of illustration, the input
scenarios were limited to four, i.e. from Figures 2 to 5. The last figure (Figure 5) was purposely
tried to include some unreasonable and uncommon scenarios to test what simulated result would
come out.
Figure 4 750 kg, 2.5 m/s, 21 stories, 80% loaded Figure 5 1300 kg, 5.0 m/s, 21 stories, 80%
loaded simulation (purposely over-designed)
The time scale in running these simulations is from 07:45 am to 12:15 pm but the first 15 and last
15 minutes are ignored. With T = 2 hours and a running increment of 15 minutes, there are nine
reasonable values (though 11 data points displayed) of simulations during a 4.5 hour time window.
The first and last points in the four figures should be ignored. The first useful value represents the
period from 08:00:00 to 09:59:59, thus centered at 09:00:00. The second value represents the
period from 08:15:00 to 10:14:59, thus centered at 09:15:00, until the last moving time period at
11:00.
During this 4.5-hour simulation, it is expected that the J/kg-m values are high during the first 2
hours due to up-peak and this can be explained through common sense. In the morning, lift riders
are rushing to commence work to their offices. During up-peak, up journeys are close to full loaded
as the motor is also working at full load. Down journeys are also close to full load. The J-
parameter value would lie in the middle range, as during the coming one-hour, staff would use the
lifts up and down as their inter-floor journeys, in particular for those offices, clinics, law firms, etc,
that occupy several floors under the same companies, and so they need to travel within their own
occupied floors. A low value of J/kg-m is expected during the last 1.5 hour in the morning session,
indicated as down-peak, such as lunch time or for the half-day workers in these lift riding cycles.
During down-peak, every journey is closed to full-loaded down, irrespective of the actual moving
direction.
5 CONCLUSIONS
After completing these few simulations by using the software [14] with the power consumption
feature at different lift operating scenarios, a series of the benchmarking parameter values in J/kg-m
was collected, though the sampling size in this exercise was not enough to secure a publicly
acceptable standard. Yet from the graphs, a preliminary conclusion could be drawn with some
confidence. Whenever the lifts are designed at rated capacity of either 750 or 1000 kg and an
operating speed of 2.5 m/s, under traveled distance at about 80 metres, i.e. 21 stories, the J/kg-m
value does not fluctuate too much. Though Figure 5 shows an average of 70 J/kg-m which seems to
be quite high comparatively, in fact, it is a special case to simulate an over designed scenario.
Apparently this small simulation work could at least demonstrate the concept of using J/kg-m as a
benchmarking parameter that could well agree to the statements as So et al said [11]. Furthermore, a
reasonable value could be suggested at 50 J/kg-m as stated in the BEC Technical Guidelines [8],
which is well supported by our simulation works here, i.e. 43 J/kg-m on average (Figure 2 and 3)
and 52 J/kg-m on average (Figure 4).
The over designed scenario as reflected from Figure 5 at an unreasonable high rated speed of 5 m/s
and rated capacity of 1,300 kg is further explained here. In this trial design figure, when a larger lift
motor is used, more power is consumed no matter under whatever traffic conditions. That is why a
higher J/kg-m value is obtained in the simulation. By the way, the pattern of curves depicted in
Figures 2, 3 and 4 is quite steady by itself, but it is quite different in Figure 5. The curve in Figure 5
shows surprisingly low J/kg-m value during up-peak, while it is flat during the inter-floor traffic,
and relatively high at down-peak. That means it is quite different from those with normal or
reasonable design scenarios, except a steady J-value appears during the inter-floor traffic period, but
it has a comparatively high J/kg-m value during the lunch time. Probably, during down-peak, cars
are not fully loaded due to its big rated capacity at 1300 kg and therefore regenerative energy is not
enough to compensate the energy consumed by the lift motor.
Further works are suggested to conduct more simulation tests with broader range of scenarios and
combinations of the input parameters/data with different design configurations in terms of rated
capacity, rated speed, zoning and passenger demands. Finally going back to the argument whether
Wh, mWh or Joule (Ws) would be used for the benchmarking parameter, the authors would like to
have a more obvious demarcation between ours and the currently used European parameters
because the concepts between these two are totally different; ours being on real-time measurement
with more complicated traffic patterns, while the European or Hong Kong ones being on fixed load
patterns.
REFERENCES
[1] Lam, Dante C.M., So, Albert T.P. and Ng T.K., "Energy conservation solutions for lifts and
escalators of Hong Kong Housing Authority", Elevator Technology 16, Proceedings of 16th World
Congress on Elevator Technologies, IAEE, Helsinki, June, 2006, pp. 190-199.
[2] Yeung K. and Lau O (2011), “Building energy code – the way towards low carbon building”,
Prof. EMSD Symposium on E&M Safety & Energy Efficiency
[3] Lift Report (2010), “Lifts energy consumption study”, http://www.lift-
report.de/index.php/news/464/56/Lifts-Energy-Consumption-Study.
[4] Schroeder J. (1986), “Energy consumption and power requirements of elevators”, Elevator
World, Vol. 34, pp. 28-29.
[5] Doolaard D.A. (1992), “Energy consumption of different types of lift drive systems”, Elevator
Technology 4, Proc. Elevcon 92, Amsterdam, pp. 214-252.
[6] Al-Sharif L. (1996), “Lift power consumption”, Elevator World, May, Vol. 44, pp. 85-87.
[7] EMSD (2015), Code of Practice for Energy Efficiency of Building Services Installation,
http://www.beeo.emsd.gov.hk/en/pee/BEC_2015.pdf.
[8] EMSD (2012), Technical Guidelines on Code of Practice for Energy Efficiency of Building
Services Installation, http://www.beeo.emsd.gov.hk/en/pee/TG-BEC_2012_r1.pdf.
[9] Buildings Department, Hong Kong (1995), Code of Practice for Overall Thermal Transfer
Value in Buildings, http://www.bd.gov.hk/english/documents/code/OTTV-01.pdf.
[10] EMSD (2012), Code of Practice for Energy Efficiency of Building Services Installation,
http://www.beeo.emsd.gov.hk/en/pee/BEC_2012%20(Rev.%201).pdf.
[11] So Albert, Cheng G., Suen W. and Leung A. (2005), “Elevator performance evaluation in two
numbers”, Elevator World, Vol. LIII, No. 1, January, pp. 102-105.
[12] ASHRAE (2016), ANSI/ASHRAE Addendum aq to ANSI/ASHRAE Standard 135-2012 Data
Communication Protocol for Building Automation and Control Networks.
[13] Strakosch R and Caporale R (2010), “The Vertical Transportation Handbook”, Fourth Edition,
by John Wiley & Sons, Inc., Hoboken, New Jersey, pp543-548
[14] Peters Research Ltd (2014), Link for ElevateTM Version 8 Manual: https://www.peters-
research.com/index.php/support/manual
1 Mr. Ricky Chan is currently a visiting lecturer in the City University of Hong Kong and a PhD
candidate of the University of Northampton U.K.
2 Dr. Albert So is an executive board member and scientific advisor of the International
Association of Elevator Engineers (IAEE). He is the honorary visiting professor of the
University of Northampton in the U.K. He serves on the Advisory Group of Elevator World,
Inc., and is based in Seattle.
3 Professor Stefan Kaczmarczyk is Professor of Applied Mechanics and Postgraduate Programme
Leader for Lift Engineering at the University of Northampton. He is currently supervisor of the
PhD candidates in the University
The Report of Thorough Examination as a Management Tool for
Maintenance
Dominic Dawson
Zurich Engineering 126 Hagley Rd,
Edgbaston, Birmingham.
B16 9PF
Keywords: Thorough Examination, maintenance.
Abstract. Legislation in the UK (LOLER 98 [1]) demands that lifting equipment in the work place1
is subjected to a Thorough Examination periodically – usually every six months. For installations
outside the work place, owners still have to satisfy the requirements of other legislation such as the
Health and Safety at Work etc Act 1974 to provide a safe lift. By following the same measures as
those in the workplace owners will in all probability satisfy legislation.
Therefore, all lift owners should regularly receive a report of Thorough Examination which will
detail the findings of the examination. The study looked at the possibility of using the report as an
aid to monitor the maintenance activity. It found it to be a useful tool in this respect; however a
level of understanding of the report along with other information regarding the maintenance activity
would be required to provide the owner with a complete and objective picture.
1 INTRODUCTION
There are many factors which contribute to make the lift a statistically safe means of transport:-
It only travels vertically within a guarded area, the range of travel is limited to its guided
path and conventionally it has no other lifts with which it may collide.
The construction and installation is generally well regulated and there are exacting standards
which should be followed.
The manufacturer should provide comprehensive instructions on its safe operation and
maintenance.
Legislation in the UK calls for periodic independent examination and suitable maintenance
of the lift.
The safety is therefore influenced by the design and manufacture, installation and after care. Clearly
the area most influenced by the owner is the after care.
Legal responsibilities are imposed on the owner or duty holder of a lift in the workplace to ensure
maintenance and examination. In all situations even where the legal responsibilities are contrived
there will be moral responsibilities which do not necessarily equate to legal accountability. In a
modern ethical society therefore the need to maintain a safe lift is paramount both legally and
morally. The (Penguin English Dictionary) defines moral as:
Moral: relating to the principles of right and wrong in human behavior; ethical. Conforming to a
standard of right behavior or to the dictates of one’s conscience.
1
The workplace Health safety and Welfare Regulations define the workplace as any premises or part of
premises which are not domestic premises and are made available to any person as a place of work, and
includes any place within the premises to which such a person has access while at work, any room, lobby,
corridor, staircase, road or other place used as a means of access to or egress from the workplace or where
facilities are provided for use in connection with the workplace other than a public road.
The same source states that responsibility is “the state of being responsible” and “a moral or legal
obligation”. Responsible is defined as “Liable to be called to account as the person that did
something – having control or care of something or somebody”.
Rather than a personal moral code, in the case of a lift owner it is more often a collective moral
responsibility which refers to arrangements appropriate for addressing widespread harm and
wrongdoing associated with the actions of groups. The key components of the basic notion of
moral responsibility are deeply rooted in the fabric of every society and are constitutive of social
life. Without some conception of moral responsibility our society would be uncivilized and
unrecognizable to that which we currently enjoy.
An example of a moral code leading from a responsibility is where within a contract of maintenance
the owner and the maintainer may agree to be bound by the LEIA2 Voluntary Code of practice,
which is not a legal requirement but provides an ethical list of responsibilities to which both parties
adhere.
The inspection body will provide periodic Thorough Examinations and will issue the owner with a
report showing the findings of that examination. The maintainer will provide the maintenance and
repair of the lift. The owner should provide monitoring of the maintenance function. Key
performance indicators should be incorporated along with a method to monitor them; the Thorough
Examination report can help to provide this. The ACOP to LOLER 98 states that the report of
Thorough Examination is a vital diagnostic aid to the safe management of lifting equipment, and the
HSE in guidance note INDG339 [2] suggest that the report of Thorough Examination may be used
to aid maintenance monitoring.
2 THE THOROUGH EXAMINATION REPORT

A well written examination report is a useful tool and can provide a wealth of information to an
informed reader. UK legislation requires a Thorough Examination report in order to satisfy
regulation 10 of LOLER 98, this is the result of the Thorough Examination required under
regulation 9 and Schedule 1 of the regulations sets out the information required in the report.
Part 8 of Schedule 1 is concerned with defects. LOLER only specifies that defects which are or
could become a danger to persons are reported in good time giving detail of the defect and
particulars of any repair or alteration to remedy it.
A typical LOLER examination report will contain three sections to part 8. Section 8a and 8b list
defects and section 8c is reserved for observations. Part 8a is confined to those defects which pose
an imminent risk to persons - both users and maintainers. These defects should be repaired either
before further use, for an issue which, in the competent person’s opinion, will manifest its-self
imminently (within the next few operations), or within a specified time. Time related defects are
those which the competent person determines will not fail imminently but within a short time – a
usual period of time is up to three months.
Other defects which are safety related but have not deteriorated sufficiently to be categorised in part
8a are listed in part 8b and should be repaired as soon as reasonably practicable. In other words, the
owner does not have to take measures to avoid or reduce the risk if they are technically impossible
or if the time, trouble or cost of the measures would be grossly disproportionate to the risk. It is
generally conceded that a “reasonable” time for attention to these issues should be at the next
maintenance visit – but definitely before the next examination.
2
LEIA – Lift Escalator Industry Association
The Report of Thorough Examination as a Management Tool for Maintenance 7-3
The legal approach to this term is well known and various cases can be sited to cover this - Edwards
v National Coal Board 1949 [3] and McCarthy v Coldair Ltd 1951 [4].
Part 8c is where other observations may be recorded; these will include defects and issues which are
not safety related to the lift and maintenance or Health and Safety issues.
3 FINDINGS
400 Engineer Surveyors were asked to participate in the study, it was completely voluntary. Each
surveyor was asked to complete a survey form for the first 3 lifts they encountered over a two week
period in order to encourage randomness.
The study found that many examination reports contained more information than is required under
LOLER 98 and many reports read like a condition survey rather than an assessment of the lift
safety. This superfluous information was mostly found in section 8b, and therefore incorrectly
categorised under LOLER 98.
The inclusion of this information is historic and dates back legislation prior to the introduction of
LOLER 98 such as the Factory Act, OSRP3 Act and the HEO4 where comments on condition and
maintenance were encouraged on the prescribed form F54.
The lift owners, their consultants or maintainers are often in disagreement with inspection bodies
regarding these comments because it may appear from the report that the lift is unsafe to use if the
issues raised are not corrected within the time periods established. The maintainer’s performance is
often measured on the outstanding defects and they may be unfairly penalised for issues which do
not affect the safety of the lift nor are included in the maintenance contract.
These issues and defects however are required in order for the report to be used in the manner
suggested as an aide to monitor the maintenance provision. Therefore, rather than be a constant
point of dispute, the information contained within the report should be embraced and be used in the
most beneficial way for all. To enable this however it is vital that the inspection industry should
ensure that the issues are categorised correctly on the report and be aware of the impact that
incorrect categorisation has on the maintainer.
For the reasons given above the information in section 8b of the reports was categorised into issues
or defects identified as maintenance, condition, installation and health and safety. For the study
those items concerned with maintenance were extracted. From a potential 1200 returns just short of
170 were returned, 159 of which were usable. There were 461 “8b” defects recorded, of which 207
were determined to be maintenance issues.
The definition of maintenance issues or defects was taken from page 15-4 of the CIBSE guide D
2015 [5] which refers to maintenance as “adjustment, cleaning, lubrication and replacement of worn
parts”.
Items critical to passenger safety either directly – such as areas relating to the passenger interface
(doors, levelling accuracy and the alarm system etc.) or indirectly – such as brakes and ropes –
applied the criteria from guide D to determine that these issues would be considered a maintenance
task under the majority of maintenance contracts – be it comprehensive or just an “oil and grease”
contract.
3
Offices Shops Railway Premises. 4Hoist Exemption Order.
Maintenance instructions imposed by the manufacturer will normally also have duties placed upon
the lift owner or duty holder to check the former, for example providing weekly checks. Most
recorded accidents happen at the interface between the lift and the landing such as contact with
moving doors, tripping due to poor levelling and crushing due to unintended movement5.
The issues were given general headings in the data analysis, and the headings and detail are shown
in chart 1 below:
Chart 1 Maintenance issues
100 65
30 26 32
50 9 13 17 4 5 6
0
Maintenance issues
Lubrication Lighting
Loose fixings Rope/Belt terminations
Doors Brake
Alarm system Levelling
Cleanliness and housekeeping Documentation
The aim of proactive preventative maintenance should be that these issues are corrected before they
become a problem and ideally they should be corrected before they are detected at a Thorough
Examination.
The examination history revealed that some issues were recorded at the previous examination (pre-
existing) and were still evident - an indication of a failure to repair and maintain the lift effectively.
91 lifts contained one or more pre-existing issues.
Using the on-site log card as a record, the maintenance dates and activities were referenced against
the findings on the report, and it was established there were maintenance issues remaining despite a
maintenance visit within the previous 6 weeks. Chart 2 below shows the number of issues recorded.
The data suggested that the type of lift, environment, maintainer or age of lift had little effect on the
maintenance performance since similar levels of performance and spread of issues were found.
The phrase, “A lift is a lift wherever and whenever it was installed” appears to be true. It should be
remembered however that this study only considers the maintenance issues, the other type 8a and 8b
defects were not recorded as they are not considered to be maintenance issues.
As BS EN 81-80 [6] has shown there are problems with older lifts from a safety point of view due
to the design and technical advancement of lifts and the safety components. This should be
considered along with maintenance.
5
The study cited an accident reported in the Epping Forrest Guardian 27/10/2010 - http://www.guardian-
series.co.uk/news/8479126.WOODFORD_LEYTONSTONE__Brakes_to_blame_for_lift_death/
27/10/2010.
The Report of Thorough Examination as a Management Tool for Maintenance 7-5
Chart 2 maintenance issues where a service visit had taken place within the previous 6 weeks
32
50 9 16 8 25 5 18 5 15 2
0
Pre -eisting maintenance issues
Lubrication Lighting
Loose fixings Rope/Belt terminations
Doors Brake
Alarm system Levelling
Cleanliness and housekeeping Documentation
The type of maintainer may have a bearing on the effectiveness of the maintenance, and selection
should not be confined to the manufacturer, as the study showed that independent maintainers
performed as well if not better.
Although on the whole it was found that the environment had little effect on the maintenance
performance there are special considerations to be made concerning certain situations such as in
hospitals. The contracts for maintenance may require out of the ordinary inclusions such as
passenger release due to the environment (HTM08-02) [7] and procurement for maintenance
services in these areas should be aware of this.
The findings suggest that there is a need for the education of lift owners and duty holders, and that
in some cases there is poor management and records of maintenance (Cooper [8]) which may be
down to a number of reasons such as time constraints, poorly written contracts, financial constraints
or inadequate understanding from those completing the maintenance (Cooper [9]).
4 CONCLUSIONS
The support and co-operation of Zurich Engineering (ZE) made this study possible, however it
should be noted that the opinions expressed are those of the author and not necessarily ZE.
Throughout the study it has been supposed that the findings on the Thorough Examination reports
are correct; work during and prior to this study did show in some cases across the inspection
industry a less than adequate standard of reporting. It was found that confusion exists in some cases
due to inspection bodies using differing terminologies and interpreting the requirements of LOLER
98 in different ways.
There is currently a call for standardisation of the current LOLER report (CIBSE Guide D 2015
p15-8 [10]). Smith [11] noted there is some misunderstanding of the report form and the defects
contained within it. It is an area that the inspection industries should investigate, and work with
clients and maintainers to resolve.
It is inadequate to just assume that the maintenance duty is being completed correctly, and some
kind of monitoring of the maintenance function should be provided. The study concluded that clear
communication between the owner and both the maintainer and the inspection body, and between
the maintainer and the inspection body should be initiated – possibly written into the maintenance
contract. It is believed that this would provide some clarity and transparency within the examination
and maintenance provision.
Some inspection bodies have now developed online platforms which will facilitate this. Further
work needs to be done by Inspection bodies possibly under the umbrella of SAFed6 to standardise
the reports.
There has long been a divide between inspection and maintenance companies, which is a major
hindrance to the effectiveness of lift safety and reliability. Owners are often frustrated by the
standoff that appears to be evident, and transparency and co-operation should be a major objective
for all.
5 LITERATURE REFERENCES
[1] LOLER Lifting operations and lifting equipment regulations1998 – approved code of
practice (ACOP) and as amended L113 Second Edition 2014
[2] HSE Thorough examination and testing of lifts – Simple guidance for lift owners –
INDG339 rev 1 01/2008 HSE
[3] B Barrett & R Howells. Occupational Health and Safety Law – text and materials 2nd
edition - Cavendish publishing Ltd
[4] Fife & Machin. Redgraves – Health and Safety in factories - Butterworth & Co
[5,10] CIBSE Guide D Transportation systems in buildings 2015 ISBN 978-1-906846-64-0
[6] BS EN 81-80 Safety rules for the improvement of safety of existing passenger and goods
passenger lifts:2003 BSI.
[7] Department of Health Health Technical Memorandum 08-02 Department of Health
[8] David Cooper Elevatori – the meaningless tick. Issue 1-2015 p82 ISSN11217995
[9] David Cooper Elevation issue 76 p96, 77 p82, 82 p56 Cyber Coms Ltd.
[11] Laura Smith The implications of the Lifting Operations and Lifting Equipment Regulations
(LOLER) 1998 in care homes. University of Salford
The author is a Senior Engineer currently employed by Zurich Engineering and has 23 years’
experience in the inspection industry. He served an electro mechanical apprenticeship in the Royal
Navy and gained an HNC in Electrical and Electronic Engineering from Highbury College in
Portsmouth before joining Plant Safety Ltd in 1993 as an Engineer Surveyor. In 2005 he joined
Zurich Engineering and was promoted to Senior Engineer. He is a member of the IET and IAEE
and has an MSc in Lift Engineering.
6
SAFed - Safety Assessment Federation
Multicar Dispatching
Stefan Gerstenmeyer1,3, Richard Peters2,3
1
thyssenkrupp Elevator Innovation GmbH, Bernhäuser Str. 45, 73765 Neuhausen, Germany
[email protected]
2
Peters Research Ltd, Bridge House, Station Approach, Great Missenden, Bucks, HP16 9AZ, UK
[email protected]
3
School of Science and Technology, The University of Northampton, UK
Keywords: Dispatcher, multicar, quality of service, waiting time, algorithm, elevator, lift
Abstract. When there is effectively no limit to the number of lifts in a shaft and the lifts can move
horizontally as well vertically, conventional dispatching operation and objectives need to be
reconsidered. This paper considers how to dispatch multicar lifts efficiently and explores the limits
of handling capacity. Quality of service cannot be measured simply in waiting time when a new car
appears at the main entrance floor almost immediately after the last car is dispatched; the dispatcher
must also consider bottlenecks in the shafts which can result in long delays in transit. The user
interface and signalling also needs consideration as ease of use may limit what information and
allocation options are available to the dispatcher. Safety distance considerations also impose limits1.
Dispatching strategies for shuttle operation and local operation are proposed.
1 INTRODUCTION
To overcome the limitations of roped lifts the concept of rope-less lifts, with cabins moving
independently in at least two dimensions, has been widely considered [1, 2, 3, 4]. The new freedom
of having multiple cabins circulating in at least two vertical shafts allows new ideas and options for
passenger transportation in buildings [5]. Vertical trains have been considered [6]. Round trip time
analysis of two-dimensional lift systems has been introduced [7].
The technology required for a circulating multi car lift system (MCLS) was introduced together
with a traffic concept where the MCLS is used as a shuttle connecting ground lobbies with sky
lobbies [8]. Linear motors, lightweight cabins, cabin guidance, vertical shaft exchanger units for
cars and certified safety systems are necessary to realise such a system. Parameters affecting
handling capacity and quality of service (QoS) in a MCLS shuttle application were discussed and
analysed [9]. Lift shafts can be used more efficiently if they are used by multiple lift cabins. But a
circulating MCLS is not limited to a shuttle application. It can also be used as local lift groups to
distribute passengers to their final destination floors. General arrangements are shown in Figure 1.
To control the operation of circulating MCLS general rules of lift behaviour [10] need to be
considered and expanded. General QoS criteria based on the psychology of waiting [11] and safety
distance constraints [12] are inputs for control algorithms.
1
Functional safety aspects as addressed by EN81-20/50 for roped lifts are necessary but not considered in this paper
Figure 1 Arrangements of a MCLS as local group
2 QUALITY OF SERVICE
Quality of service (QoS) in terms of traffic handling is mostly defined by waiting time (WT). The
interval traditionally also gives an indication of quality [13]. Other definitions of QoS exist, the
majority being based solely on interval or WT. Another factor is the transit time (TT). But QoS is
the total experience of a lift journey [11]. This includes the lift behaviour while serving passengers
requests. There is an accepted set of rules and constraints of lift behaviour [10, 13, 14, 15].
Summarised they are:
1. Do not bypass a car call/destination of a passenger
2. Do not transport passengers away from their destination
3. Only stop at a floor because of a car call or landing call
Multicar Dispatching 8-3
These rules also apply to the cabin behaviour in a MCLS as they alleviate the negative
psychological effects of reverse journeys and apparently unnecessary stops. For a circulating MCLS
rule 2 becomes less important if the cabins in the system are circulating and shafts are used only in
one direction at a time.
For MCLS these rules need to be extended to cover situations that occur if multiple cabins are
operated in the same shafts as mutual influence between cabins occurs. These additional rules
consider passengers’ perception and expectation of how lifts currently operate, taking into
consideration the additional control system options.
4. Stops at a floor without a car call or landing call are allowed if the doors stay closed and no
passenger is inside the car (an exception to rule 3).
5. Departure delays of cars with passengers inside the cabin shall be reduced to a minimum.
6. A cabin arriving at a landing and opening its doors for passenger transfer shall serve, in
addition to its cabin car call, all landing (or destination) calls allocated to this landing door
in the direction it is travelling.
Rule 4 gives controllers more flexibility, especially if a cabin ahead blocks the way for a following
cabin. With the circulating MCLS described in this paper it is necessary to stop at floors where
exchangers are located in order to change direction from vertical to horizontal.
The departure delays referred to in rule 5 can occur if loading times of cabins are not equal, the
number of stops is not equal, or if one cabin blocks the way of another [11]. The control system can
avoid such situations, although in special instances a departure delay could be the best choice.
Departure delays are a concept that can be built into the controller. They are known from the up-
peak behaviour of lifts, where a car is held in the lobby in order to wait for additional arriving
passengers so that the cabin is filled to a higher capacity factor. It is recommended that passengers
should not be held at the lobby for more that 10 to 15 sec [16]. Communicating to passengers the
reason for a departure delay can reduce passenger’s anxiety about their service, but even explained
departure delays can be annoying for passengers.
Rule 6 is related to the allocating of calls to cabins rather than to lift or cabin behaviour. It is
discussed in sections 5 and 6 of this paper.
3 HANDLING CAPACITY
The handling capacity is the number of passengers that can be transported within a specific time.
Traditionally in the lift industry the handling capacity is measured in 5 min periods (HC5). To
provide a good QoS sufficient HC5 is needed.
For a circulating MCLS maximum HC5 can be achieved if the cycle time (time between two
subsequent cabins in a two shaft system) is kept to a minimum [9]. To achieve minimum possible
cycle time the critical factors are stops made by the cabins and safety distance constraints [12]. For
a shuttle system all cabins have the same stops. If enough cabins are available, the maximum
possible HC5 is possible. This is different if a MCLS is used as local lift group. Due to different call
allocations and individual car calls (passenger destination floors) cabins will have different stops.
To avoid traffic jams caused by additional cabin stops and departure delays, the time between two
subsequent cabins (cycle time) measured at the main entrance floor needs to be increased. To avoid
collisions and traffic jams a graphical method in combination with Monte Carlo simulation was
described by Al-Sharif et al. [17]. The Monte Carlo simulation is used to simulate the different
stops of the cabins.
An increased cycle time to avoid traffic jams results in lower HC5 compared to a shuttle application
where all cabins have the same stops. If all cabins have the same stops and the distance of the stops
exceeds the minimum distance [12] between cabins the minimum cycle time (tCy) can be achieved
and the HC5 is the same as the HC5 of a circulating MCLS used as a shuttle. This is shown in
Figure 2 with cars D1 and D2.
A following car needs to be delayed if a front car has stops closer than its safe position defined by
the following car next stop. Without an additional delay safety distance rules would be violated. In
Figure 2 car D2 has two stops S21 and S22 that are closer to the safe position (S3SP(t)+dmin) defined
by the next stop S31 of car D3. Each additional stop of the front car requires a delay of the
following car.
Figure 2 Delayed cycle time of subsequent cabins
4 SYSTEM CONFIGURATION
The system configuration affects the control and dispatching strategies. In this section some system
configuration parameters impact control strategies, HC5 and QoS.
4.1 Exchanger
A circulating MCLS with at least 2 parallel shafts has at least two exchanger floors as shown in
Figure 1. One exchanger is located at the bottom floor and another at the top floor to enable the
circulation of cabins. It can help synchronisation to have the exchanger unit below the lowest
entrance floor, e.g. in a virtual landing without a door, see section 6.4. Middle exchangers between
bottom and top floors are possible and help to shortcut a round trip of a cabin. This reduces the
number of active cabins in a MCLS loop.
4.2 Linear motor
Lifts without ropes can be propelled with linear motors [8]. Coil units installed in the shafts are split
into segments. Only segments of coil units covered by the magnet yokes mounted on the cars are
involved in the movement of a specific car. Only the magnet yoke of one car is allowed to cover
one motor segment. If safety distances and controlled stopping points are calculated [12] the
segmentation of the linear motors also needs to be considered. Figure 3 shows that the minimum
distance (dmin) is possible in case A but not for case B as two cars cover the same linear motor. This
can be solved by an additional distance (dx) as shown in case C. The effect of this additional
distance to the safe position of a front car is shown in Figure 4. It shows the position over time of
two cars (DCar1(t) and DCar2(t)), the safe position of the front car 1 (DSaPo(t)) and the safe position
affected by the motor segmentation DSaPoM(t). This needs to be considered especially if the
minimum distance is needed between stops or floors.
Figure 3 Linear motor

segmentation Figure 4 Modified safety distance due to motor segmentation
4.3 3-shaft system

In a two shaft MCLS cabins are circulating. The cycle time that can be achieved between cabins
considering safety distance and QoS constraints defines the HC5. The incoming and outgoing HC5
is equal as the down direction shaft feeds the up direction shaft with cabins. If a significantly lower
cycle time can be achieved in e.g. the down direction compared to the up direction shaft a third
shaft supporting the up direction shaft can improve HC5 in both directions. As the cycle time in
shuttle applications is close to the minimum possible cycle time the effect of a third shaft will be
minimal or non-existent.
In lift groups with conventional control (collective control) the down peak HC5 can be 1.6 times
higher than the up peak HC5 [13]. The control system may choose where the cabins stop in the
down direction to collect passengers. Passengers with the same start floor are automatically grouped
together to travel to the main entrance floor. Cabins have fewer stops during a round trip. Fewer
stops leads to fewer unequal stops which enables a reduction in the time between cabins considering
departure delays. In this scenario a third shaft used in the up direction can have a benefit in HC5 in
both directions. The up direction shafts with higher cycle times are fed by the down direction
shaft’s arriving cabins with a lower cycle time. The down direction shaft with the lower cycle time
is fed by two up direction shafts each with a higher cycle time.
4.4 Express zones
High rise lift groups serve upper floors of a building bypassing lower floors as shown in Figure 1.
For traditional rope lifts, number of shafts, car velocities or cabin sizes needs to be increased in
order to achieve similar lift group performances compared to lift groups without an express zone.
For a circulating MCLS with an express zone, the number of cabins can be easily increased to
maintain a low cycle time, so HC5 and average waiting times can be maintained.
5 USER INTERFACES
The user interface of lift groups depends on the control type. Conventional control (collective
control, two button control) [13] and destination control [18] are widely applied. Their user
interfaces have different components and setups.
Lift users differ from those of other transportation systems. At train platforms serving multiple
lines, it is common for not everyone to take the train next to depart. Some passengers wait for a
following train as instructed by a departure board. Is the same scenario, breaking rule 6 of section
2, possible with lifts? If adopted, alternative means of indication would give the control system
more options to improve HC5 and QoS.
Lift user interfaces need to be as simple as possible and support passenger expectation. However,
they are likely to evolve in the future as new technologies enable new passenger guidance systems
for the wider transportation industry.
6 CONTROL ALGORITHM
6.1 Control levels
The control of a group of lift cabins to serve registered landing and car calls can be divided into two
levels [19]. The higher level (group control) lift dispatching problem can be considered as an
assignment problem. The lower level (car/cabin control) is self-contained, can be treated as a
travelling salesman problem and is traditionally solved with collective control [13]. For a
circulating MCLS using one shaft for cabin movement in one direction and the other shaft for the
opposite direction, the concept of collective control can be applied. The rules outlined in section 2
need to be applied by MCLS control algorithms. In MCLS additional control tasks need to consider
the mutual interaction between cabins. Therefore, it is necessary to expand the group control level
to introduce a third system/loop control level as shown in Figure 5. The system/loop control
coordinates multiple cabins within a MCLS loop.
Figure 5 Control levels of a MCLS
The tasks of the different control levels in a MCLS can be described as followed:
Car control: The traditional task of the car control is answering allocated calls as well as
controlling the door operation. Motion control is supported by a propulsion system.
System control: System control ensures that safety distances [12] are not violated. It specifies
speed patterns and controls the loop internal synchronization of the cabins. It also coordinates the
process of bringing new cabins in and out of the loop if the number of cabins can be adapted due to
traffic intensities. System control considers the car control behaviour.
Group control: Group control allocates landing or destination calls to cabins considering system
control behaviour and car control behaviour. It indicates to the system control how many cars are
needed and what cycle time is needed. It synchronizes different loops if necessary.
6.2 Lift control types
The control types (conventional control, destination control and mixed control) are linked to their
user interfaces. The control systems and their user interface are widely applied. Both conventional
and destination control can be an option for a circulating MCLS.
Conventional control: In conventional control systems a lift cabin can be called with an up or a
down direction push button on each landing. The dispatchers allocate lifts from a lift group to
answer the landing calls. The destination of the passenger is registered inside the cabin with car call
buttons. The advantages of using conventional control with circulating MCLS are that most people
are familiar with the user interface, especially in public places. Passengers will fill the next arriving
cabin in their travelling direction to a maximum that is culturally acceptable, and register car calls
inside the car. Individual stops of the cabins, particularly due to car calls, are not under the control
of the control system. So, to avoid traffic jams, times between subsequent cabins need to be high.
Longer cycle times reduce HC5. However, if the number of passengers per cabin is low and the
number of floors served is small, the probable number of different destinations and stops of cabins
is limited. Conventional control could be the preferred control system as it is easy to use for
passengers with the disadvantage of higher cycle times and its effect on HC5. If cycle times are too
low then traffic jams are probable.
Destination control: Destination control systems allow passengers to register their destination on
the floor. Passengers are allocated to lifts. The registration of a car call is not necessary as the
system already knows where the passenger wants to go. The benefit of using destination control for
circulating MCLS is that the control system knows the destination stops before passengers enter the
cabins. The control of movement and synchronisation of cabins using the same shafts can be
optimised to reduce cycle time and increase HC5. One of the main advantages of destination control
is that passengers with the same destination are grouped and allocated to the same lift cabin.
Passengers have fewer intermediate stops while travelling inside the car. If a lift group has two 2-
shaft systems, the MCLS dispatcher has only the choice between two shafts. The “grouping” effect
will be minimal. If in the future appropriate user interfaces (see section 5) meant that the MCLS
dispatcher was not limited to allocating the next cabin in a shaft (breaking rule 6 of section 2), its
options would increase.
Dynamic destination control: The benefit of current destination control systems is that they group
passengers together to reduce the number of stops. Dynamic destination control would require
passengers to register their destination, but then direct them to take the next lift travelling in their
direction. Car call registration would not be required. The advantage to the MCLS dispatcher would
be that it would not need to commit early to an allocation, and would have passenger destination
information in advance to help it optimise the synchronisation of cars using the same shafts.
6.3 Dispatching
Dispatching algorithms use cost functions to choose the most appropriate call allocation. Waiting
time and transit time of passengers are known cost variables. The degradation time of existing
passengers caused by an allocation is considered. Mutual interactions between cabins and departure
delays caused by the loop/system behaviour may affect costs as passengers waiting or travelling in
all cabins of a loop are affected. Every allocation may affect passenger’s satisfaction (QoS) as well
as the synchronisation of cabins and the cycle times within a loop affecting the HC5. Therefore, a
key role in multicar dispatching is the loop/system control responsible for coordinating multiple
cabins using the same shafts.
6.4 Synchronisation
If a cabin is using a shaft exclusively there is no need for any coordination between cabins to avoid
traffic jams or departure delays. In a MCLS the dispatcher needs to synchronise and coordinate
cabins to avoid traffic jams and minimise departure delays. The bunching effect [20] seen in roped
lifts causes traffic jams in a circulating MCLS as cabins using the same shafts cannot bypass each
other. Cabins need to be equally spaced with sufficient time between following cabins. Early traffic
controllers dispatched cars from the main entrance with a fixed time between departures [13]. If the
bunching effect is low and cabins are evenly distributed a spatial plot of a 3-car lift group can look
similar to 3 cars circulating in a MCLS, see Figure 6.
Figure 6 Spatial plot of three cars
Anti-bunching mechanisms need to be applied to MCLS to coordinate cabins within the same loop.
These mechanisms should not confuse passengers by breaking the rules given in section 2. To
achieve this, the car control needs to be able to receive commands to modify its standard behaviour
as follows:
Flexible speed patterns: In order to delay or speed up a cabin the speed pattern may be modified.
For example, if a cabin is ahead of schedule it can start a trip with a lower velocity to delay the
arrival at its next stop.
Modify door opening/closing times: To delay or speed up a cabin departure the door opening and
closing times may be slightly modified to vary the time of a stop without passengers noticing.
Modify door dwell: To change a departure of a cabin the door dwell may be modified. This
departure delay should be realised by an extension of the door dwell when passengers are inside the
car before the doors start closing.
If no passengers are in the cabin additional strategies can be applied:
Delay door openings: It is more confusing entering a lift cabin that does not depart than waiting in
the lobby. So, although a cabin is already at an arrival floor of a waiting passenger, the door
opening may be delayed. If the passenger is aware of the waiting cabin behind the shaft door this
strategy will not work, but will confuse and annoy.
Additional stops: Additional stops can help to delay cars during their round trip.
Departure delays: Cabins can be delayed by simply delaying their departure.
Additional means to control the synchronisation and coordinate multiple cabins are:
Passive area/stock: With an exchanger below the main lobby as shown in Figure 1, a cabin can be
ready to be dispatched to the main lobby at any time. The landing below the main lobby is a passive
area with no passenger transfer and can be used as cabin stock. If a cabin is delayed in the down
direction shaft a waiting cabin can still be used to serve the main lobby in the up direction shaft.
Middle exchangers: Exchanger units in the middle of the shaft enable cabins to short cut the round
trip.
7 CONCLUSION
Operation of multiple lift cabins in multiple shafts needs to consider lift passengers’ expectations.
Accepted rules of lift behaviour have been expanded to cover situations with mutual interaction
between cabins. Reliability is very important as if one cabin breaks down it will block other cabins.
Strategies for resuming operation after a breakdown are necessary.
Safety distance and QoS constraints affect HC5 if MCLS are used as local group. The effect of
special MCLS configurations on QoS and HC5 has been discussed. Both conventional control and
destination control with their user interfaces could be applied to a MCLS but their effect on HC5
and QoS needs to be considered and further analysed. The control system needs to be expanded by a
loop/system control. QoS, HC5, system configuration, and user interfaces need to be considered in
the development of MCLS controls.
REFERENCES
[1] Elevator World, (1996) An elevator go round. Elevator World, (January), pp. 42
[2] Jappsen, H. (2002) High Rise Elevators For The 21st Century. In: Elevator Technology 12,
Proceedings of Elevcon 2002. The International Association of Elevator Engineers.
[3] Godwin, A. (2010) Circular transportation in the 21st century (without the 'beautiful'
counterweight!). In: Elevator Technology 18, Proceedings of Elevcon 2010. The International
Association of Elevator Engineers.
[4] ThyssenKrupp Elevator AG (2014) New era of elevators to revolutionize high-rise and mid-
rise construction [online]. Available from: http://www.urban-hub.com/ideas/new-era-of-elevators-
to-revolutionize-high-rise-and-mid-rise-construction/ [Accessed 04/20, 2015].
[5] So, A., Al-Sharif, L. and Hammoudeh, A. (2014) Analysis of Possible Two Dimensional
Elevator Traffic Systems in Large Buildings. In: Elevator Technology 20, Proceedings of Elevcon
2014. The International Association of Elevator Engineers.
[6] King, F., Hesselgren, L., Severin, P., Sveder, P., Tonegran, D. and Salovaara, S. (2014) The
Articulated Funiculator. In: Elevator Technology 20, Proceedings of Elevcon 2014. The
International Association of Elevator Engineers.
[7] So, A., Al-Sharif, L. and Hammoudeh, A. (2015) Traffic analysis of a simplified two-
dimensional elevator system. Building Services Engineering Research and Technology. 36 (5), 567-
579.
[8] Jetter, M. and Gerstenmeyer, S. (2015) A Next Generation Vertical Transportation System.
In: Wood, A. & Gabel, J. (eds.), The Future of Tall: A Selection of Written Works on Current
Skyscraper Innovations. Addendum to the Proceedings of the CTBUH 2015 International
Conference, New York, 26–30 October 2015. Chicago: Council on Tall Buildings and Urban
Habitat.
[9] Gerstenmeyer, S. and Peters, R. (2015) Lifts without ropes: how many shafts and cars are
needed? In: Symposium on Lift and Escalator Technologies. Northampton:
[10] Closs, G. D. (1970) The computer control of passenger traffic in large lift systems. PhD
Thesis, The University of Manchester Institute of Science and Technology.
[11] Smith, R. and Gerstenmeyer, S. (2013) A review of Waiting Time, Journey Time and
Quality of Service. In: Symposium on Lift and Escalator Technologies. Northampton:
[12] Gerstenmeyer, S. and Peters, R. (2016) Safety distance control for multi-car lifts. Building
Services Engineering Research and Technology [online], Available from:
http://bse.sagepub.com/cgi/content/abstract/0143624416642266v1 [Accessed 06/06, 2016].
[13] Barney, G. (2003) Elevator Traffic Handbook. London: Spoon Press.
[14] Levy, D., Yadin, M. and Alexandrovitz, A. (1977) Optimal control of elevators.
International Journal of Systems Science. 8 (3), 301-320.
[15] Siikonen, M. (1997) Planning and Control Models for Elevators in High-Rise Buildings.
Research Reports A68. Helsinki University of Technology, Systems Analysis Laboratory.
[16] Strakosch, G. and Caporale, R. (2010) The Vertical Transportation Handbook, Fourth
Edition. Hoboken; New Jersey: John Wiley & Sons, Inc.
[17] Al-Sharif, L., So, A., AlZoubi, Q., Atallah, T., Hijazi, A. A. G. and T.Hammoudeh, A.
(2016) Collision prevention procedure for a rectangular two-dimensional elevator traffic system
using graphical methods. In: Elevator Technology 21, Proceedings of Elevcon 2016. Madrid/Spain:
The International Association of Elevator Engineers.
[18] Smith, R. and Peters, R. (2002) ETD Algorithm with Destination Control and Booster
Options. In: Elevator Technology 12, Proceedings of Elevcon 2002. The International Association
of Elevator Engineers.
[19] Sorsa, J. S., Ehtamo, H., Siikonen, M., Tyni, T. and Ylinen, J. (2009) The Elevator
Dispatching Problem. Submitted to Transportation Science. September 2009.
[20] Al-Sharif, L. (1993) Bunching in lift systems. In: Elevator Technology 5, Proceedings of
Elevcon 1993. Vienna/Austria: The International Association of Elevator Engineers.
Stefan Gerstenmeyer has been working as Senior Engineer and Head of Traffic and Group Control
at ThyssenKrupp Elevator Innovation GmbH. He has been involved in R&D projects relating to
group and dispatcher functions for lift controls including multi car lift systems. He is a post
graduated research student at the University of Northampton.
The 1935 Code of Practice for the Installation of Lifts and
Escalators
Dr. Lee E. Gray
College of Arts + Architecture
University of North Carolina at Charlotte
Charlotte, North Carolina USA 28223
[email protected]
Keywords: Lift and Escalator Codes
Abstract. The 1935 Code of Practice for the Installation of Lifts and Escalators was written by the
Lifts and Escalators Installation Panel of the Building Industries National Council. The thirteen-
member panel included representatives from the lift industry, insurance industry, trade unions, the
Royal Institution of Chartered Surveyors, and the Royal Institute of British Architects. Prior to 1935
there was no British national code or national legislation, beyond the Factory and Workshop Act,
which governed lift and escalator installation. Thus, the panel looked outside Britain for precedents
and they reported that they examined “all existing Codes … in force on the Continent of Europe, in
America and in several British Dominions” [1]. The new Code was described as offering “safety
and protection to all users” while also ensuring that it “would not encroach upon design and
unnecessarily or impede engineering progress” [1]. The authors’ collective goal was to develop a
system of “coordinated safety regulations having reasonable flexibility” that “would avoid the
difficulties inherent in official or departmental control per se, and would at the same time meet all
reasonable demands for safety” [1]. This, perhaps contradictory, goal was achieved in a mere 35
pages of text and one illustration. This paper will examine the membership of the Lifts and
Escalators Installation Panel, the Code’s contents, and its American and European precedents.
1 INTRODUCTION
The history of lift and escalator codes remains a relatively unexplored topic in the history of vertical
transportation. Although references to this history are often found in the introductions to new or
revised editions of existing codes, these typically consist of a brief outline of the full, and often
complex, story of the code’s origins and authors. Writers charged with revising an existing code
must, out of necessity, understand the rationale and reasoning that produced the earlier edition. This
activity often represents a pragmatic rather than a historical understanding of the prior work.
However, the decision to write a first lift code speaks to a particular moment in time. The
subsequent changes that occur in following editions constitute evidence of changes in technology,
use patterns, and the culture of vertical transportation. The publication of the Code of Practice for
the Installation of Lifts and Escalators in 1935 marked a unique moment in time for Great Britain,
as this represented the first attempt to write a British national code. At the same point in time, it was
also produced within the context of a brief, but none-the-less well established, international history
of lift codes and regulations. Beginning in the early 1900s lift codes and installation guidelines had
been or were being developed in the United States, Germany, Italy, France, Finland, Belgium, and
The Netherlands. Thus, the authors of the first British code had history on their side, with the
established precedent of the need for a national code, and they also had recent history as a guide in
the presence of existing codes, which they utilized to determine the proper content and tenor of
their new national code.
2 THE LIFTS AND ESCALATORS INSTALLATION PANEL

In September 1931 the Advisory Committee on Building Acts and Byelaws of the Building
Industries National Council established the Lifts and Escalators Installation Panel (Table 1), which
was charged with reviewing existing legislation concerning lift and escalator installation.
Table 1. Lifts and Escalators Installation Panel.
Member Representing
1
Leonard Stewart Atkinson, A.M.I.E.E. Co-opted member
2
Rendell Davies, M.I H.V.E. W. MacIntyre, Consulting Engineers
Murray Easton, F.R.I.B.A. The Royal Institute of British Architects
Alfred Harold Edwards Redpath, Brown, Ltd.
1
David W. Rolfe Green Waygood-Otis, Ltd.
Matthew T. Greenwell Electrical Trades Union
3
Edward Charles Harris, F.S.I. The Chartered Surveyor’s Institution
1
Ernest Matthew Medway J. & E. Hall, Ltd.
1
W.W. Pattinson Insurance Companies
1
Edwin Charles Stevens, M.I.M.E. Institution of Mechanical Engineers
1
John William Stevens The Express Lift Company
1
William Wellesley Weaver Waygood-Otis, Ltd.
1
Member of the Code Drafting Subcommittee
2
Chair, Code Drafting Subcommittee
3
Chair, Lifts and Escalators Installation Panel
The membership of the Lifts and Escalators Installation Panel represented an intriguing cross
section of the lift industry and engineering profession. The Panel chair, Edward C. Harris (1883-
1966), was a quantity surveyor who apparently had no direct connection with the lift industry.
However, he had founded EC Harris in 1911, one of the first multi-industry consultancy firms, and
thus he had a broad perspective on the building industry. Other non-industry members included
Matthew T. Greenwell (representing the Electrical Trades Union), Murray Easton (representing the
Royal Institute of British Architects), Alfred Harold Edwards (a structural engineer with Redpath,
Brown, Ltd.), W.W. Pattinson (representing the insurance industry) and Rendell Davies (1891-
1941) (a consulting engineer with W. MacIntyre). Of this group, only Pattinson and Davies were
selected by Edwards to serve on the Code Drafting Subcommittee.
Harris selected Davies to chair the drafting subcommittee. Davies was a member of the Institution
of Heating and Ventilating Engineers, worked as a consulting engineer in London, and was
associated with the British Standards Institution. He was also one of the youngest members of the
subcommittee, which, as will be seen, represented two distinct generations. The three other
members of the younger generation were Leonard Stewart Atkinson, William Wellesley Weaver
(1890-1947) and John William Stevens (1887-1954). Atkinson had been co-opted to the committee
from the Institute of Electrical Engineers. He had joined Waygood-Otis as an apprentice in 1914
and by the early 1930s he had advanced to the position of Assistant Chief Engineer. Weaver had
joined Waygood-Otis as an apprentice in 1907 and was appointed managing director in 1933.
Following his military service in Word War I Weaver had traveled extensively on behalf of the
company, working for one year in India, two years in Australia, and one year in New York. Stevens
had begun his career in 1900 as an office boy in the firm of Easton, Anderson and Goolden, the
successors to Easton & Anderson (who built the Mersey Railway Elevator System). In 1904 he
joined the newly founded Easton Lift Co., Ltd. who, in partnership with the General Electric
Company, Ltd., founded the Express Lift Co. in 1917. Stevens served as managing director of
Express Lift from 1923 to 1936.
The 1935 Code of Practice for the Installation of Lifts and Escalators 9-3
The older generation was represented by Ernest Matthew Medway (1875-1955), Edwin Charles
Stevens (1869-1952) and David W. Rolfe Green (1871-1942). These members also represented
three of Britain’s oldest lift firms. Medway was the son of Matthew Thomas Medway (1850-1915),
who founded the Medway Safety Lift Co. in 1878. In 1926 J. & E. Hall, Ltd. acquired a controlling
interest in Medway and by 1935 the older firm had been fully assimilated into J. & E. Hall and the
name Medway was no longer used. Stevens (no relation to John William Stevens) was the son of
John Sanders Stevens who, with Archibald Smith, had founded Archibald Smith & Stevens in 1880.
The company became Smith, Major and Stevens, Ltd. in 1909, at which time its manufacturing
plant was moved to Northampton. By 1922 Edwin Stevens was serving as Chairman and in 1930
the company was amalgamated by the Express Lift Co. Green was the son of William R. Green
(1838-1910), who had joined R. Waygood & Co. in the early 1860s (Waygood was his Uncle).
David Green began his career as a Chartered Accountant and he joined Waygood in 1886 as an
assistant to company co-founder Herbert C. Walker (1852-1939). In 1933 Green was elected
Chairman of Waygood-Otis.
Thus, the subcommittee members brought approximately 200 years of experience in the lift industry
to their assigned task. They also had experience working in six different lift companies of various
sizes: the Easton Lift Co., the Medway Safety Lift Co., Smith, Major and Stevens, J. & E. Hall, the
Express Lift Co., and Waygood-Otis. However, in 1931, Waygood-Otis clearly dominated the
subcommittee’s membership. Therefore a critical question, given the bifurcated nature of the firm:
Waygood-Otis or, as seen through another lens, British-American, concerns the significance of the
role that Otis and/or the American lift code played in writing the first British lift and escalator code.
3 PRECEDENTS
The Panel was charged with the review of “such legislation as affected the installation of lifts and
escalators in buildings and to report on the need for revision thereof and the form such revision
should take” [1]. However, they quickly shifted their focus beyond the revision of existing
legislation and, as they reported in 1935, their “enquiry was devoted to formulating a code of lift
and escalator practice” [1]. A key part of this investigation was the examination of “all existing
Codes and Glossaries available, including those in force on the Continent of Europe, in America
and in several British Dominions” [1]. The Panel also “examined the publications of the British
Standards Institution” [1].
Unfortunately, no record or list has been found of the resources examined by the subcommittee. The
only English code precedent was the Factory and Workshop Act: first drafted in 1901 and amended
on a regular basis. However, this limited set of regulations primarily concerned goods or freight lifts
and only addressed lifts in industrial settings. Other possible resources included The Protection of
Hoists, Safety Pamphlet No. 2 (H.M. Stationery Office, London: 1919: third edition 1924) and
British Standards Specification for Round-Strand Steel Wire Ropes for Lifts and Hoists, No. 329
(British Standards Institution: 1928).
Possible European precedents identified thus far include guidelines and regulations drafted in
Germany, Italy, and France. A chronological list of these works (Table 2) reveals that Germany and
Italy produced some of the earliest regulations and that most of documents addressed both the
installation and operation of lifts. The only precedent from the British Dominions discovered thus
far is the South Australian Lifts Regulation Act of 1908 (An Act to Regulate the Use of Passenger
and Other Lifts). This Act primarily concerned lift inspections and contained no technical
specifications (its only operational statute was to set a minimum age of 18 for all lift operators). The
American code precedent was the 1931 edition of the American Standard Safety Code for
Elevators, Dumbwaiters and Escalators.
Table 2. European Lift Guidelines/Regulations 1908-1927
4 ORGANIZATION
The Code Drafting Subcommittee spent approximately four years working on their assigned task,
which was completed in 1935. The Code of Practice for the Installation of Lifts and Escalators
featured a Forward by Sydney Tatchell (1877-1965) F.R.I.B.A. and President of the Building
Industries National Council, a Preface by Edward C. Harris, a brief table of contents, the code, and
a detailed index. The code was divided into three sections: a glossary that defined 94 terms, the
Code of Practice for Lift Installation with 24 regulations, and the Code of Practice for Escalator
Installation with 11 regulations (Table 3). Many of the lift and escalator regulations were divided
Table 3. 1935 Lift and Escalator Code Sections.

into sections and subsections, which resulted in a total of 182 individual rules or recommendations.
The code also included two tables and one illustration. The tables addressed the minimum top and
bottom clearances for cars and counterweights and the maximum stopping distances allowed for
cars equipped with Gradual Wedge Clamp (G.W.C.) and Flexible Guide Clamp (F.G.C.) safeties.
The illustration was a schematic section of a typical electric lift installation with the lift operating in
a stairwell (Figure 1).
Figure 1. Typical Lift Installation, Code of Practice for the Installation of

Lifts and Escalators (1935).
5 ANALYSIS
A comparative analysis of the Code of Practice for the Installation of Lifts and Escalators reveals
that the primary source for the new code was the American Standard Safety Code for Elevators,
Dumbwaiters and Escalators of 1931. A mapping of the codes’ contents reveals that 22 of the 24
lift regulations and all 11 of the escalator regulations had counterparts in the American code (Tables
4 & 5). A detailed analysis found that 89 of the 157 lift-regulation sections and subsections and 24
of the 25 escalator-regulation sections and subsections had American code counterparts. Finally, the
British code included 94 terms in its glossary and the American code defined 84 terms. It is of
interest to note that only 25 common terms appeared in these glossaries. However, while many of
the British lift regulations had American precedents, in many cases there were also key differences.
The American code’s influence included the use of identical text in the new British code, the use of
slightly modified text, and the substantial rewriting of parallel sections intended to reflect local
building and industry practices. Examples of the literal influence of the American code include
technical guidelines, illustrated by the British regulation Suspension Ropes 5d and American Rule
Table 4. Lift Code Comparison
Table 5. Escalator Code Comparison
230g Cables, both of which read as follows: “No car or counterweight cable shall be repaired or
lengthened by splicing” [1, 2]. Another essentially literal example reflects differences in attitudes
toward appropriate emergency lift use between the 20th and 21st centuries. The American Rule 100a
Fire-Resistant Hoistway Enclosures included the following:
Note: Experience has demonstrated the value of the elevator as a life-saving device in
case of fire. A simple form of fire-resistant construction (cement plaster on metal
lath) will usually resist a fire for a greater length of time than the elevator can be used
as an exit from a burning building. Fire- resistant hoistways are therefore
recommended for all elevators. [1]
This was translated into the British code as follows:
Lift wells, together with the whole of the contained equipment, apparatus, etc., shall
be rendered fire resisting to the greatest possible extent. Note. Experience in the
U.S.A. has demonstrated the value of the lift as a life saving device in case of fire. A
simple fire resisting construction will usually resist a fire for a greater length of time
than the lift can be used as a means of escape, and for this reason the above
recommendation is made. [2]
In the majority of cases the British code retained the essence of the American precedent, which was
often expressed in a simplified and edited manner. The following example illustrates this strategy:
Rule 210i Car Construction: When car-leveling devices are used the car platform
shall be provided with a substantial vertical face flush with its outer edge, extending a
sufficient distance below the car floor so that there shall be no horizontal opening into
the hoistway while the car is within the landing zone and the hoistway door is wholly
or partially open. [2]
Lift Cars 7h: Where car leveling devices are used, aprons shall be fitted to the car
floor to ensure that no space is permitted between the threshold and the landing whilst
the car is being leveled to a floor. [1]
Although there was a higher degree of synchronicity between the two escalator code sections, a
similar editing process also occurred:
Rule 602a Balustrading: Escalators shall be provided on each side with “solid
balustrading.” On the escalator side the “balustrading” shall be smooth, without
depressed or raised paneling or molding. Glass panels in “balustrading” are
prohibited. There shall be no abrupt changes in the width between the “balustrading”
on the two sides of the escalator. Should any change in the width be necessary, the
change shall be not more than eight (8) percent of the greatest width. In changing
from the greater to the smaller width the change in the direction of the “balustrading”
shall be not more than fifteen (15) degrees from the line of the escalator travel. [2]
Balustrading 7a: Escalators shall be provided on each side with solid balustrading.
On the escalator side the balustrading shall be smooth, without depressed or raised
panelling or moulding. Glass panels should not be used in balustrading. [1]
However, the British code was not simply a well-edited version of its American precedent (with 35
versus 173 pages): it included sections and information not found in the earlier code and reflected
critical differences in lift culture. The maximum speed referenced in the British code was 800 feet
per minute, while the American code referenced speeds up to 1,600 feet per minute. The British
code also included a section titled “Shafts” that referred to shafts that held sheaves and pulleys:
Shafts 18: (a) Any shaft carrying a sheave or pulley and fitted between dead eyes or
other housing must be stepped, i.e., reduced in diameter, at or near the point of entry
at each end. (b) Any shaft where stepped, i.e. reduced in diameter, must be turned to a
reasonable radius at the point of reduction in diameter.
It is unknown why the drafting subcommittee felt it was necessary to include such a detailed
recommendation on this particular aspect of lift technology.
The issue of lift inspection was also treated very differently in the two codes. The American code
stated that: “Responsibility for the care, operation, and maintenance should be definitely fixed by
statute or ordinance. Where not so fixed, it is recommended that leases for buildings specify such
responsibility as between owner and lessee” [2]. The British code stated that: “Every power
driven lift, before being put into service, should be covered by insurance, such insurance cover to
include for and incorporate regular inspections at least three times per annum by a representative of
the insurance office” [1]. The suggested preferred inspection protocol was further defined as
follows:
Rule 701 Inspection: The following is the schedule of inspections recommended:
Hoistway doors, car gates, interlocks, contacts, control apparatus, controller,
automatic stop, limit stops, car and counterweight cables, “safeties,” guide rails,
buffers, elevator machines, and the lighting of the car and of the machine room, in
passenger and freight-elevator installations, shall be thoroughly inspected at least
quarterly. [2]
Inspection, Maintenance and Insurance 8c: At least once in every three years the
safety gear and governor switch, if fitted, should be subjected to a running test under
maximum load and speed conditions, and a certificate issued on the result of each
test. Such certificate in its most effective form would be signed by the insurance
engineer supervising the test. [1]
The references to insurance companies and insurance engineers speaks to the drafting committee’s
hope that the insurance industry would play a primary role in code enforcement: “having regard to
the very deep material interest of the insurance offices in lift and escalator installation, the code of
lifts and escalator practice might, with advantage, be operated under their aegis” [1].
6 CONCLUSION
The goal of the Lifts and Escalators Installation Panel was to write a code whose implementation
would avoid the problems encountered in other countries: “It is felt … that wherever control of the
mechanical equipment of buildings is vested solely in official bodies or departments, such control
must of necessity tend to become rigid and to retard progress” [1]. The Panel also recognized that
codes “of this nature must be subject to review from time to time, as by evolution both materials
and machinery are improved and requirements change” [1]. They therefore recommended that “a
tribunal” be established “to observe the effect of the Code in operation and to suggest such interim
amendments as may prove necessary” [1]. The speed of change in the lift industry was such that the
effort to revise the 1935 code began in 1940/41 and the second edition, titled the Code of Practice
for Electric Passenger and Goods Lifts and Escalators, was published in May 1943.
REFERENCES
[1] Building Industries National Council, Code of Practice for the Installation of Lifts and
Escalators, London (1935).
[2] American Standards Association, American Standard Safety Code for Elevators,
Dumbwaiters and Escalators, New York (1931).
Dr. Lee E. Gray is the Senior Associate Dean in the College of Arts + Architecture at the University
of North Carolina at Charlotte and a Professor of Architectural History in the School of
Architecture. He received his Ph.D. in architectural history from Cornell University, his Masters in
architectural history from the University of Virginia, and undergraduate degrees in architecture
from Iowa State University. He is the author of From Ascending Rooms to Express Elevators: A
History of the Passenger Elevator in the 19th Century. Since 2003 he has written monthly articles on
the history of vertical transportation for Elevator World magazine. Current projects include a book
on the history of escalators and moving sidewalks.
London Underground Escalator Passenger Safety Strategy
Improving Passenger Safety on Escalators
Phil Harley, Kevin Seaborne
15 Westferry Circus, Canary Wharf, London E14 4HD, UK. [email protected]
Victoria Station House, 191 Victoria St, London SW1E 5NE, UK. [email protected]
Keywords: Escalator, Safety, Accident, Behaviour, London Underground.
Abstract. Safety, reliability and value for money are at the foundation of London Underground’s
(LU) service. Although LU’s safety record is one of the best in the world it is important to guard
against complacency and continue to strive for continuous improvement.
Evaluation of company incident data has identified that the largest cause of customer injuries on
London Underground is slips, trips and falls with 40% of all injuries reported occurring on
escalators.
A strategy group was set up, comprising all stakeholders across the business, to identify/shortlist
and then trial ideas to improve passenger safety. Four key risk controlling measures were identified
and from the list of ideas 12 were shortlisted for trial on more than 50 escalators at stations with
historically high accident rates. The aim was to identify ideas that would positively impact
passenger behaviour and in turn reduce the number of slips trips falls and entrapments.
Four measures were used to evaluate the effectiveness of each initiative. Seven of the twelve ideas
were found to be effective, following evaluation of the four measures.
The next step is to roll out, in a targeted manner, selected initiatives to “Top 20” London
Underground station assets where the highest number of accidents and incidents have occurred in
previous years.
1 INTRODUCTION
An escalator is an inherently dangerous machine by modern safety / engineering standards, in that
passengers are directly in contact with moving powered machine parts, with minimal guarding
between moving steps and static landings and balustrades. The high prevalence of slip / trip / fall
incidents on escalators, along with less regular but potentially far more serious entrapment
incidents, are mainly caused by customer behaviour.
There are various issues where the passenger is exposed to risk and adopts behaviours which do not
best mitigate these risks.
• To avoid entanglement, entrapment or risk of fall, the safest place to stand on the step is
with feet equidistant from the front and rear edges, and away from the edge of the step and
the brush guard.
• The most dangerous area of the escalator is the landing where the interface between the
static landing and the moving step way is protected by the comb plate.
• Holding the moving handrail whilst transiting the escalator is the most reliable way to
reduce the risk of slips, trips and falls.
• Many accidents occur on escalators due to passengers being mobility impaired, whether by
carrying luggage, holding young children, or due to age or infirmity.
2 FORMING THE STRATEGY

The intent of the Escalator Passenger Safety Strategy (EPSS) is to improve both reliability and
safety of the customer experience; the benefits are primarily social with fewer customers injured.
Financial savings are seen in reduced compensation payments, and reduced costs due to diversion
of staff from primary duties. Following an accident, escalators are often removed from service for a
period pending inspection and therefore inconveniencing other customers, and any reduction in
accidents will logically reduce this necessity. However, the primary benefit is to demonstrate
commitment to safety and high standards of customer care.
The purpose of the EPSS is to identify effective methods of influencing passengers’ behaviour to
increase their safety on escalators. The Imperial College London CoMET 2010 Case Study, states:
“It is striking that two metros – Metro A & Metro B1 – have consistently and progressively
reduced the number of falls (on escalators) as a result of systematic management attention,
appropriate investment and good campaigns to persuade passengers to avoid behaviour that
would put them at risk” [1]
A census of CoMET members was conducted in 2013 to determine best practice in other Metros
which was taken into account when determining which customer behaviours should be encouraged.
Due to the high prevalence of injuries on escalators not due to machine failure, there is a need to
inform the passenger either directly or subliminally of safe practices for transiting escalators. This
can be broken down into four main themes, or mitigating messages:
• Hold the handrail.

• Walk / stand safely.
• Be aware of the step / landing interface.
• Where possible, use lifts when mobility impaired.
These four core messages are embodied in the 12 different initiatives, technical interventions with
the specific purpose of altering passengers’ behaviour either by direct information or subliminal
coercive “nudges”.
3 FORMULATION AND DEVELOPMENT OF THE EPSS COMMITTEE PLAN

The strategy is designed as a loose framework document. It caters for ongoing initiatives to be
overseen and directed, and also for the scoping and support for as yet unstarted projects. In order to
ensure stakeholder support for this strategy the EPSS committee was convened to propose & agree
a system wide risk based approach, with representatives invited from the following stakeholders:
• LU Engineering
• LU Stations Maintenance
• Projects Directorate
• Strategy & Service Development
• Health, Safety & Environment, LU Ops
• Technical Head of Discipline, Lifts & Escalators
• LU Stations Operations
________________________________________________________________________________
1
CoMET is a group of Metros across the world which cooperate in Benchmarking. CoMET reports are subject to
confidentiality. It is not permissible to name participating organisations in public documents.
London Underground Escalator Passenger Safety Strategy Improving Passenger Safety on Escalators 10-3
A plan was formulated to develop all selected initiatives and install in suitable locations by 1st
September 2015. It was agreed at an early stage that buy in from across the network would be
essential. To this end, presentations to stations personnel informing them of the EPSS aims and
progress have been carried out, along with the production of publicity in the form of articles in “On
The Move” magazine and other internal media.
4 SUCCESS CRITERIA
The criteria for judging an initiative successful has been deemed to be a reduction of 10% of
escalator related accidents. Due to the small scale of the trials and the resultant low level of
statistical data available, this cannot be fully inferred from review of reported accidents on
escalators. To compensate for this, it was agreed that reduction of incidents can also be inferred
from an increase in safe behaviours, or decrease in unsafe behaviours. In addition, survey of staff on
stations where initiatives are installed gives a further perspective on effectiveness, and a technical
review gauges the practicality of each initiative.
5 SELECTED INITIATIVES FOR TRIAL

From a wide array of suggested alternatives, 12 asset initiatives were identified as being most likely
to affect customer behaviour and enhance customer safety. Some of these interventions are new
approaches, whilst others are industry standards.
5.1 Passenger Positional Guides (PPG)
Bi-directional footprints intended to guide passengers in their safe foot positioning whilst riding on
escalators.
5.2 Step Edge Painting
Industry standard painting of step edges to highlight safe area to stand upon.
5.3 Step Riser Messaging

Safety messages stencilled on the black step risers.
5.4 Red Lexan Combs
Red escalator combs manufactured from Lexan polycarbonate, highlighting the interface between
the moving step band and the static landing.
5.5 Under Step Lighting
White light shining up through the gaps between steps at the top and bottom landing to highlight
changing geometry of the step in the transition between the landing and the incline and nearing the
end of the moving stepway.
5.6 Over Comb Lighting
Lighting element installed in the balustrade at foot level directly over the comb to highlight
transition from landing to step band to stationary landing.
5.7 LCD Screens in Pattresses “e-Toblerone”

High definition bi-directional screens displaying safety messages mounted inside Pattresses on the
balustrade between escalators.
5.8 Embedded Handrail signage
Safety messages permanently embedded in the surface of the handrail.
5.9 Virtual Assistant Projector “Hologram”

Mobile “Virtual Assistant” silhouette projector unit to impart safety messages installed near
escalators.
5.10 PA Messaging
Modified PA announcements for particular station areas giving safety messages for a bank of
escalators.
5.11 Escalator Floor Vinyls

Temporary floor signage to encourage people to take caution when using an escalator and hold the
handrail.
5.12 Lift Floor Vinyls
Temporary floor signage to enhance awareness of station lift locations and encourage customers to
use the lift instead of an escalator if mobility impaired.
6 REJECTED INITIATIVES
The following initiatives were investigated, evaluated and then rejected during the progress of the
EPSS:
6.1 Coloured Step Brush Holders
Extruded aluminium holders of escalator brushes are normally unpainted; to highlight the risk of
entrapment in the step edge it was suggested that the brush holder be powder coated red. This was
deselected due to cost and due to insufficient numbers of escalators such an installation would be
practicable on. A report was produced detailing the decision. [2]
6.2 Step Riser Painting

As an initial proof of concept trial, escalator 4 at Heathrow T1-3 had step risers powder coated
yellow during a refurbishment to highlight the step to step gap. This was deselected due to issues
with ambience, and the requirement to remove the escalators from service for a period to effect the
powder coating. A report was produced detailing the decision. [2]
6.3 Directional Indicators
Directional indicators are “traffic light” signals which are intended to inform passengers of the
direction of travel of an escalator (by means of a green arrow or a red stop light). This concept was
investigated and deselected as no evidence was found that (a) it would prevent customers boarding
escalators in the wrong direction, or that (b) that this is a problem which merits intervention. A
report was produced detailing the decision. [2]
6.4 Onboard Train Announcements.

Operational staff suggested the inclusion of train onboard messages informing passengers of the
location of lifts prior to arrival at the station, to increase awareness of lift facilities at stations and to
direct mobility impaired passengers and passengers with heavy luggage to the nearest lift. This was
deselected due to the review highlighting the difficulty involved in the implementation of this
initiative. It also showed no evidence of its effectiveness. Furthermore overloading train passengers
with messages will effectively diminish their intent, rendering existing safety messages ineffective,
and reduction of onboard messaging is a priority for operational staff. A report was produced
detailing the decision. [2]
7 INSTALLATION
Installation of the selected initiatives was carried out by a variety of internal and external bodies,
using assets and manpower provided by either EPSS internal stakeholders or by contractors and
suppliers of specific equipment and services. The majority of installations were completed on
schedule whilst a minority of the more innovative and complex initiatives were delayed due to
supply or technical difficulties.
8 MEASUREMENT OF EFFECTIVENESS
The reason for inclusion of known systems as well as new concepts in the trial is simple: there
appeared to be very little independent assessment of their effectiveness in encouraging safe
behaviours. Therefore as part of the trial it was necessary to establish the effectiveness of both
established and new methods for altering customer behaviour. This measurement of effectiveness
was conducted in four ways: statistical comparison, survey of passenger behaviours, technical
evaluation and survey of station staff.
8.1 Statistical Comparison
The sites for the escalator initiatives trials were reviewed and across the 20 stations with trials in
place there was a reduction in incidents with injuries overall by 3%. This was a comparison of
customer injuries over the trial period of 2015/16 September - March inclusive compared with
2014/15 September - March. The same periods each year were compared to allow for seasonal
fluctuations and holiday periods etc. It was noted there was a reduction in Customer Major Injuries
by 36% (major broken bones, unconsciousness or dislocations) at the trial stations. The data only
includes incidents reported to staff and involve an injury to the customer.
8.2 Survey of Passenger Behaviours
Surveying passenger behaviours to accurately judge the effect of the initiatives was conducted
primarily by collection of CCTV data from before installation, directly after installation, and
following 3 months deployment. Where insufficient quality of image was available from CCTV we
temporarily installed GoPro cameras to record at the same periodicity.
8.3 Technical Evaluation
The technical effectiveness of the initiatives was examined throughout the trial (robustness, wear,
impact on maintenance etc), along with their impact on the Stations’ environments. This process
combined regular visual inspections of the installation sites with collation of faults reported.
8.4 Station Staff Survey

The survey of station staff gives further depth to the analysis, by gaining their subjective insight
into the effect on passengers’ behaviour over the length of the trial. Site specific questionnaires
were produced, referring only to initiatives installed in each station, and were distributed Jan – Feb
2016. Front line staff and station managers were talked to about the initiatives before the trials
started at 6 of the key stations and during the trials. All the stations were emailed and posters sent
out for comms. Over 100 staff were spoken to and there were 92 responses collated from the staff
feedback survey completed between the 3-4 month period of each of the trials. There were mixed
responses from staff and some staff were unaware of the initiatives due to a major reorganisation
having just taken place on stations, resulting in a larger than average number of new staff at the
stations visited.
All station area managers were asked for feedback and staff surveys to be completed. Feedback was
limited so a team went out to interview front line staff. A short survey of staff observations and
feedback was completed covering the following areas:
• Have there been any maintenance changes?

• Any change in failures observed?
• What change in customer behaviour has been observed?
• Is there a change in customer accidents?
• Any issues with the installation?
• Any comments of positive actions or improvements
These four methods of measurement were combined to provide a robust assessment of the effect of
each of the initiatives on passenger safety, their robustness and utility in service, and the practicality
of wider use. The assessments provide evidence to support recommendations for wider deployment
of a specific initiative where warranted. Validation of the measurement of the survey customer
behaviours was conducted by the Customer & Employee Insight Team of the Marketing and
Communications Directorate, and the methodology was found to be impartial and effective.
9 RESULTS
Listed in the table below are the scores on each of the measurements of effectiveness, giving an
overall indication of how effective each of the initiatives was found to be.
Table 1: Summary of Trial Results
Statistical
Analysis Final
(Accident Customer Overall
reduction Behaviour Staff Technical Effective
Initiative >10% 2) (>10% 2) Survey Review Decision
PPG (Blue Footprints) 27% 21% Neutral Pass Yes
Step Edge Painting -29% 2% Negative Fail No
Step Riser Messaging -20% 13% Positive Pass Yes
Red Lexan Combs 36% 15.90% Positive Pass Yes
Under Step Lighting 0% 0.08% Negative Pass No
Top Comb Lighting 0% 2.10% Negative Pass No
e-Toblerones -21% 11.20% Positive Pass Yes
Embedded Handrail Signs 23.80% 17.40% Positive Pass Yes
Hologram 13.10% 19.90% Positive Pass Yes
Speakers / PA Nil Nil Nil Fail No
Escalator Floor Vinyls -9% 7.60% Negative Pass No
Lift Floor Vinyls Vinyls 22% 1.10% Positive Pass Yes
A final report was published and disseminated on completion of the trial, detailing the findings. [2]
______________________________________________________________________________
2
>10% refers to the success criteria of reducing accidents – or unsafe behaviours – by more than 10%.
10 RECOMMENDATIONS
The following recommendations were made in the final report, and were approved by the Customer
Safety Strategy Steering Group:
10.1 Passenger Positional Guides
It is recommended that PPGs should be installed on escalators which are known to run
predominantly in one direction, with identified issues of passenger foot placement causing
accidents, and that repainting should be programmed in on a six monthly basis.
10.2 Step Riser Messaging

It is recommended that “Hold the handrail” step riser messages are installed on escalators with non-
cleated step risers which run predominantly in an upward direction.
It is also recommended that further development of a solution for steps with cleated risers is funded
and managed via the EPSS.
10.3 Red Lexan Combs
It is recommended that red combs continue to be rolled out where applicable over the LUL network
escalator fleet.
It is also recommended that the practicality of extending the use of red combs across all escalators.
10.4 LCD Screens in Pattresses “e-Toblerone”
It is recommended that further development of this initiative is funded, potentially with support
from commercial interests.
It is further recommended that the installation at Piccadilly is retained for 12 months to establish the
long term reliability of the technology.
10.5 Embedded Handrail Signage

It is recommended that message embedded handrail should be identified and approved as the
standard replacement for Shape 400 handrails (non V-type), as part of their programmed
replacement. This should be captured in the CAT1 Standard as a mandatory requirement.
10.6 Virtual Assistant Projector “Hologram”
It is recommended that a small fleet of Virtual Assistant Projectors is purchased, to be used as a
moveable safety messaging resource.
10.7 Lift Floor Vinyls
It is recommended that lift “breadcrumb” vinyls should be installed at Wide Access Gates in station
gate-lines, where confusion over direction to the lifts for mobility impaired customers has been
identified. Signage should be replaced on a six monthly basis.
10.8 Escalator Passenger Safety Strategy Committee
It is recommended that the EPSS committee is retained in its present form to act as an authorising
“clearing house” for future suggested escalator passenger safety initiatives and a management
framework for future projects; reporting to the Customer Safety Strategy Steering Group.
These recommendations now form the basis of the ongoing plan for targeted deployment of
successful initiatives to stations. This is achieved by identification of “top twenty” London
Underground stations assets by examination of historical data where the highest number of
accidents and incidents have occurred in previous years. Selection of a specific initiative for a site is
agreed through consultation between local station operational staff, engineers, maintainers and
S&SD prior to installation.
11 UNSUCCESSFUL INITIATIVES
The following initiatives were found to be unsuccessful in promoting safe passenger behaviours on
escalators
11.1 Step Edge Painting
Painting of yellow step edges on single piece cast steps had no significant measured effect on
customer behaviour or safety. Allied to the restrictions of painting cast steps only and the
requirement for access to the step band in the machine chamber, there is no advantage in pursuing
this initiative.
11.2 Under Step Lighting

Installation of lighting beneath the ends of the moving stepway was found to have no significant
measured effect on customer behaviour or safety. There is no evidence to support any further
deployment of this initiative. However whilst there is no benefit to further installations, there is
little point in removing any under step lighting installed, as there is no disbenefit to their continued
use.
11.3 Over Comb Lighting
Installation of lighting above the escalator landing combs was found to have no significant
measured effect on customer behaviour or safety. There is no evidence to support any further
deployment of this initiative. However whilst there is no benefit to planning further installations,
there is little point in removing any under step lighting installed. There is no disbenefit to their
continued use, and to do so would require the escalators’ removal from service whilst the removal
was effected.
11.4 PA Messaging
Recorded PA announcements specific to escalator safety were installed in the PA system for the
northern concourse of Kings Cross St Pancras. Initially the recordings were played, but were found
by station staff to be “cluttering” the PA, and the messages were removed. These initiatives were
not fully submitted to trial, and results did give a representative view of their effect on customer
behaviours. However the intent of station staff to minimise “clutter” on PA systems in stations is
widespread, and as such precludes further development of this initiative.
11.5 Escalator Floor Vinyls
Although initially promising, escalator floor signage had only a temporary effect on the behaviour
of passengers. An initial 15.9% increase in safe behaviours (holding the handrail), reduced over the
course of the trial to 7.6%. Over the trial period reported accidents increased by 8%. Concern with
passengers being distracted when approaching the escalators, and likelihood of the message being
obscured in busy periods was raised. There is insufficient evidence to support any further
deployment of this initiative.
12 CONCLUSION
The collaboration between the wide and disparate grouping of stakeholders across London
Underground resulted in the delivery of twelve discreet initiative designs which have been trialled
on operational escalators and stations. The EPSS committee is a useful conduit for the review,
discussion and approval of any suggested improvements to customer safety on escalators, under the
authority of the Customer Safety Strategy.
In development of the EPSS involvement has been sought from a wide array of external
organisations, including the Community of Metros (CoMET, the international forum for
metropolitan rail transport providers), the Construction Industry Research & Information Agency
(CIRIA), Health & Safety Laboratories (HSL, an agency of the Health & Safety Executive) and the
Office of Rail & Road (ORR), along with a number of manufacturers and suppliers.
It is intended that the final report will act as a record of the selection process of all suggested
initiatives, both those trials undertaken and those which were rejected. This will enable the
minimising of future nugatory effort by concisely stating the reasons for selection or deselection,
and where trialled recording evidence of effectiveness.
Photographs of all initiatives of the EPSS trial are included at Annex A.
REFERENCES
[1] CoMET 2010 Case Study – Station & PTI Safety, Final Report, 01/07/2011
[2] STN 271-6 Escalator Passenger Safety Strategy, Final Report, 01/04/2016
Phil Harley IEng MSOE
Phil Harley is an Incorporated Engineer of the Society of Operations Engineers. He holds the post
of Senior Engineer of the Ops LU Engineering Stations Equipment Team. Following a 25 year
career in the British Army, Phil joined London Underground in 2010. He has been involved with a
wide array of innovative projects since, and is the Engineering Lead and Chair of the Escalator
Passenger Safety Strategy committee.
Kevin Seaborne CEng MIMechE

Kevin Seaborne is a Chartered Engineer of the Institution of Mechanical Engineers. He is the Head
of Technical Discipline for Lifts and Escalators for London Underground, accountable for London
Underground Lift and Escalator standards and responsible for managing an engineering team which
supports all parts of the business. He joined London Underground in 1981 on the Graduate training
scheme; having graduated from Sheffield University with a B.Eng (Hons) in Mechanical
Engineering. In 1983 he moved into a substantive post in Lifts and Escalators and has undertaken
various posts in the intervening period. Notably he was part of the team which delivered over 100
escalators as part of the Jubilee Line Extension.
Kevin is a member of BSI MHE 4 committee and has been a BSI delegate member of CEN
TC10/WG2 for escalators works for over 20 years.
Pilot for Standing on Both Sides of Escalators
Celia Harrison1 Neera Kukadia2, Paul Stoneman3 and Grant Dyer4
1
55 Broadway, London, SW1H 0BD, [email protected]
2
3
4
Keywords: escalators, safety, congestion, flow, customer behaviour, standing on both sides,
London Underground (LU)
Abstract. A pilot was carried out at Holborn Station on London Underground (LU) for standing on
both sides of escalators exiting the station. The aims were: to improve safety by reducing slips,
trips and falls; to reduce congestion by using the escalators more effectively; and to change
customer behaviour. The data collected was both qualitative and quantitative. The data collected on
safety was statistically insignificant. It was shown that using both sides of the escalators to stand on
did reduce congestion and increased escalator capacity by approximately 30%. There was a change
in customer behaviour for the duration of the pilot.
1 INTRODUCTION
In seeking to improve safety on LU’s escalators, a series of initiatives were put into place. One of
those initiatives was a three week series of tests at Holborn Station to encourage customers to stand
on both sides of certain escalators. The aims of the tests at Holborn focussed on Congestion and
Flow, and Customer Behaviour, in addition to Safety. Previous research affecting these areas
underpins the methodology described. Calculations were made to predict how many extra
customers might be carried by escalators with a vertical rise of 24 metres (as at Holborn): an
increase in the region of 25-30%. This would be sufficient to increase flow in the station and reduce
congestion and the associated station control measures normally in place. Many methods of data
collection were available and it was decided to collect as much data as possible and analyse it both
qualitatively and quantitatively. The outcomes are discussed and followed by a summary of the
outcomes and conclusions.
2 AIMS
The aims of the tests were to see if changing the way that escalators were used could:
Improve safety by reducing slips, trips and falls
• Accidents happen every day on LU escalators. LU aims to improve safety by reducing
accidents on escalators.
• Most accidents occur when customers have heavy luggage, or are mobility impaired.
• Walking on escalators exacerbates the risk of accidents.
Improve the flow of customers through the station in order to reduce congestion.
• LU places emphasis on safe evacuation from stations, focussing on customers exiting
stations and managing numbers of customers entering.
• With an increasing frequency of trains passing through stations as the service improves,
congestion is an issue at older stations with limited space and new, cost effective solutions
are needed to ease the congestion.
Achieve customer behaviour change
• For more than 100 years LU customers have been requested to stand on the right and walk
on the left of escalators. A significant change in behaviour would be necessary for the
proposed tests to be carried out.
3 PREVIOUS RESEARCH
3.1 Improve Safety by Reducing Slips, Trips and Falls
The Safety Assessment Federation’s 2011 paper providing guidelines on BS EN 115, which defines
escalator safety requirements in the UK, stated that
“Slips trips and falls are the most common incidents on escalators … There are a number of
reasons why they occur, which include: poor lighting, location of the installation, crowding,
distraction, inappropriate footwear, poor judgment by users, horseplay, use of alcohol and drugs,
loss of balance, spillages, debris, environmental conditions, use as a static staircase, or by
unsupervised minors.” [1]
In an article [2] it was identified that the highest risk group of slips, trips and falls on escalators
were those aged 65 and over, and those aged 5 and under.
According to the South China Morning Post in August 2015 [3] it is now mandatory to stand only
on both sides of the escalators on metros in Hong Kong and Japan. The practice was brought in to
improve safety. “According to the MTR, in the first seven months of 2015, 382 escalator accidents
were recorded – about 12 per cent fewer than in the same period last year. Some 51 per cent of the
accidents involved seniors and children due to loss of balance, standing too close to the step edge,
or carrying heavy luggage.”
3.2 Improve the flow of customers in order to reduce congestion.
People need more space than the size of their physical bodies and how much space is needed varies
from country to country. [4]
LU escalators have width of 1.01m and depth of 0.41m and height of 0.4m. These dimensions mean
that it is uncomfortable for people to stand side by side. Two people, side by side, will require
1.22m width, where LU escalators have 1.01m available. One person on a step requires 0.457m,
where LU escalators have step depth of 0.41m available. Again, this will make a person in this
position very uncomfortable.
“…escalator utilisation and capacities are closely related to human factors such as shoulder width,
personal space preferences, and ability to adjust to system speed. Even under heavy queuing,
vacant steps can be observed on most escalators…” [5]
This is described as “the empty step phenomenon” and Fruin [4] explains this as why capacity is
never as high as two people on every step would be. The two reasons he gives for this is the slight
hesitation that people have when getting onto an escalator, and the innate desire for personal space.
Fruin [4] also studied movement on stairs and observed that, in general, people keep two vacant
steps in front of them when walking on stairs.
Davis and Dutta [6] carried out a study of escalator capacity on LU which observed that escalators
with a greater vertical height have fewer people walking up them. Other factors apart from vertical
height affect how escalators are used: where there is more than one escalator, and where escalators
are next to a corner which reduces the approach space to the escalator. Non-commuters also have an
effect, as they tend to stand rather than walk up escalators.
3.3 Achieve customer behaviour change
Larcom et al [7] looked at the effects of forcing behaviour change on commuters by LU workers
strike action, where commuters under-experiment with routes in normal times. The implication is
that people do not naturally seek change for improvements in their journeys i.e. do not want to
Pilot for Standing on Both Sides of Escalators 11-3
change their behaviour. However, if forced to change their behaviour, people can recognise benefits
and make changes.
In work carried out by Dolan et al [8], which drew on academic evidence of what influences
behaviour, suggestions for innovative interventions were made:
“…much of behaviour change is about battling habits…Habits …usually develop when actions are
repeatedly paired with an event of context (e.g. drinking coffee after waking up)… …the most
effective way of changing…habits is by going with the grain of behaviour: harnessing the same
automatic effects to nudge people onto a different, self-sustaining, track, without always explicitly
stating the need to pursue a particular goal.” [8]
LU customers’ habits of walking, or standing, are very entrenched. A gradual progression on tests
with one escalator only, followed by two, then three over the three weeks was decided on in an
attempt to introduce the standing on both sides slowly, leaving the option to walk open until the
third week of tests. It was decided to use staff to “encourage” customers to stand on the left of the
escalators instead of walk.
4 METHODOLOGY
A start date for three weeks of testing was agreed for the 23rd November as this would permit two
weeks of tests before Tottenham Court Road re-opened to Central Line trains, which was expected
to result in a reduction of customer numbers at Holborn.
4.1 Calculations for a Theoretical Increase in Capacity of Escalators at Holborn

Simple calculations were made to show escalator capacity. LU escalators have a speed of 0.75m/s
and a step height of 0.4 m which gives the number of steps/minute as 112.5. With customers
standing on both sides of the escalator and occupying every step this gives a theoretical maximum
of 225 customers/minute. However, looking at the right hand, stand only, side; and taking into
account the empty step previously discussed, this gives a capacity of 56.25 customers/minute.
On the left hand, walking side, with an assumed walking speed of 0.5m/s, an escalator speed of
0.75m/s gives a walking speed of 75m/minute. Given a step rise height of 0.4m this gives a walking
speed of 187.5 steps/minute. Taking into account the two step vacancy described by Fruin for stair
walking, the speed for walkers on the left hand side of the escalator is calculated to be 62.5
customers/minute.
These calculations do not take into account the vertical height of escalators. It is assumed for the
purpose of this calculation that there is a decreasing percentage of passengers willing to walk up a
high machine.
The percentages given below are partly based on observations at Canary Wharf with a 10m vertical
rise, together with observations of customer walking behaviour on escalators with a greater vertical
rise than 10 metres. The graph below shows vertical height vs % of passengers willing to walk. At
Holborn the escalator rise is 24 m which gives an estimated amount of 40% of customers willing to
walk.
% of Vertical
120 Passengers height (m)
% of passengers willing
100 100 5
80 100 10
to walk up
90 12.5
60
80 15
40 70 18
20 60 20
0 50 22
40 24
0 10 20 30 40
30 26
Vertical height of escalator (m) 20 27
10 28
0 30
40% of customers walking on the left is 40% of 62.5 customers/minute, which totals 25
customers/minute. If customers stand on both sides of the escalator this gives a rate of 112.5
customers/minute. The difference between rates for standing on both sides of the escalator, or
leaving one side for walking is 31.25 customers/minute. In theory, passengers standing on left and
right of escalator at Holborn should increase number of passengers per minute by 27.8%
4.2 Data collection

It was decided to collect data from as many sources as possible:
• Numbers of customers counted off escalators
• Observers to note crowd behaviour and use of escalators
• Staff de-briefs after each test
• Dwell times and headways of all services
• Gate line exits
• Timed walks from platforms
• Incident comparison
• Customer feedback
The data was to be analysed both qualitatively and quantitatively.
4.3 Mechanics of tests and staffing
Two or three members of staff were placed at the bottom of escalator 5, 6 and 7 to encourage
customers to stand on both sides of the escalator/s being used to stand on both sides. An observer
stood at the back wall of the mid-circulating area to monitor crowd behaviour and assist as
necessary. People were located at the top of the escalators to count customers leaving the escalator
using a “clicker” counter. A second observer was located where they could observe crowd
behaviour and assist as necessary.
Staff: The tests were carried out by a combination of the Special Requirements Team (SRT) and
“volunteers” from Lifts & Escalators (L&E) and Strategy & Service Development – Customer
Strategy (S&SD). Non-operational staff were identified by pink hi-vi tabards. Station staff were not
to be taken from their normal duties.
4.4 Variations to the Planned Tests

Variations to the tests were made over the three weeks. After the first days of tests loud hailers were
used for three days, followed by use of the local PA system in order to be heard. After suggestions
from various sources, including customers, some staff in plain clothes volunteered to stand on the
left of the test escalators to stop people walking up. This had the added benefit of the plain clothes
staff hearing comments from customers on the escalators.
4.5 Unplanned Incidents
Day one: escalator 7 had been chosen as “stand only”, but was out of service. The test was not
carried out on that day.
Day four: 58 minutes suspension on the Piccadilly Line (smoke from a train at Kings Cross).
Day six: escalator 6 taken out of service because a fault at 08:32. Escalator 6 was used as a walk
down staircase, escalator 4 reversed to “up”, with standing on both sides “encouraged” on escalators
5 and 7.
4.6 Service Provision
Leading into the tests, both Central Line and Piccadilly Line Fleets had technical problems
requiring a large number of cancellations. The Piccadilly Line had up to 13% cancellations the first
week, 8% the second week, falling to a maximum of 4% on the third week. The Central Line had a
steady maximum of 4% cancellations on all three weeks. Both lines have 78 trains per hour
scheduled at this time of day.
5 QUALITATIVE OUTCOMES
5.1 Observations on Safety
Observers noted that there were several issues around customer behaviour that posed a potential
safety risk. Many customers began to prepare themselves for exiting the station on the escalator,
but on leaving the escalator, would drop items, such as ticket holders, etc., and would stop to pick
them up without regard for the surge of people behind them. The same effect was caused by
customers with wheeled suitcases, where they would lift the case off the escalator in front of them,
hesitate while they extended the handle and then move forward around their case so as to pull it
behind them. These little interruptions to the flow of customers exiting escalators had the potential
to cause a “pile up”. There were no customer injuries.
Tottenham Court Road, which is the next station from Holborn on the Central Line, had no Central
Line trains stopping while upgrade works were being done. This caused increased numbers of
customers at Holborn which led to congestion: the station response to this was to implement
“station control” by holding customers exiting from the Piccadilly Line in the lower circulating area
at the bottom of escalators 2 and 3 while congestion cleared in the mid-circulating area. During the
escalator tests, “station control” was only implemented once and this was during the first week.
There were few gate line problems over the three weeks and none of them led to over-crowding of
the ticket hall.
5.2 Customer Feedback

5.2.1 Customer Contact Centre and Email
Six customers gave feedback received via the Customer Contact Centre and seven customers from
other sources (e.g. phone or direct email). Recurring themes were that the tests would not work (to
relieve congestion); people feel deprived of the choice to walk and/or exercise; and that it delays
their journey. Three customers understood and supported the tests, but felt that the choice to walk
up at least one escalator should remain.
5.2.2 Twitter
Twitter comments were selected based on certain keywords: Holborn, both side, escalator, pilot,
stand, test and trial. The date range was selected to include a period prior to the first day that
escalator tests were carried out on Tuesday 24th November. There was a large increase in tweets on
the first two days of tests. The number of tweets fell sharply at the weekend to none and then rose
sharply on Monday, but not to as high a level as the previous week. Again, the number of tweets
fell gradually over the week, briefly reaching zero over the weekend. On the final week there was a
peak on the Monday which fell over Tuesday and Wednesday.
Holborn Pilot Tweets per day

(from legend - Majority and Highly Possible only)
60
40
20
0
16/11/2…
17/11/2…
18/11/2…
19/11/2…
20/11/2…
21/11/2…
22/11/2…
23/11/2…
24/11/2…
25/11/2…
26/11/2…
27/11/2…
28/11/2…
29/11/2…
30/11/2…
01/12/2…
02/12/2…
03/12/2…
04/12/2…
05/12/2…
06/12/2…
07/12/2…
08/12/2…
09/12/2…
10/12/2…
11/12/2…
Each week the tweets peaked on the first day of tests. These coincide with the expansion of
numbers of escalators included in the tests. The tweets fall off over the week as customers became
accustomed to the new restrictions placed upon them. Less than half of the tweets looked at were
negative. Others were humorous, neutral or questioning.
5.2.3 Media
On the second day of the test period the media began to take an interest, with journalists going to
Holborn station and taking covert footage and interviews. The intense media interest had an impact
on the tests. The most positive impact was that customers were given an explanation of how the
tests were aiming to improve flow and reduce congestion.
Once media reporting began, customers affected by the tests began to verbally express their
opinions, both positive and negative, in an uninhibited way and to take films of their ride on the
escalators on their mobiles. Customer behaviour changed as they felt observed.
5.2.4 Customers at Holborn
Customer response directly given at Holborn during the tests was wide-ranging. There was frequent
non-verbal communication in the form of head-shaking, particularly if the person concerned met the
eyes of a member of staff. Many people gave short, negative feedback, e.g. “This is a stupid idea”;
“This is not working”; “You are making me late”. Initially there was a high frequency of people
asking for information and saying that it would not work. After the first week, the comments
changed from saying that it would not work, to saying that they did not like it or did not want to do
it, implying a level of acceptance and compliance. Another theme that was mentioned frequently
was that customers felt they were being deprived of exercise and the choice to walk.
There was also a significant amount of positive feedback with customer comments that the flows
from the platforms had improved and suggestions on how the tests could be improved. Some
customers suggested that staff/students be used to “enforce” the standing, by standing on the left in
front of customers. By the third week, SRT staff reported that some regular customers said good
morning and made a point of standing on the left of the escalators.
5.3 Observations on Customer Flow, Congestion and Customer Behaviour

The first day of week one brought the most resistance from customers and it took the longest to gain
compliance. By the third week, most customers were compliant by Tuesday. From the first days of
the tests it was observed that the mid-circulating area cleared much more quickly. Apart from one
day during the first week, no “station control” was required.
Over the three weeks there were various staff, with different styles, assisting with “encouraging”
customers to stand on both sides of the escalators concerned. Most noted that humour worked best
in achieving compliance. One member of staff encouraged couples to stand side by side and hold
hands. It was observed that if customers stood side by side and talked, or held hands, customers
behind them did not attempt to pass them.
It was observed that those customers who really wanted to walk found a way to do so e.g. weaving
between other customers on both sides of escalator. One man pushed a child aside so that he could
walk, demonstrating how strongly ingrained the habit of walking can be that overcomes the social
norm that prohibits the touching of other people’s children. Standing on both sides of the escalators
was most effective when the mid-circulating area was congested.
6 QUANTITATIVE OUTCOMES
6.1 Safety
Incident reports from LU Safety and Environmental Analysis (LUSEA) were run which allowed
comparison between the three weeks of tests and the previous three weeks, and the three
corresponding weeks from the previous year. There were only two customer related escalator
incidents reported: one on the 22/11/15 and one on the 29/11/14, with none reported during the trial
period. With such small numbers this is not considered significant.
6.2 Congestion and Flow
To compare escalator usage of standing and walking, simple calculations were completed to
understand if there was an improvement in customer throughput. In week 2, Escalator 5 gave
customers the option to walk up the escalator; the total amount of people that used this escalator
was approximately 12,745 customers. In week 3 when escalator 5 was standing only, approximately
16,220 customers used it. This is around a 30% increase in the throughput of customers, matching
our predictions.
On Tuesday 8th Dec (Week 3, Day 2) all escalators were standing only, meaning the gate line data
and physical counting was very similar. Human error accounts for an approximate 8% discrepancy
between the two. Graph 5 shows the counter data of the three escalators. Graph 6 shows the gate
line data. The peaks in customers are at 8:45, 9:05 and 9:15 and low flows at 8:35, 8:55, 9:10 and
9:25. The headways show that the Piccadilly line had delays between 8:55- 9:00. Trains from both
directions came in at 9:01 after a five minute gap in the service, which explains the dips and the
peaks at 9:05. Between 9:08 and 9:11, there were delays on both lines in both directions, causing
the exaggerated dip on the graphs.
1500 Counter data - Throughput of customers

Number of customers
1000
500
0
08:30 08:35 08:40 08:45 08:50 08:55 09:00 09:05 09:10 09:15 09:20 09:25
Time
Graph 1 Throughput of customers on all 3 escalators
1500 Gate line data - Throughput of customers

Gateline Oyster counts
1000
500
0
08:30 08:35 08:40 08:45 08:50 08:55 09:00 09:05 09:10 09:15 09:20 09:25
5 minute time slots
Graph 2 Throughput of all customers exiting the station
6.3 Customer Behaviour

It was observed that customers exiting the Central line would normally use escalator 7, and
customers exiting the Piccadilly line would use escalator 5. Escalator 6 is between them and was
used by customers from both lines, but mainly the Central Line. Using headway data for 08:30-
09:30 from the final week of the trial it can be seen that delays in the service of the different lines
demonstrate customer’s preference for particular escalators.
Graph showing escalator counts on 08/12/15

500
400
300 Escalator 5
200 Escalator 6
100
Escalator 7
0
08:30
08:35
08:40
08:45
08:50
08:55
09:00
09:05
09:10
09:15
09:20
09:25
Graph 3 Counts for each escalator on 8/12/15
Graph 3 has a high peak between 9:05 - 9:10 on Escalator 5. A train on the eastbound Piccadilly
line came in after a 4 minute delay. These customers arrived at the escalators at 9:06, showing that
customers from the Piccadilly Line tend to use Escalator 5.
Between 9:10 - 9:15, there were very few customers on escalators 6 and 7. Only one Central Line
train arrived at this time; there were no Central line customers exiting the station for 4 of the 5
minutes, demonstrating that Central line customers tend to use Escalators 7 and 6.
7 SUMMARY OF OUTCOMES
7.1 Safety
During the tests over the three weeks at Holborn, no injuries relating to escalators were reported.
When compared to the period prior to the tests and against the same period last year, the data was
not statistically significant.
Some customer behaviour was observed which posed some risk to themselves and others. When
exiting escalators, customers tended to show a lack of awareness regarding the flow of customers
behind them e.g. stopping to pull a case in a different direction, etc. With increased flows the need
to keep customers moving becomes more of a priority.
7.2 Congestion and Flow

Observations by station staff and those implementing the tests confirmed that encouraging
customers to stand on both sides of escalators does improve the flow of customers and relieves
congestion. These observations were confirmed quantitatively with an approximate increase of
30% matching the prediction of increased capacity for standing on both sides of escalators at
Holborn.
“Station control” was only implemented on one day during the first week of tests. Prior to the tests
implementing “station control” was something which happened on an almost daily basis. This is a
good indicator that flows had improved. Service provision had a significant impact on customer
flows.
7.3 Customer Behaviour
There was a wide variety of customer behaviours during the tests which were exacerbated by the
intense media interest in the tests. Customers expressed concerns about prevention of exercise,
lateness, not believing that improving the flow in this way worked. The media attention appeared to
make customers feel less inhibited in expressing their feelings, but also had a major benefit of
explaining what the tests were trying to achieve. There were a significant number of customers who
were interested and/or positive about the tests. Some observed that they could see that the flows
from the platforms had improved and others suggested ways to improve the tests. Very few
customers submitted feedback to TfL: there were 13 submissions from an approximate 130,000
customers affected by the tests.
It was noted by staff that humour worked best in achieving compliance and when customers stood
side by side and talked, or held hands, customers behind them did not attempt to pass them. Those
customers who really wanted to walk found a way to do so. One man pushed a child aside so that he
could walk, demonstrating how strongly ingrained the habit of walking can be that overcomes the
inhibition of touching of other people’s children. Standing on both sides of the escalators was most
effective when the mid-circulating area was congested and minimal encouragement was used to get
customers to stand on both sides of the escalators.
8 CONCLUSIONS
Regarding safety there were no significant incidents or injuries reported. Customer behaviours at
the exit points of escalators do present some concern where interruption to customer flows are
concerned.
The tests were successful in easing congestion and improving customer flows. However, the tests
required a large number of staff to implement. This is a consideration in how to take this forward.
Customer behaviour was only changed for the duration of the tests, with “normal” escalator usage
resuming when the tests were over. Some strong emotions were displayed by customers who
wished to continue in their habitual routine, although most were compliant. Significant numbers of
comments related to wanting a “walking” escalator for the purpose of speed, exercise and in case of
lateness.
It is clear that implementing “standing only” escalators would not be suitable for all locations given
that shorter escalators achieve greater efficiency when walking is permitted; not all locations have
congestion issues which would benefit from this approach; and, each location varies in physical
characteristics which could affect the efficiency of how the escalator is used.
REFERENCES
[1] “Guidelines for the Safe Operation of Escalators and Moving Walkways” (2011), Safety
Assessment Federation http://www.safetyresearch.net/blog/articles/escalator-injuries-and-deaths-
and-role-design
[2] “Escalator Injuries and Deaths and the Role of Design” (2011), Safety Research and
Strategies, Inc. http://www.safetyresearch.net/blog/articles/escalator-injuries-and-deaths-and-role-
design
[3] “Stop walking on escalators: Hong Kong divided over MTR plea for people to stand still ...
but what do you think?” (2015), South China Morning Post, http://www.scmp.com/news/hong-
kong/education-community/article/1853813/stop-walking-escalators-hong-kong-divided-
over?page=all
[4] J.J. Fruin (1987), “Pedestrian Planning and Design”, Revised Edition, Elevator World, Inc.
Educational Services Division.
[5] J.J.Fruin (1992) “Designing for Pedestrians”, Chapter 8 of “Public Transport United States”,
http://ntl.bts.gov/DOCS/11877/Chapter_8.html
[6] P. Davis, G. Dutta (2002), “Estimation of Capacity of Escalators in LU”, File URL:
http://www.iimahd.ernet.in/publications/data/2002-11-01GoutamDutta.pdf
[7] S. Larcom, F. Raunch, T. Willems (2015), “The Benefits of Forced Experimentation:
Striking Evidence from the LU Network”, University of Oxford, Department of Economics
Discussion Paper Series, Number 755.
[8] P. Dolan, M. Hallsworth, D. Halpern, D. King, I. Vlaer (2010), “MINDSPACE: Influencing
Behaviour Through Public Policy”, Institute for Government, UK.
Celia Harrison is a Duty Reliability Manager, recently seconded to Customer Strategy Analyst in
Strategy and Service Development at London Underground.
Neera Kukadia is a Graduate Engineer with London Underground.
Paul Stoneman is a Design Engineering Manager in Lifts and Escalators at London Underground.
Grant Dyer is a Customer Strategy Manager in Strategy and Service Development at London
Underground.
An Overview of India, Travelling Tall
Anandi Khandekar
TAK Consulting Pvt. Ltd. A303, Galleria, Hiranandani Gardens, Powai, Mumbai 400 076, India
[email protected]
Keywords: tall habitable building, post occupancy, morning peak, building population distribution,
economic impact
Abstract. This paper offers an insight which should go into the design of tall buildings and the
potential factors that may influence it, keeping India and Indians in mind. The paper explains how
Indian tall buildings are unique vis-a-vis the tall buildings constructed or under construction
worldwide. It also details the cultural / lifestyle impact that the buildings have and highlights the
precautions the designers need to take to successfully construct, habitable tall buildings. The author
also reviews “Transportation systems in buildings, CIBSE Guide D: 2015” in context of India, the
vertical transportation practices followed here and how cultural aspects affect the theories.
1 INTRODUCTION
There are many papers being presented across the world, on the latest trends and technologies for
construction of a tall building and that there is not yet a technology to change the core of the
building midway or post construction. Hence, this paper tries to cover the essential aspects of
constructing a technically right tall building with a focus on vertical transportation, which according
to the author is the essence of a tall habitable building.
Every country has its own growth curve, and so does India. India did not see tall residential
buildings (around 40 floors and around 120m) with lift speeds 4.0m/s (meter per second) and above
until 2002. Hence, we are in a very interesting phase where we, as designers, are debating between
what the ideal tall building should be and what the precedents are. The learning curve is a very
interesting phase to be in. Hence it all begins with understanding the major design requirements of
the tall building and then designing it to be habitable post occupancy. Habitable is the word we all
need to concentrate on, because a building’s success depends on the building services being
designed based on the occupant’s needs.
2 KEY DESIGN PARAMETERS

This section identifies the key inputs which need to be looked at while designing a habitable tall
building. These inputs are listed below:
1. Type and class of the building
2. Location of the building and nearby places
3. The building’s population distribution
4. Other service requirements
5. Emergency evacuation
6. Economic impact
7. Project planning
8. Selection of the right type of equipment (lifts/escalators/moving walks etc.)
The following paragraphs try to detail the above inputs with respect to different types of tall
buildings from an Indian context.
2.1 Type and class of the building

As per the author’s experience, we, as designers, need to firmly decide what type and class of
building we want to construct. For example, a pure residential tower, office, commercial tower,
hotel tower or mixed use building, luxury or low-income residential building, class “A” or class “C”
office building. Of course the Indian market is fluctuating as this is a developing country but we as
designers have to take a call on these aspects before the construction starts and achieve the goal of
constructing a habitable building. We need to understand that every type of building has its own VT
(vertical transportation) requirements. For example, residential building requirements are totally
different to mixed-use buildings, and hotels have different requirements to commercial buildings.
Also, we need to keep in mind that a 60 storey building is not equal to three 20 storey buildings.
Hence, designing a VT system for a tall building is totally a different ball game. As we go tall the
VT requirements get complex and all the more difficult to alter. Clarity on what type and class of
building we are designing is therefore required.
Some of the VT requirements followed in India are as follows:
1. For any residential building, a minimum of 7.5% of up-peak handling capacity (as per NBC
(National Building Code of India) 2005) is to be designed for.
2. Office buildings could be of many types: single tenant, multi-tenant, 24/7 buildings, call
centres, etc. For a single-tenant building a minimum of 15% up-peak handling capacity (as
per NBC (National Building Code of India) 2005) is to be designed for, whereas for a multi-
tenant building a minimum of 10% handling capacity is to be designed for.
Note: Further details are mentioned in heading 3, page 4.
2.2 Location of the building and nearby places
The geographical location of the building is crucial in designing the VT system, particularly for tall
buildings as this influences the type of VT system we need to design. This also includes the cultural
impact that the location of the building has on the VT system. Even the population of the building
can be influenced by the location of the building. For example, the service staff figures could vary
depending on the location of the building. Other factors which influence the VT system are: tier of
the city, the target customers, exact location of the building (for example prime locality, near to the
airport, near to the metro station etc.), wind loads, seismic zone, developer of the building, etc.
2.3 The building’s population distribution
The distribution of a building’s population primarily comes into consideration in the case of
residential and mixed-use buildings. The VT system is influenced by the lifestyle choices of the
building’s population. While designing, the building population needs to be systematically
bifurcated so that there is no mixing of different constituents and they travel hassle free once the
building is constructed. VT system compliments the requirements of bifurcation of different
constituents.
1. A residential building’s population could be bifurcated into residents and service staff which
includes the floating and resident service staff. Further details are mentioned below in the
section entitled “design considerations in reference to India”.
2. In a mixed-use building, bifurcating the population of different sections of the building is
crucial. In a hypothetical mixed-use tower where the lower floors are a shopping mall, the
middle section is a hotel and the upper floors are residential floors, we cannot expect the
residential population to travel with the hotel guests or the hotel guests with the shopping
mall visitors. They have to be physically bifurcated with separate VT systems designed for
all these 3 constituents so as to avoid any inconvenience to other passengers. In mixed-use
An Overview of India, Travelling Tall 12-3
buildings proper thought even needs to be applied while designing the multi-level car
parking. Not many Indian buildings have shuttle lifts for parking floors. The debate is
always on whether to provide separate parking shuttle lifts or main lifts serving the parking
floors. But with time this scenario is changing.
2.4 Other service requirements

As the buildings get taller, all the services (mechanical, electrical, plumbing, air-conditioning,
firefighting etc.) get complex in design and need to be coordinated together. For example, a
transformer which needs to be carried to the 60th floor has to have a lift which can accommodate it.
Also, an observatory deck, which needs to be serviced from the basement, requires pre-planning.
Freight lifts are crucial in tall buildings, since these will be used for renovation, material movement
(to transfer upright pianos, modular kitchens, marble/granite pieces, furniture etc.). The author has
experienced that if meaningful co-ordination between different service requirements does not
happen then it ends up with tall structures rather than habitable tall buildings.
2.5 Emergency evacuation
We usually read “do not use lifts in case of fire”, which means that only the fireman’s lift is to be
used by authorised personnel to evacuate the entire building. In many buildings, we only have one
dedicated fire lift for the entire building, which could be inadequate. Hence for a tall building more
thought needs to be given to designing at least one or two lifts per group (depending on the size of
the group) as fire lifts. Also, we need to answer whether the fire lifts are capable of evacuating the
entire population of the building. How will the elderly, children, pregnant women or disabled
people use stairways or fire chutes in the case of fire lifts being unavailable for immediate rescue?
Fire lifts can evacuate people in turns and hence the designers need to address this issue as a priority
when designing tall buildings, especially in India. Thought needs to be given to providing protected
lobbies which in turn will make the lifts available in case of fire on any floor. This thought is being
applied in few buildings in India, after mishaps in fire lifts were noticed. The capacity of these lifts
is also an important factor to be looked at. The evacuation plan will have to be designed from the
beginning, in tandem, with VT design. To achieve this, all the services (mechanical, electrical,
plumbing, air-conditioning, firefighting etc.) need to work together and not in isolation, as is usually
the case in India.
Another important parameter to be designed for is that the fireman’s “lift shall work at or above
1m/s so as to reach the top floor from ground level within one minute” (Indian Standard 14665 (Part
2/Sec 1): 2000 Electric Traction Lifts). The local (particular to a state) fire norms also need to be
accounted for. This aspect of design has more significance in designing tall buildings, which the
designers need to understand. Another important aspect for emergency evacuation is the seismic
zone of the location of the building. A VT system that is designed according to the seismic zone
will not be helpful if the structure does not support the same. The author has experienced cases
where the VT system has been designed according to the seismic zone but the structure is not.
Hence it has to be a combined design effort.
2.6 Economic impact

Being a developing country, economic considerations are key factors since a project has to be
financially viable for it to see the light of day. We Indians are still concerned about the cost of the
equipment vis-a-via the VT technology required for the project. We still do not feel the need to
adopt the best of the technologies available if it is high on the cost scale. We keep wondering
whether high speed lifts, the cutting-edge of lift technology, are worth spending money on. Having
said this, there are projects which demand huge freight lifts (as high as 18tons) and escalators
travelling up to 13.0m, which is still new for India. However, in many cases the financial
implications are still the factor that ultimately decides the VT system in the building.
2.7 Project planning

Project planning is another aspect where designers in India are in the learning process when related
to the construction of tall buildings. The designers may fail to understand the intricacies involved
while the construction is underway. They tend to ignore the critical aspects such as construction
accuracy, vendor’s involvement in the project, installation process etc. for tall buildings. Design
changes happen at times even mid-way through the construction, and designers try to manipulate
the existing systems to work for the changes. In this process most of the times buildings end up
having compromised VT systems. It is beneficial that the designers understand the adverse effects
the tall buildings have due to improper project planning.
2.8 Selection of the right type of equipment (lifts/escalators/moving walks etc.):
The type of equipment and their detailed specifications are crucial, and designers need to
understand this as the environment for maintaining the equipment could be unique. For example:
1. A lift car provided with a top hatch is dangerous in India, since most of the times the rescue
operation is performed by the security personnel who are not completely trained or educated
2. The Indian climatic condition requires blowers/fans inside the car, which is usually located
in the false ceiling.
3. Due to the type of attire majority of Indians wear (sarees, dupattas etc.), a saree guard is a
must in escalators.
3 DESIGN CONSIDERATIONS FOR INDIAN BUILDINGS AND REMARKS ON

CIBSE GUIDE D: 2015
Below is the gist of a study made by the author on a few important categories of buildings.
3.1 Residential Buildings
Most Indians are used to getting their daily milk, newspaper, laundry and other items delivered at
their doorstep. The important aspect designers need to note is that VT design requirements are
crucial for allowing access by service staff such as maids, drivers, deliveries, garbage disposal et al.
when the buildings are tall. Additional service staff movement needs to be accounted for as in a tall
building they will definitely use the lifts. This usually happens in the morning, which sometimes
overlaps with school children going downstairs and office workers leaving home early, which
impacts the overall VT performance if dedicated service lifts are not provided. Hence the most
important design period for residential buildings across India is morning peak as against the practice
followed in the other parts of the world. In continuation to the peak period in residential buildings,
the author remarks as follows:
CIBSE Guide D 2015, clause 3.15.9 on Residential buildings, states that “the commonly
used design period for a residential building is afternoon, 5-minute, two-way traffic
condition, which is considered the most demanding traffic period”. In India the afternoons
mostly witness very light traffic.
The majority of the buildings still have gas cylinders being used for cooking, which are taken up the
building by lifts. The lifts are also used for garbage disposal from every floor, since garbage chutes
are rarely designed. These are the parameters which need planning from the design stage itself to
make a building habitable.
The population figures assumed are in Table 1, below:

Table 1 Indicative population figures for residential buildings
Service
Service Staff:
Apartment Type Residents Staff:
resident
floating
1 BHK 2 to 3 NA 0.5
NA
2 BHK 2 to 4 (0.5 to 1 if servant’s 1
room provided)
3 BHK NA
4 to 5 (0.5 to 1 if servant’s 1.5
room provided)
4 BHK 4 to 6 1 to 2 2
5 BHK 4 to 6 2 to 3 2.5
6 BHK 4 to 6 2 to 3 3
Penthouse (5 to 7
4 to 6 3 to 4 5
BHK)
Notes:
1. The above figures are just indicative (as per the author’s experience) and could vary
depending on the location of the building.
2. The figures in the table usually hold good for tier 1, 2 and few tier 3 cities.
3. Resident service staff includes full time maids/helpers.
4. Floating service staff includes maids, drivers, cleaners, milkmen, paper men etc. Maids,
cleaners, milkmen, paper men could be common for the floor(s) or shared between
apartments, hence the decimal figures.
5. BHK is “Bedroom, Hall and Kitchen”
Another parameter to be considered in residential buildings is the stair factor (some percentage of
population using the stairs, maybe floor 1 and 2 residents). Below is the author’s remark on CIBSE
Guide D: 2015 on stair factor (clause 3.14.3: Stairs, Page 3-17).
What if the floor height is a double height entrance lobby or a floor with height of 6 to 8m?
Most of the tall/premium buildings in India are designed with higher heights which unable
passengers to go up by stairs. Hence, the stair factor has to be a function of the floor height.
In a luxury/high-end/tall building the ratio considered for self-driven to chauffeur driven cars could
sometimes be 50:50. Hence, with multi-level car parks being designed, this factor needs to be given
thought while designing the shuttle lifts dedicated to parking floors. Buildings in India rarely have a
concierge desk, so an additional count for the drivers, cleaners etc who go up the building to collect
keys and then go back to the parking floors for cleaning the car or getting the car at the main lobby,
needs to be accounted for. The drivers’ and cleaners’ movements happen in the morning peak itself,
hence morning peak is crucial in a residential building as explained above.
3.2 Office Buildings

The population is usually assumed based on the area per person. The area can vary from 4m2 per
person (for small offices) to 25m2 per person (for single tenant office space) on carpet area. If an
office building is near to a metro station, then it goes without saying that this could have a great
impact on the arrival rate. This is all the more crucial an aspect in a tall building in order to meet the
target VT requirements. Regular working hours are from 10:00hrs to 18:00hrs (excluding multi-
National Companies, call centres, airports etc.).
Below mentioned are few target VT requirements:
Table 2 Recommended handling capacity as per NBC 2005
Type of Building Handling

Capacity
Office- Diversified 10 to 15 percent
Tenants
Office- Single 15 to 25 percent
Tenant
Table 3 Recommended Quality of Service in office buildings as per NBC 2005
20 to 25 seconds Excellent
30 to 35 seconds Good
34 to 40 seconds Fair
45 seconds Poor
Over 45 seconds Unsatisfactory
Note: NBC 2005 is silent on the requirements of Average Waiting Times for both residential
and office buildings.
The author’s another remark on “Clause 4.4.2 Mixed traffic, Page 4-2” of CIBSE Guide D: 2015 is
as follows:
1. From an Indian context, there are different patterns observed in office buildings during
lunchtime. In Multi-National Companies (MNC), occupants usually travel to the common
food court or main lobby (to go out of the building). Few buildings also have common areas
on every floor for employees who carry their own lunch. Another pattern observed is that
meals are served on every desk, and service staff movement is quite high.
a. The importance of morning peak and lunchtime peak is still what we are trying to
understand, hence even today many of the high end buildings have longer waiting
periods at lunchtime – as high as 15-25 minutes.
A common solution to avoid waiting in the queues is that people pack their lunch and
go to their respective floors and have lunch on the desk itself. Alternatively, a few
people have longer lunch breaks or a multi-tenant building has staggered lunch
breaks. Unfortunately, NBC 2005 does not emphasise on the waiting periods at lunch
time.
b. Another source of traffic includes smokers exiting the building during morning and
afternoon peak periods, making 2 trips per person. They usually go up the building to
keep their belongings, then travel back to have a smoke and then travel up to their
respective work stations. As such, 3 trips per person happen during peak periods.
3.3 Hotels
While the international norm of 1 lift per 100 keys does work in India, additional service lifts need
to be provided since a few of the items such as masala chai, fresh ginger tea etc. cannot be prepared
at the room and so require room service. Also, the usual check in and check out times vary
depending on the hotel. If the hotel operator is known at the design stage itself (which rarely
happens in India), the VT design could incorporate this. Another aspect includes the fact that
Indians do tend to celebrate their weddings / engagement-ceremonies / birthdays in a grand way and
the needs of a huge crowd (which prefer to travel together) need to be considered when designing
the VT system for banquets, wedding halls etc.
Note: Generally, escalators are provided for banquet / wedding halls if the hall is on the 1st or 2nd
floor (with nominal floor heights). Lifts are preferred for travel above 3 floors. However, it is
advisable to service these banquet / wedding halls by lifts since the attires could be flowy (sarees,
dupattas etc.) which could get stuck in escalators or moving walks and are difficult to manage on
escalators in a hurry and huge crowd.
3.4 Mixed-use Buildings

With the rapid urbanization and scarcity of land, mixed-use buildings are the way forward for
primarily tier 1 and 2 cities. India still has very few tall buildings in this category as precedents and
hence Indian designers are in the learning phase of understanding the art of mixed-use
developments. An important aspect in designing VT systems for mixed-use buildings is the
bifurcation of all the constituents of the building and utilizing the space available in the best
possible way. It is known that the ideal VT system changes as per the type of building, hence the
same has to be applied to mixed-use buildings. Physical bifurcation of all the constituents is a must
for the success of a mixed-use building and a proper VT system facilitates this. Most of the
buildings have poor VT systems catering to the service movement (maids, helpers, drivers etc.).
Servicing a small building could be managed without specific VT arrangements, but how to service
tall buildings without a proper VT design is the question which needs to be answered at the design
stage itself.
3.5 Hospitals
Due to the space constraint and the healthcare sector getting better, India has hospitals with more
than 5 to 8 floors in tier 1 and 2 cities. These definitely require proper VT systems to cater to the
various activities in a hospital, such as emergencies, health camps, visitor movements, etc. The type
and size of the lifts are crucial when designing VT systems for hospitals. The author’s remark on
CIBSE Guide D: 2015 is as follows:
Clause 3.15.5: “Hospitals”, provides that “In Britain, most hospitals are designed on a 2-3
storey low-rise principle, although many city hospitals have high rise elements. Lifts are
provided in UK low rise hospitals mainly as a means of moving bed bound patients and for
service activities moving floor to floor as staff and visitors use the stairs.”
In hospitals, in the west or far-east very few visitors visit or are allowed; whereas in India we
tend to visit the patient or new-born with the entire family. Hence separate visitor lifts
(sometimes with specific visiting hours) need to be provided. With the healthcare sector
blooming in India, tier 1 and 2 cities have many hospital buildings which are 6 to 7 storeys
high as against the info quoted above. In tall buildings visitors cannot be expected to use the
stairs. These are crucial design considerations which can have a major impact on the VT
system.
4 SUMMARY
With the world getting smaller day by day, all thanks to the internet and social media, there are still
very unique characteristics of the way people live and it is called the culture or lifestyle of the city /
country. While one tends to follow the precedents for the latest VT design or technology, designers
also need to appreciate the fact that cultural differences cannot be changed / ignored when
designing the VT system for any building. Of course, tall buildings have some fixed design criteria
which need to be adhered to, but this needs to be addressed keeping in mind the cultural impact the
VT system has. Though the aim is to construct world class buildings with respect to the
architecture, design, technology etc., the additional services factor (maids, helpers, drivers, etc.)
cannot be ignored, which primarily differentiates Indian buildings from buildings across the world.
The goal needs to be set at constructing habitable tall building and not a tall structure. In this paper
the author has tried to point out a few crucial aspects in designing tall buildings in India.
[1] Cultural Challenges to Elevatoring – Case Study India by TAK Mathews and Nalini
Raghavan
REFERENCES
[1] Transportation systems in buildings CIBSE Guide D: 2015
[2] National building code of India (NBC), 2005
[3] IS 14665 (Part 1 to 5): 2000 and the subsequent amendments
[4] Gina Barney, Elevator Traffic Handbook, Spon Press, 2003
Anandi Khandekar is the Regional Director-South at TAK Consulting Pvt. Ltd., currently located in
Bengaluru, India. She has been involved in E&E design for various prestigious projects including
super high rises, including IndiaBulls Sky at Mumbai, Supernova at Noida, Krrish Square at
Colombo and Nathani Heights at Mumbai, to name a few. She has published articles in Elevator
World and Elevator World India. She is also the committee member for drafting the National
Building Code of India 2015 with specific responsibility for drafting the chapter on building design
guidelines for elevators and escalators. She holds a bachelor of engineering degree in Electrical and
Electronics discipline.
How Current Technology Trends are Empowering Us All to Drive
Innovation
Ben Langham
London Underground AP JNP Escalator Services,
Frank Pick House, Bollo Lane, Acton, London W3 8RP
[email protected]
Keywords: Innovation, London Underground, escalators, lifts, maintenance, technology, web
Abstract. Recent technological trends have given those outside the Information Technology
industry access to increasingly sophisticated products and the ability to contribute to their
development. These trends, such as cloud computing, democratisation of the web and ubiquitous
embedded devices are breaking down the separation between creator and end user. One effect of
this is the ability for users to drive the development of their own innovative solutions, informed by
essential domain knowledge. This paper explores the implications of this on the maintenance of
lifts and escalators and the associated challenges and risks. Some of the ways in which these
opportunities are improving the maintenance of London Underground’s assets shall also be
presented. These include a web-based solution which integrates multiple data sources to facilitate
effective maintenance and asset management decisions.
1 INTRODUCTION
The last decade has seen technological changes that have significantly changed the way we live our
lives. Forty years on from co-founder of Intel, Gordon Moore’s observation that computer processor
speed was doubling every two years [1], this exponential trend has continued. The reduction in size
and cost of processing power, coupled with improvements in networking technology and
transmission speeds opens up many possibilities. The Internet is evolving and growing, through
global collaboration and the sharing of knowledge and ideas, with many developments expected in
the future.
Today, it is not only those with an in-depth understanding of programming or computer science
theory who can actively contribute to the development of advanced technical solutions.
Improvements in hardware and software, plus the availability of cloud computing, have put tools
previously reserved for specialists in the hands of small businesses and individuals. User-friendly
interfaces provide an abstract layer around complex tools, and mobile devices and modern web
browsers enable the deployment of applications to multiple users with minimal investment.
London Underground is increasingly looking for ways to make cost savings and improve efficiency.
Being a large public service organisation responsible for a complex and aging transport network,
the scope for improvement is significant but so too are the challenges in making changes to existing
processes. With growing passenger numbers placing more demand on the service and the
introduction of a 24-hour operation at weekends reducing the number of hours available for
maintenance, the optimisation of resources is vital. One of the areas in which London Underground
is innovating is in the maintenance of its assets. Extracting value from data and making it available
to the right people in a useful format is possible due to the capabilities now afforded by recent
developments in technology, which, if applied appropriately, will enable maintenance to take an
increasingly predictive, condition-based approach and optimise the use of available resources.
This paper outlines the main technological trends influencing our ability to drive innovation as end
users, from prototyping of connected devices to the development and deployment of mobile
software solutions. Two examples of innovations in lift and escalator maintenance at London
Underground, which have been made possible by these trends, shall be presented. In each case a
specific problem was identified which existing systems were not capable of solving, presenting an
opportunity to develop a cost-effective solution with the tools and technologies available.
2 INNOVATION
To innovate is to make changes in something established, especially by introducing new methods,
ideas, or products [2]. As well as the use of new technology, innovation can also involve applying
existing technologies in new ways.
It has been understood for some time that innovation can come from different sources, and that each
can have contrasting functional relationships defining the benefits to be obtained from the
innovation [3]. These sources can include users (both firms and individuals), product
manufacturers and distributers. It is also suggested that the interests of the user and the producer
are misaligned. Where the former aim to meet their exact requirements, the latter tend to generalise
solutions or adapt their own existing products to maximise profits [4].
In many fields, product users are the major sources of innovation, and research shows that there are
advantages to users with unique needs in developing products for themselves [3]. The context in
which the innovation sits, and the associated domain knowledge and expertise, are essential for a
product to meet user requirements. This is a common challenge for manufacturers of technical
solutions, particularly in software development, which adds cost and risk to projects [5]. These
added costs can be minimised by user-centric innovation, which ensures the solution is based on an
in-depth understanding of the domain as well as the problem to be solved.
3 TECHNOLOGY TRENDS
A number of high-level trends are continuously improving the ability of users to develop their own
innovations, and these shall now be discussed.
3.1 Embedded devices
Recent development of small, affordable computers with sufficient processing power and
functionality to make them useful has empowered people to have a go at making rather than
consuming technology. This trend, known as the Maker Movement, has led to the creation of a
myriad of innovative products incorporating embedded computing [6].
Many of the suppliers of embedded devices release the schematics under open source licences.
Some of these require derivatives to be released under a similar license; however, there are options,
such as MangOH [7], which permit derivative products to be released commercially. This opens up
significant possibilities to industry.
2005 saw the release of the Arduino [8]; a small prototyping board consisting of an ATmega128
microcontroller, an integrated development environment based on Processing; a language for non-
programmers, and library functions to easily program the microcontroller. This was an open source
product but required derivative products to be licenced with the same permissions, which has
generated a thriving community and culture of both technical and non-technical innovators.
Since then, devices incorporating microprocessors capable of running a Linux-based operating
system have offered even greater possibilities. The Raspberry Pi is a credit card sized device first
introduced in 2012 and now on its 3rd iteration which boasts a 1.2 GHz 64-bit quad-core processor
How Current Technology Trends are Empowering Us All to Drive Innovation 13-3
and built in Wi-Fi at a price tag of just £25. Since its introduction, over 8 million units of the Pi
have been sold as well as similar offerings from other suppliers [9]. The evolution continues, with
products now in development that are ten times faster than the current Pi, such as the UDOO [10].
3.2 Connectivity
A trend that has very much influenced our way of life is the increase in the number and variety of
devices connected to the Internet. Mobile devices now take many forms, with 4G Long-Term
Evolution now a globally adopted standard, facilitating the use of Internet Protocol (IP) services
completely wirelessly. This makes it easier than ever for us to be connected 24 hours a day.
Machine-to-machine communication, now referred to as the Internet of Things, is becoming
increasingly prevalent, largely due to cost effective embedded systems, improved data transmission
via a variety of protocols and the ability to process and extract value from data sets previously too
large to manage – known as Big Data.
With most mobile devices now comprising fast processors and IP connectivity, deployment of
applications is as straightforward as sending a link to anyone with a compatible device. Hosting
and authentication can be handled via an abstract layer provided by cloud services.
3.3 Cloud computing
The delivery of on-demand computing resources over the Internet on a pay-for-use basis has now
evolved to the stage where it is considered a viable and secure option, with many major
corporations and banks now relying on the cloud for mission-critical applications [11].
A number of models have become popular: Software as a Service (Saas) gives access to innovative
applications and scalable computing power, while Platform as a Service (PaaS) models provide the
environment in which applications can be built and delivered without the need to provision and
maintain hardware or software licences. Infrastructure as a Service (IaaS) makes servers, data
storage and networking possible with no need to provision or maintain physical equipment. With
security features, automatic scaling and flexible pricing structures, cloud computing can offer
significant value with minimal investment.
3.4 Web technologies

The evolution of the Internet has transformed our interactions with each other and our access to
information. The term Web 2.0 was coined in 1999 by Darcy DiNucci [12] and popularised in 2007
by Tim O’Reilly [13] to represent the shift towards a new paradigm of openness and
democratisation for the web. This encompasses social media, applications that run in browsers and
closer integration of data [14].
Originally used solely by scientists and governments in the 1980s, the Internet has come a long way
to the interactive dynamic web we know today. HTML5, the current specification, was finalised
and published on 28 Oct 2014 by The Worldwide Web Consortium (W3C), with close collaboration
between browser vendors ensuring maximum compatibility of features. JavaScript (JS), originally
developed by Netscape in 1995 to make websites more dynamic, has also evolved significantly and
the 7th version of the defining standard, ECMAScript-262, was finalised in June 2016. Many
frameworks and libraries, almost all open source, have been built on top of the language, opening
up vast possibilities for developers. Facebook’s React and Google’s Angular.js are just two of
many open source JS frameworks for building dynamic web applications, and a vibrant community
of developers is continuously creating libraries and tools for accomplishing various tasks from
maths functions to interactive visualisations, providing significant flexibility.
With mobile devices now capable of running advanced web applications, a trend towards
responsive web design enables applications to adjust according to the device on which they are run
to behave like native applications.
3.5 Democratisation
It can be clearly seen from the examples discussed that technology is having a democratising effect.
The fact that the average person carries a device in their pocket that would outperform early
supercomputers supports this statement. This move towards democratisation is having an impact on
business. Wolf [14] defines four pillars of democratised business, all of which are enhanced by
technology:
Democratised knowledge
Access to information on a wide range of subjects is available via the Internet and is ever growing.
Massive Open Online Courses (MOOCs), many of which are free and provided by respected
institutions are increasing in number, providing a wealth of knowledge. Specific questions can
often be answered by peers through online forums, if they have not already been answered and
made available online. Individuals and organisations regularly upload tutorials and how-to videos.
Creative Commons licencing makes it possible to release open source material with clearly defined
permissions from the creator, facilitating the sharing of information.
Democratised creation
An open source approach to software enables large scale collaboration and feedback as well as
democratised tools for creativity. As well as the examples previously discussed, open source
software packages such as Computer Aided Design (CAD) software make advanced product design
tools accessible to all, and low cost 3D printers have also democratised the production process.
Democratisation of funding
Not only is the initial outlay to launch an advanced technical product vastly reduced thanks to the
democratised creation process, but there is the ability to efficiently scale up funding acquisition
through the process of crowd-funding, where a product has potential value for a large number of
users.
Successful projects, such as the +POOL project to build a filtered floating swimming pool in New
York, and GoldieBlox, a construction toy and book series to promote engineering amongst young
females, each raised over a quarter of a million dollars through crowdfunding and brought their
ideas to fruition. These are just two of many such examples.
Democratisation of distribution and commerce
Distribution of software is now easier than ever via the web or through app stores. E-commerce
enables direct payment from mobile devices, providing instant access to a global market.
The combination of the above factors results in the ability for firms or individuals to create products
that effectively satisfy a requirement with minimal investment. In the commercial world there are
many examples of small business that have developed disruptive technology which has taken over
existing markets, often despite the competitors being large corporations with excellent management
processes [15]. Businesses such as Uber and Air BnB have successfully used technology to open up
new markets and rapidly scaled up and built on their initial success to capture long established ones.
4 APPLICATIONS IN LIFT & ESCALATOR MAINTENANCE

Over the past two years, the author has attempted to take advantage of the technology trends
previously described to improve the capability of London Underground’s Jubilee, Northern &
Piccadilly line (JNP) maintenance organisation to make effective decisions to optimise safety,
reliability and cost. The following examples, although within the domain of lift and escalator
maintenance, demonstrate the value of user-centric innovation based on an understanding of a
specific need and the context in which a solution to this need must fit.
Two solutions shall be presented, both of which involve the management of data; an area which was
identified as having potential for improvement. The first is a mobile solution to provide vital
escalator asset data to staff at remote locations and the second is a web-based application which
aggregates data from multiple sources for desktop, smartphone and tablet.
4.1 Mobile Escalator Asset Register

Asset component data for escalators, such as part numbers, drawing numbers and configurations,
are held in a database which is managed by an in-house design team. The database is used to
inform maintenance activities and asset management, and has the ability to generate updates to a
Computerised Maintenance Management System (CMMS) which holds a subset of the data. This
presented two areas for improvement. Firstly, the full database was only accessible via a PC on the
company network, limiting access at remote sites or in meetings. In some cases this caused delays
in carrying out site work. The other area for improvement was in the process by which the data was
kept up to date. The existing process to add corrections to the database involved ad-hoc chains of
communication which were unreliable.
All maintenance team leaders are issued with iPhones and iPads so it was decided to utilise these to
make the full database available remotely. To enable a solution to be developed, a software
package called Filemaker was used. Filemaker, owned by Apple, is marketed as “Powerful, easy-
to-use software used to create custom solutions that run on iPad, iPhone, Windows, Mac, and the
web”. This software was chosen as the use case was achievable with the built in features, including
mobile support, security and authentication. The solution retained the original database as the
master but enabled it to be uploaded periodically to a cloud server which could then be synced to
mobile devices. The solution was designed such that data could be located via an easy to use
interface, and developed through close collaboration and regular feedback from other end users of
the application. The data is stored locally on each device, with no signal is required to view it, and
it can be synced with the master database when signal is available. Staff can submit data
corrections via the app itself, which go directly to the design team who can update the master
database. The app has now been deployed within the JNP maintenance organisation and is
regularly used for emergency and planned maintenance and asset management activities.
The initial investment for software licences was approximately £2,400, which included the ability to
host the master database on a server with up to 10 concurrent user connections (only required for
the syncing process) and the option of increasing this at a later date. The mobile version of the
software was free to download from the App Store by individual users, who could access the app
via an authentication process. A cloud server was set up with Amazon Web Services at a cost of
approximately £20 per month, although for the first 12 months there was no cost due to the use of
the introductory free tier offered by Amazon to new customers. The design of the system was
intuitive and self-explanatory, and no training was required for the users.
The author’s experience in setting up and learning the software was positive, with a variety of
learning materials provided by Filemaker. The debugging functionality provided by Filemaker was
found to be invaluable during the development process.
Although the gentle learning curve make this an effective option for managing data and presenting
it via custom interfaces, the limited options for data visualisation and difficulty in connecting to
non-standard data sources mean it is less effective for complex applications. Scalability is also
limited, so the platform would not be suitable for an enterprise-wide solution to be rolled out across
London Underground. It has, however, enabled a low cost solution to be designed and implemented
which solves a real problem, adding value to maintenance. Should the business decide to
implement an enterprise-scale version, this provides a successfully tested prototype, reducing the
amount of development work that would be required.
Figure 1: Mobile Escalator Asset Register
4.2 Managing multiple data sources

As part of London Underground’s condition-based approach to maintenance of lifts and escalators,
data from various sources are captured and analysed to gain a better understanding of the condition
of the assets and identify potential failure modes. The following data sources are used:
• Remote temperature and vibration monitoring systems

• Offline vibration monitoring with a handheld device
• Thermography of electrical and mechanical components
• Gearbox oil analysis
• Ultrasonic monitoring
• Visual inspections and site reports
• Asset component data
• Work order and failure history
• Photos, videos, audio, drawings and station layout diagrams
Some of the above require specialist proprietary software for importing and processing data, and
other data comes from databases hosted by third parties and emailed files. It was found from
experience that accessing separate systems and databases to obtain the full picture of the assets was
time consuming and detracted attention away from the analysis process. The aim of the solution
was to provide a high level overview of all relevant data via a single user-friendly interface,
enabling the users of the system to focus their attention on value-adding tasks.
Due to the complexity required it was decided to take the approach of developing a custom web
application. This would make it possible to build a scalable and modular solution that would work
on any device with a modern web browser, with scope to incorporate more assets and data sources
as well as analytics and visualisations at a later date. The JS framework Angular.js was selected
due to its flexibility and open source licencing. An active community of users and the fact that it
was backed by Google showed that it was well supported. Using open web technologies meant that
there was zero investment required in new software or tools, the only cost being development time
and the cost of cloud hosting. However, knowledge of web development as well as the frameworks
used required more in-depth learning. The vast amount of information available online facilitated
the learning process, but to keep up to date with web technologies is an on-going task. An
alternative approach would be to invest in the support of one or more web developers.
After identifying the required data sources, it was necessary to make the data available in a
consistent and web-friendly format. The format that was chosen where possible was Javascript
Object Notation (JSON) which is both computer-readable and human-readable and popular with
web applications. Initially, each data source was converted manually to enable the interface to be
designed and demonstrated. The next stage, which is now in progress, is automating the extraction
of the data from each system, format conversion and storage, ensuring up to date information is
presented to the user.
Figure 2 shows a screenshot of the interface on a desktop web browser.
Figure 2: Web front end for access to multiple data sources
Other potential users of this solution include site fitters, maintenance and asset managers as well as
senior management, who would benefit from an overview of the assets. End users were consulted
and observed as part of an iterative design process whereby feedback was incorporated to ensure the
design satisfied the user requirements as closely as possible.
Although the development of the system requires further work, it is already being used successfully
to support group discussions such as failure review meetings, where it is displayed on a large screen
and referred to when needed. It is also useful in the event of a fault to provide instant visibility of
the history of the asset, known issues and monitoring data to aid the fault finding process. The ease
of access to data also facilitates informed decisions relating to the frequency and extent of planned
maintenance activities to be carried out. With maintenance driven by this data, smarter decisions
can be made when planning and scoping work, focusing resources and investment where they are
most needed.
5 DISCUSSION
With customer expectations increasing and pressure on the business to improve efficiencies, there is
certainly a drive towards more innovation in London Underground. The available technology is
developing at a faster rate than that at which it is utilised and embedded in the business so there is
significant potential to take advantage of the trends presented in this paper. London Underground is
currently developing an innovation strategy to co-ordinate Research & Development projects,
which is promising for staff within the business wishing to innovate and develop new technical
solutions. In the wider industry, it is in the interest of businesses to have processes in place to
support small-scale user-driven innovation, with many having successfully embraced this approach
in the form of sponsored incubation programs. Without this, many ideas with potential value will
not have the opportunity to be taken forward. This also provides visibility of technology across the
business to ensure compatibility, avoidance of duplication and shared benefits as well as the ability
to make commercial and strategic decisions regarding the organisation’s technology assets.
Although the increased accessibility of technology can have significant benefits, the adoption of
new technology should be done with consideration of the limitations and risks involved. One major
risk when working with data is security, with accessibility of public cloud services presenting a
possibility for sensitive or confidential data to be made public. Although cloud providers can
encrypt data, it is often up to the user of the service to specify the security settings and configure
authentication as well as scheduling backups to avoid data loss. With the number of devices
connected to the internet growing rapidly, the number of potential vulnerabilities is also increasing,
so cyber-security is a vital consideration, particularly where a security breach could have severe
consequences. Provision of guidance and support for small-scale development projects would
ensure compliance with the necessary standards and protocols.
The introduction of new processes, tools and technologies requires change management, and this is
a common challenge to innovation. In a research report by Reliabilityweb.com, Bentley Systems
identified organisational culture to be the biggest single obstacle to improving asset performance,
but stressed that the proper tools, training and leadership, incorporating data-driven process
management, can make it possible [16].
6 THE FUTURE
Improvements in connectivity are already making it much easier to share data, so it is feasible that
this will develop to the stage where all data that could be of use to businesses will be instantly
accessible, either manually or programmatically by scripts running seamlessly in the background.
A shared data approach requires a paradigm shift in the way in which organisations operate.
Departments previously separate from each other require open data policies and a collaborative
approach in order for the benefits to be realised. Transparency across departments and with
external bodies where appropriate can allow organisations to harness collective knowledge and
maximise the benefits. Deciding on what data to capture and how to process it to extract useful
knowledge and insights is a key challenge that will influence how much value is gained from data
in the future.
Collaboration with the developer community is already happening at Transport For London, who
have organised a number of hackathons, where large previously unused datasets have been provided
to teams who have generated innovative and original solutions in a single day workshop. This
approach is expected to continue with more developers becoming involved and more datasets made
available, gaining useful knowledge and improving the experience for the travelling public in
London.
Software solutions incorporating machine learning and artificial intelligence are becoming more
effective, as are products that enable these services to be incorporated into other applications.
These are made accessible due to the data storage and distributed processing afforded by the cloud,
and it is expected that these solutions will mature to provide maintenance organisations with
intelligent insights to support fault-finding and fault prediction. These tools have already proved
themselves in the aerospace and manufacturing industries and are now being considered on London
Underground lifts and escalators which present a more varied asset base with less standardisation of
design types adding complexity to the modelling.
Improvements in mobile technology are making virtual and augmented reality accessible to all with
minimal investment. This could also have applications in maintenance, for example in visualising
the movement of equipment in confined spaces, overlaying vital information about assets to staff
during maintenance activities and for training. Therefore, this is an area where user-driven
innovation could have significant potential.
An increased expectation for customised solutions to specific problems means that suppliers are
changing their design and development processes. The ability to release software on the cloud is
resulting in much shorter and more agile development cycles, with user feedback informing the
scope of the design. Many software products are supplied as micro-services with open protocols
enabling the customer to incorporate components in a modular fashion to create solutions that meet
their specific needs.
7 CONCLUSIONS
It is now possible not only for ideas to be generated by users of technology, but also for them to be
seen through to completion or to working prototypes of solutions. The cost savings that can be
achieved by this approach are clear, as well as the likelihood of an effective solution being found,
due to the advantage of an in-depth understanding of the problem and the domain or context in
which it should fit. Where the skills or resources are not available to develop a solution,
collaboration across the organisation or with suppliers can be an effective alternative.
This paper has presented two examples demonstrating the value that has been gained from user-led
innovation within lift and escalator maintenance at London Underground. These have required
minimal investment and have each addressed specific problems effectively, adding value to the
business.
As technology continues to evolve, the potential for innovation that is available to all of us will
increase further, and this will present greater opportunities. For businesses to harness the ideas that
may arise within their organisation, a policy which provides the necessary support, tools, and some
freedom to experiment, will enable these ideas to be taken forward, whilst ensuring the necessary
strategic and commercial overview and management of risk.
REFERENCES
[1] Intel, 2015, Moore’s Law, [Online], Available at:
http://www.intel.com/content/www/us/en/silicon-innovations/moores-law-technology.html
[2] Oxford English Dictionary, Innovate Definition, Available at:
http://www.oxforddictionaries.com/definition/english/innovate#innovate__2
[3] von Hippel, E., 1988, The Sources of Innovation, Oxford University Press
[4] von Hippel, E., 2005, Democratizing Innovation, The MIT Press
[5] Box UK [Online], Available at: https://www.boxuk.com/insight/blog-posts/importance-
domain-knowledge
[6] Techopedia, Maker Movement [Online], Available at:
https://www.techopedia.com/definition/28408/maker-movement
[7] MangOH [Online], Available at: http://mangoh.io/
[8] Arduino [Online], Available at: https://www.arduino.cc/
[9] ELinux, History of Raspberry Pi, [Online], Available at:
http://elinux.org/RPi_General_History
[10] Kickstarter, UDOO [Online], Available at: https://www.kickstarter.com/projects/udoo/udoo-
x86-the-most-powerful-maker-board-ever
[11] IBM, Cloud Computing, [Online], Available at: https://www.ibm.com/cloud-
computing/what-is-cloud-computing
[12] DiNucci, D, Fragmented Futures, [Online], Available at: darcyd.com/fragmented_future.pdf
[13] O’Reilly, Web 2.0, [Online], Available at:
http://www.oreilly.com/pub/a/web2/archive/what-is-web-20.html
[14] Wolf [Online], Available at http://www.forbes.com/sites/michaelwolf/2012/11/05/the-new-
era-of-democratized-business/#cb93106190e1
[15] Christensen, C. M., 2006, The Innovator’s Dilemma, Collins
[16] Reliabilityweb.com, 2014, Research Report on Asset Management Practices, Investments
and Challenges 2014-2019. www.reliabilityweb.com
Ben Langham has a BEng in Mechanical Engineering from the University of Reading and an MSc
in Advanced Engineering Design from Brunel University. He has worked in maintenance on the
London Underground since 2006 when he joined the Metronet Rail engineering graduate scheme.
For the past 8 years he has been based in lift and escalator maintenance at London Underground,
first as a Performance Engineer and currently as a Condition-Based Maintenance Engineer.
Changed Requirements In The International Lift Market Ask For
New Pulley Types With Better Tension Equalization Features
Jawk Meijer1, Lothar Sieber²
1
Jawk Meijer, Global Sales Director, Schwartz GmbH Technical Plastics, Hagdornstr. 3,
Xanten/Germany, e-mail: [email protected]
2
Lothar Sieber, Key Account Manager, Schwartz GmbH Technical Plastics, Hagdornstr. 3,
Xanten/Germany, e-mail: [email protected]
Keywords: modernisation, plastic coated ropes, rope compensation, easy-flex, light weight pulley,
cost down, suspension, tension compensation, increased life time, reduce installation time
Abstract: There is a clear trend in the lift construction: Due to the changes in the world
population’s composition as well as the trend toward larger cities the need for new installations as
well as modernization is rising.
In these volume driven “standard” lifts total cost is the major driving force. In these lifts, small
drives and plastic coated ropes or other traction media and pulleys of innovative character and
flexible design are requested. This dimension reduction leads to problems in the suspension systems
such as prolonged installation times and reduced traction media life time. In this presentation the
newest product, which is developed to comply with these new circumstances is introduced. In
practice a new type of deflection pulley is presented which improves installation and life time of
plastic coated ropes.
1 INTRODUCTION
Today’s situation: World population grows, average life expectancy raises (see figure 1) and people
continue to move into cities. This leads to increased and changing requirements for vertical
transportation. In most countries there is, driven by legislation, an absolute need to support
handicapped persons and to make available the right transportation means for the elderly citizens.
The installation of new lifts and the modernization of existing ones is a must.
The task: Public authorities, architects, consultants, lift planners, manufacturers and service
companies have to design, produce, install and service lifts which offer more space, a higher level
of ride comfort and cost benefits. Driven by the fact that elderly people stay at higher age in their
own house, modernized lifts with a larger car space to accommodate wheelchairs and other mobility
aids are required.
Figure 1: Today’s demographic change.

That means in terms of lift engineering that cars need to have a larger floor space to be accessible
for people with restricted mobility. As a result, counterweight and further parts traveling in the shaft
need to be especially space saving. ‘MachineRoomLess’ does no longer mean "No machine room"
but best shaft space utilization possible. In combination with smaller drive units (Figure 2) to reduce
costs, these facts lead to traction media like belts and plastic coated ropes with smaller steel rope
diameters, smaller gearless drive units and finally to smaller pulleys of 240, 160 or even as small as
120 mm diameter.
Figure 2: Reduction of sheave diameter
This dimension reduction in the traction system leads to unwanted effects in the suspension of the
lift. Due to the growing amount of deflection / support rollers in a system as well as the growing
amount of ropes the effects of small dimensional changes can lead to reduced life time of the
traction media.
2 THE PROBLEM
2.1 Installation and Rope tension
In smaller traction systems typically a higher number of ropes are used in order to retain the same
payload. This is often done in combination with a higher number of deflection, car and
counterweight pulleys in order to distribute the loads over more shafts to remain the required life
time for the smaller diameter bearings.
This creates several problems in the installation, as well in the operational phase of the lift. The
modern traction media such as plastic coated ropes bring higher trip numbers but are more complex
in installation and servicing. Due to their design, which is focused on creating high friction between
the surface of the traction sheave and the rope outer surface, the complexity is mainly to achieve
Changed Requirements In The International Lift Market Ask For New Pulley Types With Better Tension 14-3
Equalization Features
and maintain a good equal rope tension over the life time of the ropes. As a general rule all traction
media in a set should be tensioned equally (+/- 5%). This tension needs to be checked right after
installation of the ropes and it needs to be rechecked after some weeks or a maximum of 3 months
after bringing the lift into service. This can be done with the help of electronical tensioning devices
or continuously balancing end termination (hydraulic) devices.
Figure 3: Typical plastic covered rope tension deviation after 100-500 hours of operations
for a 3 rope suspension system in classic rope suspension system.
However due to the high friction surface of the rope (or belts) it is already difficult to install all
ropes equally in a short time since all pulleys will be blocked by the friction between the pulley and
the first rope after this has been pulled through the system. This leads to extended installation times
(and cost) and to problems in equalizing the rope tension - even between section of the same rope.
Experience in the field shows that not only the rope tension of all ropes in one set needs to be
checked and equalized, but also that the tension of one and the same rope in this set can vary
significantly in 2:1 or higher suspensions between pulley and traction sheave, pulley and
neighboring pulley or pulley and end termination. In daily practice it is seen that the equalization is
generally not done properly due to the complexity of these systems.
Between the ropes in a section tension differences can exist (figure 3) as a result of these friction
forces which are directly related to the tension in the two sections on either side of the pulley for a
specific rope and its friction coefficient between the pulley and the rope surface. Especially when
operating with a steel rope tension differences can lead to extra wear in the traction sheave grooves
where the highest loaded ropes are running.
2.2 Length differences

With a classic drive of 320 mm and a 8 mm rope, the wear induced groove diameter reduction of 1
mm in combination with the tolerance in rope diameter of -1/+5% can already lead to a run length
difference of 4.65 mm per rotation (Table 1). With a smaller pulley of 160 mm this effect is the
same, but this pulley has 2x the rpm to follow the same car speed. For 10 meter height the 320 mm
pulley can "give" a run length difference of 42.77 mm, for the 160 mm pulley this will be 83.68 mm
max.
rope length difference per 10 meter movement per pulley compared to nominal
320 160 240 120
319 320 320,1 159 160 160,1 239 240 240,1 119 120 120,1
effect of pulley diameter -30,58 0,00 3,05 -59,88 0,00 5,95
rope diameter in mm
nominal tolerance effective
6 -1% 5,94 -2,45 -2,44 -2,44 -4,79 -4,76 -4,76
5% 6,3 12,23 12,20 12,19 23,95 23,81 23,79
max pot. delta 42,77 83,67
effect of pulley diameter -30,58 0,00 3,05 -59,88 0,00 5,95
8 -1% 7,92 -2,45 -2,44 -2,44 -4,79 -4,76 -4,76

5% 8,4 12,23 12,20 12,19 23,95 23,81 23,80
max pot. delta 42,77 83,68
Table 1: Rope length differences /10 m. movement per pulley compared to nominal diameter
of pulley and rope
These effects can be increased due to the fact that there are more pulleys in the system however the
grove diameter differences are typically only to be expected at steel pulleys such as the traction
sheaves.
These potential extreme length differences between the different ropes in a section of the
suspension will lead to tension differences, especially when the traction media have no possibility
to “slip” over the pulleys to compensate.
This effect can be extra damaging at so called double wrap sheaves. The diameter differences can
lead to an extreme loss of traction due to “loose” ropes.
If the tension is out of balance there will be different forces on each rope, on the rope structure and
on the outer coating. Depending on the rope constructions unequal tension may lead to protruding
wires or strands or cracking of the plastic coating before the end of the expected life time is
reached. Additionally extra loads on the traction media are to be considered which may occur from
design and installation quality out of buffer movements, wrong alignments or high deflection
angles.
3 NEW PULLEY SOLUTION
The solution for the abovementioned problems is to create pulleys which neutralize the friction
and/or the length differences. The smart technique combines the traditional pulley with the benefits
of polyamide. Groove segment rings are mounted on a polyamide-basic body (figure 4) with good
sliding characteristics. The separated rings with grooves on the support body allow each and every
rope to move independently from each other.
Figure 4: Cross section new pulley solution
This design brings a number of benefits to elevator producers and service companies.
The installation is far easier, especially in a lift with multiple pulleys. Due to the separate
independent movement it is possible to tension each rope from each end termination to the traction
sheave without the effects of the friction on each pulley. During the operation the independent rings
(figure 5 and 6) allow each rope to move over the pulley with slight speed differences compared to
the average speed of the rope set. This way rope tensions and/or length differences are neutralized.
Due to this compensation in the rope tension differences the life time of the ropes are increased up
to 1,8 times. It also supports the lift ride comfort and reduces the lift Life Cycle Costs.
Figure 5: New pulley solution with plastic coated ropes

Figure 6: Segments moving the basic body
The new pulleys are running in various test installations, in house and in demanding field
applications. Test in lifts which have been equipped before with solid pulleys show that the ropes
run within a closer tension tolerance field and do compensate the length between the end
termination and the traction sheave.
At present a standard program with grooves for 6.5 and 8.1 mm are used.
In Figures 7 and 8 the marks on the pulley grooves indicate the movement of the different grooves
when in operation. In figure 9 it is shown that the tension deviation between the three ropes is
clearly better than in the classic system as is shown in figure 3.
Figure 7 / 8: New solution before at installation / after 1 week operation

Figure 9: Typical plastic covered rope tension equalizytion after 100-500 hours of operations
for a 3 rope suspension system in a rope suspension system with new pulleys installed.
4 CONCLUSION
Well tensioned ropes are one of the major steps in creating a good drive system. Due to the trend to
use smaller diameter traction and support pulleys and at the same time increase the amount of these
pulleys per system the tension and length equalization problems increase. This leads to a decreased
life time of the traction media, longer installation times and reduced ride comfort. A pulley which
allows speed differences between the different ropes in a set is introduced.
The new solution supports an easy installation, and allows to ropes to compensate for length and
tension differences. This light pulley will not only reduce total life time cost but also improve the
whole elevator behavior and ride comfort.
5 REFERENCES AND FIGURES

[1] Design and drawings by Thomas Schonlau, Jawk Meijer, Karl-Heinz Bosmann, Lothar
Sieber / all Schwartz GmbH Technical Plastics, Xanten/Germany
[2] Figure 1: Demographic change: c.f. Wahlgang (2013):
http://www.wahlgang.de/wahlgang13/?p=1889; 10.02.2016
[3] Figure 2: N.Imbimbo (2004)
[4] Figure 3 and 9: out of Henning Weight Watcher Readings
[5] Table 1 / Figure 4, 5, 6, : Table and drawings by Schwartz GmbH Technical Plastic,
46509 Xanten/Germany
[6] Figure 7,8, : Pictures by Schwartz GmbH Technical Plastic, 46509 Xanten/Germany
6 BIOGRAPHICAL DETAILS
1) Jawk Meijer, Schwartz GmbH Hagdornstr.3, 46509 Xanten/Germany, meijer@schwartz-
plastic.com
2) Lothar Sieber, Schwartz GmbH Hagdornstr.3, 46509 Xanten/Germany, sieber@schwartz-
plastic.com
Jawk Meijer has been with Schwartz GmbH Technical Plastics since 2005 with long experience in
engineered plastics used in cranes, offshore, lifts and many other industries. He is representing the
company Schwartz as Global Sales Director. Mr. Meijer has been in the plastic industry since 1988
and has handled marketing developments and sales in different regions all over the world. He holds
an Msc engineering degree at the Technical University of Delft / Netherlands.
Lothar Sieber for more than two years has been active as a key account manager in the offshore,
crane and lift industry at Schwartz Technical Plastics. Before that he was heading product
management at Pfeifer Drako elevator ropes unit. He has been in the industry for more than 18
years active in international sales and key account management on all continents. His background is
industrial plants turnkey project coordination holding a degree as BSc in Business Administration /
Industrial Systems.
Modelling of a Rope-Free Passenger Transportation System for
Active Cabin Vibration Damping
Jonas Missler1, Thomas Ehrl2, Benedikt Meier2, Stefan Kaczmarczyk3 and
Oliver Sawodny1
Institute for System Dynamics (ISYS), University Stuttgart, Waldburgstr. 19, 70563 Stuttgart,
1
Germany, Email: [email protected], Email: [email protected]

2
thyssenkrupp Elevator Innovation GmbH, Robert-Bosch-Straße 5, 72124 Pliezhausen, Germany,
Email: [email protected], Email:
3
School of Science and Technology, University of Northampton, St. George’s Avenue,
Northampton NN7 3NG, UK, Email: [email protected]
Keywords: Dynamic Modelling, Multi-Body System, Vibration Damping, Control Engineering.
Abstract. Conventional vertical passenger transportation is performed by lifts. Conventional traction-

drive electrical lifts use ropes to transfer the rotational motion of an electrical motor into a vertical
motion of the cabin. The vertical passenger transportation system discussed in this paper does not use
any ropes, the motor directly provides a driving force, which moves the cabin. This new propulsion
is realized through an electrical linear motor. The use of the linear motor requires a new design of the
passenger transportation system (PTS), which includes reducing the weight of the car through
lightweight construction. The reduced stiffness of the lightweight design renders the construction
more vulnerable to vibrations. In order to improve ride quality of the transportation system it is
necessary to develop new concepts to damp the vibrations. One way to increase stiffness
characteristics of the system is to introduce active damping components to be used alongside passive
damping components. It is essential to derive a dynamic model of the system in order to design and
also later control these damping components in the best possible way. This paper describes the
fundamental steps undertaken to derive a dynamic model for designing and controlling active
damping components for the new type of vertical PTS. The model is derived as a Multi-Body System
(MBS), where the connections between the bodies are modelled as spring damper elements. The
derivation of the MBS is demonstrated on a transportation system, consisting of three main
components: a sledge, holding the rotor of the linear motor; a mounting frame, which is used to
provide support for the cabin; and the actual cabin. The modelling of the propulsion system, thus the
electrical part of the PTS, will not be the focus of this work.
1 INTRODUCTION
The novel way of lift designs is PTS without ropes, where a linear motor directly provides the vertical
motion. The rope-free propulsion offers several benefits, like vertical and horizontal travel, the
possibility of multiple lift cars that may ride in a single shaft and the reduction in construction space.
The ability of horizontal movement of the PTS allows the connection of several lift shafts and hence
enables the design of more complex lift shaft networks. A disadvantage of the new propulsion is the
lesser weight that can be carried, therefore a lightweight design of the PTS system is required.
Lightweight construction render systems more susceptible to vibrations, due to the reduction in
stiffness. In general the riding comfort of passengers in lifts decreases in the presence of vibrations,
the vibrations should therefore be kept at an acceptable level. In conventional lifts vibrations have
been reduced by passive damping elements that are placed in the mounting frame around the exterior
of the cabin. This construction decouples the lift cabin from the rest of the car and leads to an
improved riding quality for the passengers. The new structure of the rope-free PTS omits the
mounting frame around the cabin, due to the new propulsion. The high passenger comfort of
conventional lifts must be kept for the new transportation system and therefore it is essential to
develop a new damping concept that fits the new requirements. The new design of the rope-free PTS
is shown in Fig. 1, it consists of three main parts: a sledge, a mounting frame and the cabin. The
sledge holds the passive elements of the linear motor and the mounting frame connects the cabin and
sledge, and provides support for the cabin. One way to reduce the vibration is to accompany the
passive damping elements with active damping components. Active damping components demand a
controller, in order to work and reduce the vibrations in the best possible fashion. The design of such
a controller is a complex engineering task; one commonly used way is to design it using a dynamic
model. The dynamic model is used to simulate the motion of the real system that shall be actively
damped. An important step in developing a controller is the derivation of such a model. This paper
will focus on the derivation of a model for the rope-free PTS system. The derivation will be performed
on a simplified version of the transportation system.
Figure 1 Rope-free passenger transportation system (PTS) in exchanger position.

Source: multi.thyssenkrupp-elevator.com
Active damping components are already implemented in conventional lifts, when passive dampers
are not sufficient to damp vibrations arising in the system. This is especially the case for high-rise
lifts, where the basic structure of cable lifts is kept the same, but the increase in speed also increases
the vibration felt by the passengers inside of the cabin. The primarily used active components in
conventional lifts are active roller guides. An active roller guide has the advantage that vibrations
induced by the guidance can be directly compensated at its source. The vibrations inside the cabin
can be reduced by the factor of five by an active roller guide [1]. Even if not all rollers of the roller
guide are actively actuated the active roller guide shows to be beneficial for the reduction of
vibration [2]. Whilst not directly applicable to the rope-free PTS, because of the different design of
conventional lifts, active roller guides show the potential of active damping components.
The implementation of an active cabin damping concepts will always include the design of a
controller for the active components, thus the active damping is a branch of control engineering. One
field of control engineering is the model-based control design, where a dynamic model of the real-
world system is derived to develop a controller. In the environment of model-based control there is
an important distinction between two types of models. The first model is a very detailed model of the
real-world system, which should include as many properties of the real system as possible. The second
model is the design model, which covers only the most important properties and effects of the real
system. The design model is used to design a controller for the real system. This controller will then
be tested on the detailed model. In many cases, the design model is derived by reducing a detailed
model to the effects that are crucial to achieve a good performance of the controller. Even in case that
the controller is not designed using a model of the real system, it is still valuable to simulate, thus
test, the controller implementing it on a real-world system. For more details on model-based control
and control engineering see [3, 4].
Modelling of a Rope-Free Passenger Transportation System for Active Cabin Vibration Damping 15-3
A commonly used technique to derive a dynamic model of a mechanical system is by using Multi-
Body-System (MBS) techniques. The method of MBS is especially applicable for rigid systems which
experience large rotational and translational displacements. The key point of modelling via MBS is
to divide the real-world system into several bodies and use connection elements, like joints, springs
and dampers to connect these bodies. The technique of MBS is an internationally standardised method
for the derivation of an idealised dynamic model of a mechanical system and is a part of the classical
mechanical engineering [5]. The dynamics of the MBS can also serve as a basis for vibration analysis
and model-based control design [6]. The application of MBS for the active rope-free PTS is therefore
a natural one; the separation in rigid bodies’ results from the three main parts of the lift car. Another
advantage of MBS is their extensibility. If the flexibility of a body is crucial to the overall dynamic
behaviour of a system this body can be replaced by a flexible body, resulting in a MBS [7].
The goal of this paper is to presents the modelling for active cabin vibration damping on the example
of a simplified version of the rope-free PTS. The following chapter will give a brief outline of the
expected vibrations in the environment of a rope-free transportation system. The third chapter will
give a short overview of the basic steps needed to derive a MBS and also give a selection of modelling
elements used in the context of MBS. The succeeding chapters will give a short overview over the
complex model and display the derivation of the dynamic model on a two-dimensional simplified
version of the passenger transportation system. The influence of periodic imperfections in the guide
rails on the cabin will also be simulated with the simplified model. In the conclusion, an outlook over
the further steps in the design of active vibration damping will be given.
2 VIBRATIONS
The aim of active cabin damping is the improvement of the passenger’s riding comfort. One crucial
step for an improvement of the passenger comfort is to reduce the vibrations inside of the cabin,
because these vibrations are directly sensed by the passengers. The vibrations that are induced by
external effects have different sources, but the most significant vibration source are imperfections in
the guide rails. The vibrations induced by the rails can be separated in rail joints, e.g. gaps between
two rails, and periodical imperfections in the rail itself. The frequency of both vibration excitations
by the rails depend on the travelling speed of the passenger transportation system. The excitation by
rail joints happens over a very short period of time, thus they can be very high-frequent, while the
periodically induced vibrations are lower than 2 Hz, even for the top speed of the passenger
transportation system. Another source of high-frequent vibrations is the linear motor. The vibrations
excited by the linear motor lie mainly in the driving direction of the passenger transportation system,
whereas the vibrations induced by the guide rails are oriented in all directions except the driving
direction. Additional to the lateral vibrations, the guide rails induce rotational vibration, because of
the offset in the placement of the roller guides, which is similar to the placement in conventional lifts.
Depending on the sort of the active component, the vibrations that they are able to damp are restricted.
Most active components cannot damp high-frequency vibrations, like the frequencies induced by the
linear motor. Usually high-frequent vibrations have small amplitudes and can be efficiently damped
by passive damping elements, therefore a combination of passive and active components is desirable.
In this paper the main focus lies in the deriving of a dynamic model. The vibrations used for the
simulation correlate with periodically induced vibrations by the guide rails.
3 THEORY OF MULTI-BODY SYSTEMS

The following procedure is the standard approach of deriving a dynamic model using MBS. This
chapter will give a basic summary very closely related to the treatment presented in [5]. It will mainly
describe the elements of an MBS that are necessary for the modelling of the rope-free PTS. In the
context of MBS a few idealisations are made for the model. The model consists of rigid bodies with
inertia; there exist reference points, thus explicit points on the bodies, like the centre of gravity, where
e.g. forces act on, and all of these points have their own coordinate system. Coupling elements are
massless and generate applied forces and torques following a known law. Joints elements are also
massless and are frictionless in the motion direction and rigid in the locking direction, and are
therefore ideal joints. A selection of modelling elements is shown in Table 1.
Table 1 Selection of modelling elements
Element Representation Properties

Mass, inertia
Rigid Body
C: Centre of gravity
Coupling Spring Stiffness, free-length

Elements Damper Damping coefficient
Binding Spherical
Elements Joint Rotation around a point
Prismatic Translational movement along

Joint one axis
Rigid
No movement
Clamping
The complete MBS will describe the dynamical behaviour, and thereby the motion of the real system.
The motion of real mechanical systems can be described by a finite number of degrees of freedom.
In the MBS a single body in a three-dimensional space (3D) has six degrees of freedom, thus it can
move along all three direction in space and rotate around the three axes in space. In a two-dimensional
plane (2D), the degrees of freedom reduce to three, two movements along the axes in the plane and
one rotation. Therefore an unconstrained mechanical system with 𝑛 bodies has 6𝑛 degrees of freedom
in three dimensions and 3𝑛 in two dimensions. The movement of the whole system can be described
by the vector 𝒙 of the size 6𝑛 × 1, or 3𝑛 × 1 respectively for two dimensions. A MBS consist of
several connected bodies; the connections are described by constraints. The constraints reduce the
free motion and they are represented by binding elements like joints, see Table 1. A spherical joint
for example reduces the degree of freedom in 3D by three and in 2D by two. The degree of freedom,
thus the number of directions in which the MBS is able to move is than determined by
𝑓3𝐷 = 6𝑛 − 𝑞 and for two dimensions by 𝑓2𝐷 = 3𝑛 − 𝑞, (1)
where q is the number of constraints on the MBS. The motion of the system can therefore be described
by 𝑓3𝐷 or 𝑓2𝐷 independent generalised coordinates; one for each degree of freedom. The coordinates
are summarised in the vector 𝒚 of the size 𝑓3𝐷 × 1, and 𝑓2𝐷 × 1 respectively.
The topology of MBS can basically be divided into two categories: trees and loops. A MBS is
categorised as loop, if its bodies or part of its bodies form a closed circle. Without a closed loop the
MBS is categorised as a tree. In general, the equations of motion of MBS can be derived independent
of its topology. In the presence of a loop in the MBS, the loop has to be cut open and the dynamic
equations are first derived for the open loop, which then forms a tree. Cutting the loop means that the
dynamic of the model is described by the coordinates 𝒚𝑏 of the size 𝑓𝑏 × 1 more coordinates than are
actually needed to describe the motion of the closed loop MBS. The difference between the number
of degrees of freedom of the closed and the open loop is denoted by 𝑛𝑐 = 𝑓𝑏 − 𝑓, where 𝑓 denotes
the degree of freedom of the closed loop MBS. The additional coordinates are reduced by a closing
condition, which restricts the motion of the open loop to the possible motion of the closed loop. The
motion of the closed loop can then again be described by a minimal set of coordinates 𝒚 of the size
𝑓 × 1. The closing condition is denoted by a vector 𝒄(𝒚𝑏 , 𝑡) = 0 of the size 𝑛𝑐 × 1.
The generation steps; definition of bodies, definition of constraints and generalised coordinates and
if necessary the closing conditions of the MBS, are all part of the description of the kinematics of the
MBS, thus describe the possible motion of the MBS. The next steps are the derivation of the equations
of motion under the influence of external and internal forces and torques. These steps, are for the lack
of space, only sketched, for more details see [5, 6].
First the Newton-Euler equations are used to derive the unconstrained motion of the MBS by
establishing the principle of linear momentum and angular momentum. The equation of motion of a
free body was introduced by Euler in 1755 [8]. The description of a constraint MBS with minimal
coordinates is found using the principle of d’Alembert, which is the key for the derivation of equations
of motion from the Newton-Euler equations. The first consistent formulation was derived by
Lagrange in [9]. Using all this, the equation of motion can be written in the general form:
𝑴(𝒚, 𝑡)𝒚̈ + 𝒌(𝒚, 𝒚̇ , 𝑡) = 𝒒(𝒚, 𝒚̇ , 𝑡) (2)
In this form, the constraint forces between the bodies of the MBS are eliminated and the dynamic of
the MBS is described by a minimal set of independent generalized coordinates 𝒚. The matrix
𝑴(𝒚, 𝑡) is the symmetric 𝑓 × 𝑓-inertia matrix, which contains the mass moments of inertia and
masses of the bodies of the MBS. The vector 𝒌 of the size 𝑓 × 1 inherits the generalized Coriolis
forces and elastic and damping forces, thus the remaining forces after the constraint forces are
eliminated. The 𝑓 × 1-vector 𝒒 are the generalized applied forces, which contain the external forces
acting on the MBS, such as the gravitational force. The scalar 𝑡 represents time. It should also be
mentioned that the equations of motion can also be obtained by using the Lagrange’s equation of
the second kind formulated in [9].
The MBS includes a range of different parameters, e.g. the stiffness of the connecting springs and the
inertia of the bodies. These parameters are required for a numerical simulation and should match the
parameters of the real system. One way to choose these parameters is to perform measurements on
the real system. A way to achieve useful parameters if no real world system is available, is to use the
geometric data from the mechanical model of the system via CAD or perform additional numerical
simulations, like finite element analysis.
4 SIMPLIFIED MODEL OF THE ROPE-FREE TRANSPORATAION SYSTEM

The example presented in this paper is a model of the rope-free PTS, as shown in Fig. 1. The rope is
replaced by a linear motor, where the active elements of the motor are placed inside the lift shaft and
the passive elements are placed on the lift. The PTS consists of three main elements, which are moved
through the lift shaft. As mentioned before the three components are the sledge, the mounting frame
and the cabin. The sledge holds the passive elements of the linear motor, which provides an
electromagnetic force that is directly used to drive the transportation system through the shaft. The
mounting frame connects the sledge and the cabin. The principle structure of the PTS travelling in
the vertical direction is shown in Fig. 3. The relative motion between the three components is
constrained by connection elements, springs and dampers. The vibrations induced by the guiding
system directly affect the sledge and deviate the movement of the sledge. The novel design demands
a new vibration analysis of the system and therefore the derivation of a dynamic model of the
passenger transportation system.
In this paper, a two-dimensional model of the PTS is derived, in order to simplify the modelling
process. The main focus of the model lies in the investigation of the cabin vibrations, therefore in the
first approach the desired motion of the passenger transportation system is neglected and only the
relative movement between mounting frame and cabin is investigated.
Figure 3 Principle structure of the passenger transportation system, consisting of the sledge,
holding the armature of the linear motor; a mounting frame; and the cabin. The desired
movement 𝒗 is along the 𝒛-axis. The connection elements have a spring/damper dynamics.
Further, the sledge can be neglected in this approach, because it mainly conveys the deviations to the
mounting frame. The resulting model consists only of the mounting frame and the cabin. The
undesired vibrations induced by the imperfections in the rails are modelled as disturbances. These
disturbances directly act on the suspension of the mounting frame, thus the connection between sledge
and mounting frame. The resulting model is derived as a MBS, whose structure is shown in Fig. 4(a).
The MBS consists of four rigid bodies: the mounting frame, the cabin and two active damping
actuators underneath the cabin. The vibrations that are conveyed from sledge to mounting frame are
represented by force 𝐹𝑑 and the torque 𝑇𝑑 . The torque 𝑇𝑑 induces a rotational displacement as the
force 𝐹𝑑 induces a translational displacement on the mounting frame. The forces that can be applied
by the actuators are represented by the forces 𝐹𝐴1 and 𝐹𝐴2 . The actuator forces are used to move the
cabin. The force 𝐹𝑏 and torque 𝑇𝑏 are needed to directly influence the motion of the mounting frame
in the simulation. This force and torque pair is only a virtual input to the model without having a real
world equivalent. The torques 𝑇𝑑 and 𝑇𝑏 act on the origin O, which represents the sledge-sided
connection point between mounting frame and sledge. The forces 𝐹𝑑 and 𝐹𝑏 act on the point 𝐵𝑠 , that
is the connection point between mounting frame sledge on the mounting frame. The actuator forces
𝐹𝐴1 and 𝐹𝐴2 act on the centre of gravity of the actuators, thus 𝐶𝐴1 and 𝐶𝐴2 respectively. The centre of
gravity of the mounting frame and cabin are denoted by 𝐶𝑏 and 𝐶𝑐 , respectively. The gravitational
forces of each body act on the respective centres of gravity. Additional to the gravitational forces, in
all joints the internal coupling denoted by springs and dampers is implemented. The spring and
damper elements represent the elasticity of the connection between the bodies and are parameterised
to simulate the dynamic behaviour of the real system, especially to determine the natural frequency
of the system.
The connection of cabin, mounting frame and the two actuators form a closed loop. This closed loop
has to be cut open in order to derive the equations of motion. The closed loop is cut open at the points
𝑃1 and 𝑃2 , which represent the contact point between cabin and the respective actuator, see Fig. 4(b).
This cutting attains that the motion of the cabin can be described, as that of a free body in the plane,
by the two coordinates of the body-fixed point 𝑃0 underneath the cabin, denoted by 𝑥𝑐 , 𝑧𝑐 , and the
angle 𝛽𝑐 . Because the motion of the mounting frame is only constrained by connection elements, its
motion is described by coordinates of the body fixed point 𝐵𝑠 , denoted by 𝑥𝑏 , 𝑧𝑏 , and the angle 𝛽𝑏 .
The actuators can change their length and rotate around the contact points 𝐵1, and 𝐵2 respectively.
Therefore the movement of the actuators is described by their length and an angle, that is 𝑙1, 𝛽1 for
the first and 𝑙2 , 𝛽2 for the second actuator. Summarised this leads to 𝑓 𝑏 = 10 degrees of freedom for
the open loop system represented by the vector of the generalised coordinates 𝒚𝑏 =
[𝑥𝑏 , 𝑧𝑏 , 𝛽𝑏 , 𝑥𝑐 , 𝑧𝑐 , 𝛽𝑐 , 𝑙1 , 𝛽1 , 𝑙2 , 𝛽2 ]𝑇 . The generalised coordinates are also displayed in Fig. 4(b).
(a) Simplified two-dimensional Multi- (b) Generalised coordinates of the cut

Body system. open simplified model.
Figure 4 Simplified two-dimensional model consisting of mounting frame with centre of

gravity 𝑪𝒃 and cabin with centre of gravity 𝑪𝒄 . In (a) are the disturbances from the guidance
represented by the force 𝑭𝒅 and torque 𝑻𝒅 , the forces 𝑭𝑨𝟏 and 𝑭𝑨𝟐 represent potential
actuators forces and in blue are the gravitational forces. In (b) the generalised coordinates of
the open loop are displayed, namely 𝒚𝒃 = [𝒙𝒃 , 𝒛𝒃 , 𝜷𝒃 , 𝒙𝒄 , 𝒛𝒄 , 𝜷𝒄 , 𝒍𝟏 , 𝜷𝟏 , 𝒍𝟐 , 𝜷𝟐 ]𝑻 .
The closing condition is given by connecting the end point of the actuators with the points 𝑃1 and 𝑃2
respectively. For the first actuator the closing condition is given by
𝐼
𝒄𝐴1 (𝒚𝑏 ) = 𝒓𝐼𝑃1 (𝑥𝑐 , 𝑧𝑐 , 𝛽𝑐 ) − 𝒓𝐴1end (𝑥𝑏 , 𝑧𝑏 , 𝛽𝑏 , 𝑙1 , 𝛽1 ). (3)
In the closing condition the vector 𝒓𝐼𝑃1 (𝑥𝑐 , 𝑧𝑐 , 𝛽𝑐 ) describes the position of 𝑃1 with the coordinates of
the cabin and the vector 𝒓𝐴1end
𝐼
(𝑥𝑏 , 𝑧𝑏 , 𝛽𝑏 , 𝑙1 , 𝛽1 ) describes the end point of the first actuator with the
coordinates of the mounting frame and the coordinates of the first actuator. The loop for the second
actuator is closed in the same fashion. The loop closure reduces the degrees of freedom by 𝑞 = 4,
because two spherical joints are closed, which leads to 𝑓 = 6 degrees of freedom. The minimal
coordinates can then be chosen to 𝒚 = [𝑥𝑏 , 𝑧𝑏 , 𝛽𝑏 , 𝑥𝑐 , 𝑧𝑐 , 𝛽𝑐 ]𝑇 .
The next step is to derive the Newton-Euler equations of motion for each body separately depending
on the generalised coordinates 𝒚𝑏 . The Newton-Euler equations are derived in one point for each
body, in the present example, the points are 𝑃0 for cabin, 𝐵0 for mounting frame and 𝐶𝐴1 , 𝐶𝐴2 for the
actuators. The contact points of the forces have to be described with respect to these points in the
generalised coordinates 𝒚𝑏 . In the present case, the contact points are the positions of the joints and
the centres of gravity of the bodies. For the mounting frame for example, the distances to the points
𝐵1 , 𝐵2 , 𝐵𝑠 and 𝐶𝑏 from the point 𝐵0 have to be formulated, given by the vectors 𝒓𝐵0 𝑏 𝐵0 𝑏
𝐵1 (𝒚 ), 𝒓𝐵2 (𝒚 ),
𝐵𝑠 (𝒚 ) and 𝒓𝐶𝑏 (𝒚 ). The forces that act on these points of the mounting frame are the negative
𝐵0 𝑏
𝒓𝐵0 𝑏
actuator forces 𝐹𝐴1 and 𝐹𝐴2 , the disturbance force 𝐹𝑑 , the gravitational force 𝑭𝑔,𝑏 = [0, −𝑚𝑏 𝑔]𝑇 and
the force 𝐹𝑏 . Additionally, the disturbance torque 𝑇𝑑 and the torque 𝑇𝑏 attack at the mounting frame.
Additionally, the here not listed spring and damper torques and forces in the joints have to be
considered. Using the vectors from 𝐵0 to the contact points, the forces and torques the Newton-Euler
equations can be formulated. This procedure has to be repeated for all bodies. Afterwards the principle
of d’Alembert is used to derive the equations of motion of the overall system in the form Eq. 2.
In this step it was assumed that all vectors were already in the same orientation. Normally the vectors
have to be rotated by the respective angle, which describes the rotation of the body, in order for them
to have the same orientation.
5 SIMULATION
In this chapter a simple simulation is performed to investigate the influence of periodic imperfections
in the guide rails. The simulation is performed using the software environment MATLAB/Simulink.
The travelling speed of the PTS is assumed to be 𝑣 = 10 m/s and the period of the excitation is
assumed to correspond to the length of two rails given as 10 m, thus has a frequency of 1 Hz. Further,
it is assumed that the excitation results in a torque 𝑇𝑑 and a force 𝐹𝑑 in 𝑥-direction acting upon the
mounting frame, shown in Fig 5, that corresponds to a rail unevenness of 3mm. The torque 𝑇𝑑 and
force 𝐹𝑑 are given by sinusoidal signals with amplitudes of 𝐴𝑇𝑑 = 17084 Nm, 𝐴𝐹𝑑 = 3871 Nm
and a frequency of 1 Hz. Furthermore the torque was filtered to achieve a smooth input for the
simulation with the filter 𝐹(𝑠) = 517.8/(𝑠 + 5)3 . The internal dynamics in the rotational axis around
𝐵𝑠 is chosen, such that it matches the eigenfrequency of 10 Hz along the rotation of the mounting
frame. The spring stiffness has been assumed as shown in Table 2 and the damping parameters have
been chosen to have the numeric value of the square route of its respective stiffness. Table 2 also
states the initial distances used in the simulation. The simulation is started from the equilibrium point,
hence the input forces are chosen to compensate the weight of the bodies, leading to:
𝐹𝐴1 = 𝐹𝐴2 = 5003 N, 𝑭𝑏 = [0,1.2 ⋅ 104 ]T N, Tb = −9479 Nm (4)
The important factor for the passenger comfort are the vibrations felt by the passengers inside the
cabin; the absolute value of vibration inside the cabin is shown in the Fig 6. It is visible that in this
simple simulation the acceleration exceeds the 10 milli-g border, which is here used to define
acceptable ride quality for a vibration with a frequency of 1 Hz.
Figure 5 Disturbance torque 𝑻𝒅 representing the periodic excitation through imperfection in

the guide rails.
Figure 6 Absolute value of the acceleration measured inside of the cabin.
Table 2 Parameter and initial distances for the simulation
Parameter Value Parameter Value

Mass [kg] Inertia [kg m²]
Cabin 𝒎𝒄 = 𝟏𝟎𝟎𝟎 Cabin 𝑰𝒄 = 𝟏𝟎𝟎𝟎
Mounting 𝒎𝒃 = 𝟐𝟎𝟎 Mounting 𝑰𝒃 = 𝟐𝟎
Actuator 𝒎𝑨 = 𝟏𝟎 Actuator 𝑰𝑨 = 𝟎. 𝟎𝟕𝟓
Initial distances [m]
𝑩 𝒔 → 𝑪𝒃 𝒓𝑩𝒔
𝑪𝒃 = [𝟎. 𝟏𝟔𝟓, −𝟎. 𝟔]
𝑻 𝑶 → 𝑩𝒔 𝒓𝑶
𝑩𝒔 = [𝟎, 𝟎]
𝑻
𝑩 𝟎 → 𝑪𝒃 𝒓𝑩𝟎
𝑪𝒃 = [−𝟎. 𝟕𝟓, 𝟎. 𝟑]
𝑻 𝑷𝟎 → 𝑪𝒄 𝒓𝑷𝟎
𝑪𝒄 = [𝟎, 𝟎. 𝟕𝟔𝟑]
𝑻
𝑩𝟎 → 𝑩𝟏 𝒓𝑩𝒔
𝑩𝟏 = [−𝟎. 𝟓, 𝟎]
𝑻 𝑷𝟎 → 𝑷𝟏 𝒓𝑷𝟎
𝑷𝟏 = [−𝟎. 𝟓, 𝟎]
𝑻
𝑩𝟎 → 𝑩𝟐 𝒓𝑩𝒔
𝑩𝟐 = [𝟎. 𝟓, 𝟎]
𝑻 𝑷𝟎 → 𝑷𝟐 𝒓𝑷𝟎
𝑷𝟐 = [𝟎. 𝟓, 𝟎]
𝑻
Translational Stiffness [N/m] Rotational Stiffness [N/rad]

Point 𝑩𝒔 (𝒙) 𝒌𝒙𝑩 = 𝟏𝟎𝟔 Point 𝑩𝒔 𝒌𝜷𝑩 = 𝟕. 𝟕𝟒 ⋅ 𝟏𝟎𝟔
Point 𝑩𝒔 (𝒛) 𝒌𝒛𝑩 = 𝟏𝟎𝟔
Actuator 1 𝒌𝒍𝟏 = 𝟏𝟎𝟔 Point 𝑩𝟏 𝒌𝜷𝟏 = 𝟏𝟎𝟔
Actuator 2 𝒌𝒍𝟐 = 𝟏𝟎𝟔 Point 𝑩𝟐 𝒌𝜷𝟐 = 𝟏𝟎𝟔
6 OUTLOOK AND CONCLUSION

In this paper the modelling procedure for active vibration damping has been investigated. The basic
steps for the mechanical modelling as a Multi-Body system were given. A simplified model to be
used for the vibration analysis of the rope-free PTS was derived. The model consists only of the
mounting frame, the cabin and actuators. The dynamic behaviour of this model was simulated using
a periodic excitation, which represents the imperfections in the guide rails. It was shown that for this
simple simulation unacceptable vibrations are induced inside the cabin.
The next steps for the model will be to implement the actuator dynamics in the model. A crucial
further step is the validation of the model by measurements on the real-world system. These
measurements are especially important to identify the parameters of the MBS so its behaviour
matches the behaviour of the real system. In the context of active cabin damping the dynamic model
forms the basis to design a control concept for the active cabin damping. The dynamic model will
also be used to test these concepts in simulation before they are implemented on the real system.
Another application for the model can be the generation of a smooth guiding, thus suitable paths for
the real system. The trajectories could be designed in such a way that excitation of vibrations are kept
to a minimum.
7 REFERENCES
[1] A. Hamdy, “Active damping of vibrations in elevator cars". J. Struct. Control, Vol. 6, No. 1,
53–100 (1999).
[2] N. Noguchi, A. Arakawa, K. Miyata, T. Yoshimura, and S. Shin, “Study on Active Vibration
Control for High-Speed Elevators". JSDD, Vol. 5, No. 1, 164–179 (2011).
[3] G. F. Franklin, J. D. Powell, and A. Emami-Naeini, Feedback control of dynamic systems, 6th
ed.: Pearson. Upper Saddle River, NJ (2010).
[4] N. S. Nise, Control systems engineering: John Wiley. New York NY (2014).
[5] W. Schiehlen, “Multibody System Dynamics: Roots and Perspectives". Multibody System
Dynamics, Vol. 1, No. 2, 149–188 (1997).
[6] W. Schiehlen, N. Guse, and R. Seifried, “Multibody dynamics in computational mechanics
and engineering applications". Computer Methods in Applied Mechanics and Engineering,
Vol. 195, No. 41-43, 5509–5522 (2006).
[7] A. A. Shabana, “Flexible Multibody Dynamics: Review of Past and Recent Developments".
Multibody System Dynamics, Vol. 1, No. 2, 189–222 (1997).
[8] Euler L, “Nova methodus motum corporum rigidorum degerminandi". Novi Commentarii
academiae scientiarum Petropolitanae 20, 208–238 (1776).
[9] J. L. Lagrange and J. Bertrand, Mécanique analytique: Mallet-Bachelier (1855).
Jonas Missler received his bachelor’s degree in Engineering Cybernetics from the University of
Stuttgart, Germany. He also obtained his master’s degree in Engineering Cybernetics from the
University of Stuttgart. Since 2015, he is working towards his Ph.D. at the Institute for System
Dynamics at the University of Stuttgart. His current research interests are the developing of an active
damping concept and the respective control scheme for rope-free PTS.
Thomas Ehrl, Mechanical Design Engineer (degree in 1994), Part-time PhD student with the School
of Science and Technology of The University of Northampton. Professional career started in 1994.
Since 4/2008 with thyssenkrupp Elevator: Head of Research & Innovation Center of TKE Innovation
GmbH, Pliezhausen;Engineering Training Manager at Corporate Level of TKE AG; Manager R&D
Project Standards at Corporate Level of TKE AG. Interests: Travelling, sports, soccer, music, vintage
English motor cycles, model building, photography.
Benedikt Meier received his Diploma in Mechanical Engineering from the University of Hannover,
Germany. In 1992, he obtained his doctorate in Cold Testing of combustion engines. In thyssenkrupp
Elevator AG, he is leading the Global Project Management Office (PMO). Since July 2015, he serves
as Visiting Professor in the School of Science and Technology at the University of Northampton. His
expertise is in the area of horizontal and vertical transportation and material handling systems. In
addition, he is an internationally recognized expert in Project and Program Management. Professor
Meier has published several journal and international conference papers in his area of expertise.
Stefan Kaczmarczyk has a master’s degree in Mechanical Engineering and he obtained his doctorate
in Engineering Dynamics. He is Professor of Applied Mechanics and Postgraduate Programme
Leader for Lift Engineering at the University of Northampton. His expertise is in the area of applied
dynamics and vibration with particular applications to vertical transportation and material handling
systems. He has been involved in collaborative research with a number of national and international
partners and has an extensive track record in consulting and research in vertical transportation and
lift engineering. Professor Kaczmarczyk has published over 90 journal and international conference
papers in this field. He is a Chartered Engineer, being a Fellow of the Institution of Mechanical
Engineers, and he has been serving on the Applied Mechanics Group Committee of the Institute of
Physics.
Professor Oliver Sawodny received his Dipl.-Ing. degree in electrical engineering from the University
of Karlsruhe, Karlsruhe, Germany, in1991 and his Ph.D. degree from the University of Ulm, Ulm,
Germany, in 1996. In 2002, he became a Full Professor at the Technical University of Ilmenau,
Ilmenau, Germany. Since 2005, he has been the Director of the Institute for System Dynamics,
University of Stuttgart, Stuttgart, Germany. His current research interests include methods of
differential geometry, trajectory generation, and applications to mechatronic systems. He received
important paper awards in major control application journals such as Control Engineering Practice
Paper Prize (IFAC, 2005) and IEEE Transaction on Control System Technology Outstanding Paper
Award (2013).
Dynamic Lift Control for Improvements in Energy Efficiency
Vicente Pacheco
IMEM LIFTS, R&D Dept. Adarzo Str. 167B. Santander. 39011. Spain. [email protected]
Keywords: Energy efficiency, Dynamic lift control, Direct to floor, PMSM motors, ISO 25745-2.
Abstract. A lift’s energy behaviour is an important issue and R&D departments are constantly
searching for ways to improve results. Focusing on the electrical-electronic area, it is already well-
known that the use of 3VF inverters and PMSM motors allows better energy results to be achieved.
The combined use of real-time communications between lift control and inverter and the use of
“direct approach to floor” function (we suggest to call this feature Direct To Floor in the paper)
allows the realisation of an energy decision-making control panel and improvements to traditional
energy consumption. From this point onwards, our objective is to present an improved concept for
energy-efficiency based on the development of a new dynamic control. To achieve this, the
following is important: 1 - Identify the different behaviours of the lift with regard to energy-
efficiency in each different stage of the journey, taking into account: the number of people
travelling, the direction of travel, the distance to be travelled and the lift’s speed. 2 - Propose
energy-saving improvements for each stage, always using DTF & sharing information in real time
as a basis. 3 - Develop an intelligent control capable of taking decisions affecting energy-efficiency
in real time. This allows the best energy-saving profile to be selected for each journey, adapting the
curves as well as the motor and brake control in any situation. 4 - Using a certain energy profile and
incorporating a certain set of proposals can produce good results in some circumstances and only
acceptable results in others. For this reason, the smart lift control must always select the most
suitable option. 5 - Show a comparative analysis of the results obtained with the Smart-ECO mode
with traditional solutions, as well as a comparison with current regenerative systems. In this paper,
results with ISO25745-2:2015 are also shown. The aim of this paper is to make an in-depth
presentation of the studies carried out for the journey stages, the proposals and the obtained results.
All the results shown have been taken from real lift installations.
1 INTRODUCTION
Given the increasing importance of energy efficiency to the lift sector, there is already a large
number of studies and reports on this subject, of which we reference some of the best-known
[1,2,3,4,5,6,7]. There are also several kinematic analyses, such as [8,9]. To date however, we have
found no papers with the concentrated focus on energy efficiency which we are offering here.
What we set out to do was to analyse the energetic behaviour of a traction lift with a gearless
permanent-magnet synchronous motor (PMSM) during the various stages of its travel, starting with
analysis of its performance in DTF (Direct To Floor or Direct Approach) mode. In [10] we present
evidence of greater energy-saving in DTF mode as compared to that achieved with the traditional
speed curve profile and the standard approach speed.
All our data was generated by tests carried out in our testing tower, using a lift system with a travel
distance of 15.31 metres, a 1000 Kg lift car, a gearless PMSM, a 2:1 roping ratio, 50%
counterweighting and a compensation chain, at a travel speed of 1 metre per second.
We carried out several hundred tests (511 measurements), taking measurements with a FLUKE 435
II Power Quality & Energy Analyzer and software applications for 3VF frequency inverters which
included NCDrive trace and precision data logging functions[11].
Our first step was to carry out tests to measure the energetic behaviour of the lift under different
conditions, using the following variables: the number of passengers inside the lift car, the speed
reached during the journey, the direction of travel and the distance travelled. We also used the
motor with different standard control settings to see how each one affected the amount of energy
consumed.
Once we had recorded the detailed variations in the lift's energetic behaviour, we tested out
different speed curve settings and various electrical configurations aimed at lowering energy
consumption, while at all times maintaining the optimum passenger ride comfort, thereby
generating an extensive database produced by the settings and configurations for each individual
journey type.
With the energy consumption improvements achieved, we created a simulator which allows us to
configure the lift traffic in different types of buildings and during different time periods. This
simulator enables us to predict energy consumption by each particular lift system.
Using real-time communications (RTC) and the Direct Approach mode, our aim is to develop an
intelligent lift control system that, right before the start of the journey, can use the detailed data
mentioned above to select the most energy-efficient consumption profile to carry out the required
task.
We have also calculated the impact that these improvements could have on lift system rating
according to ISO 25745-2.
We carried out tests to analyse the lift's energy efficiency when fitted with a regenerative drive
system (a regenerative kit very easy to connect to the control panel) which feeds electricity back
into the building's power grid, using an up-to-the-minute device produced by a European firm with
an excellent reputation in the lift sector. The resulting data was incorporated into our simulator
system.
2 OVERALL APPROACH
During lift travel, electricity is consumed by the control system itself as it manages the functioning
of the lift and also by the traction machine with energy wasted in the brakes and by the motor itself.
Given this, as a general rule less travel time implies less waste of energy by the inverter and the
motor as well as by the control system.
However it is not always the case that a shorter travel time will result in energy-saving, because it
may be accompanied by increased energy use despite the reduced time frame.
2.1 Phase 1: Control system at 0 hz: motor start-up, brake release & control of roll-back
(rb) effect
During this phase, it is important to avoid causing unnecessary delays, given that they waste energy,
and to ensure a smooth and comfortable start-up.
A PMSM does not consume active electric power until it is required to generate torque. One option
in the 0 Hz phase is to control the start-up based in induction motors control (using the
magnetisation times that these motors require). This reactive power is not present at the lift's
connection to the mains, but the consequent activation of the control system and the frequency
inverter does lead to energy wastage. Dispensing with this period would therefore save energy.
Fig.1, note 1 in the text.
Once the brake release command, MecBrakeOpen at Fig.1 in the text, is delivered, the rotor is
unlocked and it is crucial that the system accurately determines the exact degree of torque to be
applied immediately in order to avoid any RB effect.
Dynamic Lift Control for Improvements in Energy Efficiency 16-3
Irrespective of the direction of travel, the drive unit will always consume electrical power
proportionate to the difference between the lift car and the counterweight loads (Fig.1, note 2 in the
text.)
Car load 300 Kg, going up, 6 floors. For demostration purposes, lenght of processes 1,2,3 and 4 is
augmented
1400
1200
MotorPower (W)
1000
1. Unnecessary delays previous to the brake release command
Power (W)
800 2. Motor consumption with brakes released and without car movement. MC_Run (signal
3. Motor consumption during smooth start phase (to minimise effects from to command the
frictions with the guide rails) IGBTs to feed the
600
4. Excessive time after brakes engagement motor)
400 2 4 MecBrakeOpen
3
200
1
FreqRampOut
0
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
5.5
6
6.5
7
7.5
8
8.5
9
9.5
10
10.5
11
11.5
12
12.5
13
13.5
14
14.5
15
15.5
16
16.5
17
17.5
18
18.5
19
19.5
20
20.5
21
21.5
22
22.5
Time(s)
Figure 1. Lift going up, 6 floors, car load 300 kg.

It is well-known in the lift industry that to control the RB effect, a weighting device in
communication with the control system can be an effective solution, although depending on its
mechanical design and in-car load placement, this could be a not very accurate operation.
An alternative way: The control of the brake release time. It is good practice to control this using
brake switches in order to determine the exact motor setting required to avert the RB effect.
Both solutions (weighting device and brake switches) have an added cost. A slight increase in
electric power consumption is produced with the weighting device option (significant when the lift
is idle or on standby). Brake switches, unless they are inductive, tend to experience malfunctions
that cause unnecessary energy usage.
Once the rotor is completely unlocked, it is essential to terminate the 0 Hz control phase (Fig.1,
notes 1 and 2 in the text) and initiate a smooth start-up speed curve (Fig.1, note 3 in the text).
Reducing the duration of 0Hz phase is significant because 1) it reduces travel time and improves
ride comfort, and 2) it cuts down the time spent generating wastage (Fig.1, note 3 in the text).
Energy-saving proposal: to detect and interpret (without brakes switches and without weighting
device) the electrical variables in the motor as well as the encoder signals, and determine the precise
degree to torque required.
The study of the electrical variables also enables us to calculate optimum brake release timing at
any moment (as this can also change).
A fast and accurate energy-supply response avoids unnecessary delays in PMSM operation. As soon
as the brake has been released, the smooth start-up phase must begin immediately.
2.2 Phase 2: Smooth start-up at low speed

Potential disturbances in the ride comfort, because of any possible mechanical frictions on the guide
rails, are minimized leaving the floor at low speed during a brief period of time.
The energy consumption depends on the difference between car load and the counterweight
balance, without taking into account any losses because of the mechanical and electrical
efficiencies. Independently of the in-car load, this phase always involves energy consumption. Even
when the motor is operating in generative mode, mechanical and electrical wastage in the system
obliges the 3VF inverter to feed in electrical power.
Energy-saving proposal: In this phase as well, the time taken to execute a smooth start-up must be
cut to a minimum, while guaranteeing ride comfort. The shorter the phase, the more efficient it is.
It is important that this phase is designed specifically to suit the specific mechanical characteristics
of the particular lift system. Otherwise there will be small travel delays and higher energy
consumption, with no improvements in ride comfort (Fig1, note 3 in the text).
2.3 Phase 3: Acceleration curve (initial jerk, or jerk 1 constant acceleration ramp
second jerk, or jerk 2 to achieve constant speed)
As is well-known, the lift's energy consumption is directly proportional to the torque produced
during the acceleration ramp.
2.3.1 Journeys where the motor works in generative mode

During the acceleration phase, the motor can achieve generative mode either very shortly after ramp
start-up or later on. Also, sometimes the motor can alternate between generative and motor modes
during the acceleration ramp. This produce different energy profiles even when the acceleration
curve is exactly the same, depending on the load in the car.
Both Fig.2 and Table 1 in the text show how the motor's electricity demand has a direct relationship
with the in-car loading (both the peaks in consumption and the form of the fluctuations).
Fig. 2 in the text shows the power demand to the mains when the lift is going up and the
counterweight is heavier than the car. The area created under the power curve is the electrical
energy used. It is seen that the lift reaches the generative mode (and no electrical energy is
demanded any longer by the motor) when the power becomes stable to 600 W.
Going up. Different loads in the car

2000
1800
1600
1400
0 Kg DTF
1200
Power (w)
1000 150 Kg DTF
800 300 Kg DTF

600 350 Kg DTF
400
400 kg DTF
200
0 450 kg DTF
0
1.5
4.5
7.5
12
15
18
21
0.75
2.25
3.75
5.25
6.75
8.25
9.75
10.5
11.25
12.75
13.5
14.25
15.75
16.5
17.25
18.75
19.5
20.25
21.75
22.5
Time (secs)
Figure 2.Going up with different loads in the car (Measurement taken at connection to the
mains). For clarity's sake, the timer to feed the brakes at 200/100 volts has been switched off.
They were connected to 200 VDC during the complete journey. Real power consumption is lower.
Table 1: Time to achieve performance as generator

Nominal load : 1000 [kg], Time elapsed since start of car movement until electrical power supply
upwards travel, Accel. = to motor ceases (measured from frequency inverter output)
0.55[m/s2], Jerk1= [0.55m/s3]
0 [kg] 1.679[ms]
150 [kg] 1.748[ms]
300 [kg] 2.702[ms]
450 [kg] 3.449 [ms]
So, it is shown how the time (Table 1 in the text) and the energy profile to reach this generative
mode depend on the car load. During this period, the motor is demanding electrical energy from the
mains.
On the same acceleration ramp, the lesser the difference between car and counterweight loading, the
longer this period lasts.
Energy-saving proposal: Generally speaking, in terms of reducing electricity consumption, the
motor should achieve electricity production capability (generation mode) as soon as possible;
although, when the car/counterweight load difference is very small, we occasionally observe the
opposite effect (we analyze this later in the paper). Of course, the acceleration and jerk values must
keep the ride into the levels of comfort already well-known in the lift industry.
In the majority of cases, the electrical energy generated by the motor will simply be converted into
heat by the brake resistor. However, the lift will have made its journey drawing less electrical
energy from the mains supply.
2.3.2 More detailed observations: car loads between 30 & 45% (ascending) and 55 & 70%
(descending)
From 300 Kg upwards, the lift tested required electrical energy almost until only just before it
reached its nominal speed.
Energy-saving proposal: in this case, reducing the rate of acceleration results in the motor
generating its own electrical energy earlier and without causing significant increases in journey
time. During a journey of 15.31 metres at 1m/s, the consequent delay is less than half a second
(Fig.3 in the text).
Going up 6 floors 1 m/s

1200
1000
Power (w)
800
600
400 300 Kg ECO
200
300 Kg DTF
0
0.25
1.25
2.25
3.25
4.25
5.25
6.25
7.25
8.25
9.25
10.25
11.25
12.25
13.25
14.25
15.25
16.25
17.25
18.25
19.25
20.25
21.25
Time(sec)
Figure 3. Areas under the power curves represent the electrical energy demand. This Fig.
shows how energy generation begins earlier (about 1.5 s earlier approx.) and energy
consumption is reduced if ECO profile is used. For clarity's sake, the timer to feed the brakes at
200/100 volts was configured to keep the brakes at 200 VDC during 7 seconds approx. Real
power consumption is lower. Because of this, the power in the graph goes down beyond instant
7.25 sec.
When the loading difference between the car and the counterweight is even smaller, modifying the
Jerk2 (from constant accel. ramp to constant speed) value and reducing the rate of acceleration,
results in the motor's generation of electrical energy being delayed even longer, rather than starting
earlier.
However, this also results in a reduction in the lift's energy consumption (Fig. 4 in the text).
Going up 6 floors 1 m/s

2000
Power (w)
1500
1000
500 450 kg DTF
0 450 Kg ECO
0.25
1.25
2.25
3.25
4.25
5.25
6.25
7.25
8.25
9.25
10.25
11.25
12.25
13.25
14.25
15.25
16.25
17.25
18.25
19.25
20.25
21.25
Time (sec)
Figure 4. Areas under the power curves represent the electrical energy demand. It is shown
how energy generation begins later (about 1 s later approx.) using the ECO profile. For
clarity's sake, the timer to feed the brakes at 200/100 volts was configured to keep the brakes at
200 VDC during 7 seconds approx. Real power consumption is lower. Because of this, the power
in the graph goes down beyond instant 7.25 sec.
Energy-saving proposal: Acceleration and Jerk2 have to be selected at appoint that journey time is
not prolonged too much. Also this would increase electricity wastage and energy consumption
rather than reduce it.
2.3.3 Journeys with the traction machine working in motor mode

We found some acceleration and jerk values which, while always maintaining the levels of ride
comfort standard in the industry, allowed us to adjust the travel time and reduce power wastage.
Differences for the selected values exist depending on the loads in the car and speed to be achieved.
Exceeding these values (in addition to having an unacceptable effect on ride comfort) sometimes
resulted in higher energy consumption despite the shorter journey time, and going below them also
led to increased consumption.
2.4 Travel Velocity: Surpassing motor's rated speed by up to 20%
In [10] we explain how, thanks to real-time communication (RTC) between the control system and
the 3VF frequency inverter, the rated speed of the motor can be surpassed when the imbalance
between car and counterweight loading is not at its maximum. This cuts travel time, which in turn
reduces energy wastage.
Reaching a travel speed above and beyond the lift's rated speed requires the delivery of greater
kinetic energy, which generally results in increased energy consumption until that particular higher
speed is reached.
Thanks to RTC, the control system can select a particular speed at the start of the journey on the
basis that it is the most energy-efficient. In order to do so, given that kinetic energy is a function of
mass and velocity, it has to evaluate the distance to be travelled and the load to be carried. The
decision to exceed rated speed during the journey is only justifiable in energy terms when the
distance to be travelled is far enough to mean that the savings in energy wastage are greater than the
initial expenditure in energy supply.
2.5 Phase 4: Car-to-landing approach manoeuvre & approach speed
Fig. 5 in the text, taken from [10], reveals that, when not in DTF mode, the motor starts to consume
electricity as soon as approach speed is reached. This is always the case, independently of the in-car
load and of the direction of travel.
DTF mode overrides the standard approach speed, results in faster journey times and so the lift
spends less time generating wastage (in the control system, brakes and motor).
Going down and going up with empty car
5000
4500
4000 With DTF
3500
3000 Without DTF
Power (W)
2500
2000
1500
1000
500
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
Time (sec)
Figure 5. Lift going down and going up with empty car.

2.6 Phase 5: Arrival at stop: jerk4 (from constant deceleration to 0 hz), 0 hz phase &
demagnetisation of motor
Fig. 6 in the text shows that when the lift reaches speeds approaching 0 Hz, the motor always
consumes electricity, even if it has been operating in generative mode throughout the rest of its
journey.
DTF: Going up. Different loads in the car

2000
1500 0 Kg DTF
Power (w)
150 Kg DTF
1000
300 Kg DTF
500 350 Kg DTF
400 kg DTF
0
450 kg DTF
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Time(sec)
Figure 6. 1000 Kg lift, at 1m/s, ascending. Measurement at connection to the mains. For
clarity's sake, the timer to feed the brakes at 200/100 volts has been switched off. They were
connected to 200 VDC during the complete journey. Real power consumption is lower.
Also, for a few moments supply to the motor remains at 0 Hz, both before and after the brakes are
switched off. See also Fig.1, note 4 in the text.
One way of achieving energy savings is to reduce the duration of this phase to an absolute
minimum without interfering with ride comfort. Once the brake is blocking rotor movement, it is
reasonable to start to demagnetize. To do this efficiently, it is very important to identify the exact
timing of brake engagement.
Fig. 6 in the text also shows how consumption is conditioned by car/counterweight load imbalance.
So it is vital to reduce this phase to a minimum, but it is counterproductive to make the cut in power
too abrupt, as this can make motor operation too noisy.
3 RESULTS
After carrying out hundreds of controlled tests, we have developed speed curve profiles and sets of
electrical adjustments which correspond to each of the phases described above: according to
percentages of rated load (0%, 15%, 30%, 45%, 55%, 70%, 85% or 100%), according to direction
of travel (upwards or downwards), according to the number of stops (2, 3, 4, 5 or 6) and according
to the speed of travel (1m/s, or 1.2m/s where this is possible).
For reasons of space in this paper, here we will present only the energy expenditure and savings
recorded (real measurements) for journeys of two stops (6.31 metres) and six stops (15.31 metres).
Table 2. Energy expenditure and savings identified for journeys of two stops and six stops.
Energy demand (Wh) Savings (%)

Approach speed
(0.05 m/s) DTF Dymamic control DTF Vs Approach speed Dynamic control Vs Approach speed
[kg] Up [Wh] Dw[Wh] Up[Wh] Dw[Wh] Up[Wh] Dw[Wh] Up (%) Dw (%) Up (%) Dw (%)
6 floors
0 3.336 30.422 2.154 29.592 1.971 29.292 35.432 2.728 40.917 3.714
150 2.739 21.793 1.981 21.338 1.624 20.966 27.674 2.088 40.708 3.795
300 2.371 13.75 1.882 13.572 1.59 13.01 20.624 1.295 32.940 5.382
450 2.611 7.036 2.166 6.982 2.005 6.825 17.043 0.767 23.209 2.999
550 7.036 2.611 6.982 2.166 6.825 2.005 0.767 17.043 2.999 23.209
700 13.75 2.371 13.572 1.882 13.01 1.59 1.295 20.624 5.382 32.940
850 21.793 2.739 21.338 1.981 20.966 1.624 2.088 27.674 3.795 40.708
1000 30.422 3.336 29.592 2.154 29.292 1.971 2.728 35.432 3.714 40.917
2 floors
0 2.477 8.303 1.357 7.500 1.171 7.215 45.216 9.671 52.725 13.104
150 1.857 5.921 1.132 5.436 0.943 5.297 39.041 8.191 49.219 10.539
300 1.516 3.835 1.025 3.589 0.903 3.656 32.388 6.415 40.435 4.668
450 1.742 2.329 1.320 2.218 1.194 2.278 24.225 4.766 31.458 2.190
550 2.329 1.742 2.218 1.32 2.278 1.194 4.766 24.225 2.190 31.458
700 3.835 1.516 3.589 1.025 3.656 0.903 6.415 32.388 4.668 40.435
850 5.921 1.857 5.436 1.132 5.297 0.943 8.191 39.041 10.539 49.219
1000 8.303 2.477 7.5 1.357 7.215 1.171 9.671 45.216 13.104 52.725
3.1 Simulating traffic types: results obtained

We created a simulation programme which uses the extensive database generated by the tests to
select the optimum energy-use configuration for each specific type of journey, taking the following
factors into account: travel distance, number of passengers on board and direction of travel. In each
case, right at the start of the journey, the control system recommends a specific energy use profile.
Some of the simulated traffic types are dealt with below. We decided not to apply the standard
traffic type definitions used in articles, studies and simulation software in the lift sector.
The sequences and the journeys defined in this chapter are achievable, thus they have been studied
and the results are shown. However, we haven’t dedicated time to use definitions and traffic
simulations that are defined and explained in very well-known books such as [12,13,14].
We are fully aware of them and a comparison of our differences in criteria may well be the subject
of a future study. But the results achieved by our simulator have been notably effective.
Note: As it is not the purpose of this paper, these examples does not show energy consumption
during idles and standby periods.
Example A: High-traffic lift in a metro station. 2 stops, 2000 journeys/day, 12.31 metres travel
distance, 1000 kg, 2:1, gearless PMSM, 50% counterweight:
Percentage of journeys depending on the car loading (example: 28% of the 2000 journeys are made
with 450 kg in the car): 0 kg: 2%, 150 kg:4 %, 300 kg: 10%, 450 kg: 28%, 550 kg: 28%, 700 kg:
18%, 850 kg: 6%, 1000 kg: 4%.
Power consumption per day: standard system: 12.09 kWh.

Power consumption per day: proposed dynamic control: 11.02 kWh 8.86% saving.
Example B: Low-traffic lift in residential building. 6 stops, 15.31 metres travel distance, 1000 Kg,
2:1, gearless PMSM, 50% counterweight:
We randomly generated the following sequence of journeys in a low-traffic context:
Lift on Floor 0. Empty Called to Floor 5 (empty). 4 passengers board. Called to Floor 2. 2
passengers board (6 now in car). Travels to Floor 0. Waits empty on Floor 0. 2 passengers
board to travel to Floor 5. Lift ascends to Floor 5. 2 passengers get out. Waits empty on Floor
5. 2 passengers board. Lift travels to Floor 4 where 2 passengers board (4 now in car). Lift
travels to Floor 0. Lift waits empty on Floor 0. 4 passengers board. 2 get out on Floor 4. 2
get out on Floor 5. End.
Table 3.Example B: consumptions and savings.
Traffic example B Magnets (standard DTF Dynamic

approach speed) [Wh] [Wh] solution [Wh]
Ascent from ground to 5th (empty) 3.336 2.260 1.971
Descent from 5th to 2nd (4 passengers) 8.585 8.579 8.367
Descent from 2nd to 0 (6 passengers) 3.510 3.413 3.291
Ascent from 0 to 5th (2 passengers) 2.739 1.981 1.562
Descent from 5th to 4th (2 passengers) 5.921 5.436 5.125
Descent from 4th to 0 (4 passengers) 11.268 11.075 10.809
Ascent from 0 to 4th (4 passengers) 2.156 1.668 1.476
Ascent from 4th to 5th (2 passengers) 1.857 1.132 0.902
TOTAL [Wh] 39.372 35.544 33.512
Energy saving (%) - 9.7 14.9
Example C: (includes results from tests using the regenerative drive system): Medium-traffic lift in
hospital, 5 stops, 1000 journeys/day, 12.31 metres travel distance, 1000 Kg, 2:1, gearless PMSM,
50% counterweight.
Journeys (ascents, 40% of total): 0 Kg: 15%, 150 Kg: 25%, 300 Kg: 25%, 450 Kg: 15%,
550 Kg:10%, 700 Kg: 7%, 850 Kg: 2%, 1000 Kg: 1%.
Journeys (descents, 60% of total): 0 Kg: 5%, 150 Kg:18%, 300 Kg: 24%, 450 Kg: 19%,
550 Kg: 15%, 700 Kg: 14%, 850 Kg: 3%, 1000 Kg: 2%.
Journeys 2 stops: 40%, Journeys 3 stops: 30%, Journeys 4 stops: 20%, Journeys 5 stops: 10%.
Daily power consumption–standard solution: 4.30 kWh/day
Daily power consumption–dynamic solution: 3.65 kWh/day - 14.82% saving
Daily power consumption-REGEN but without dynamic solution: 3.00 kWh/day - 30.04% saving
3.2 The regenerative drive system tested
The measurements with the FLUKE 435 II were taken at the regenerative unit power terminals. The
savings figures were reached by subtracting the energy generated from the energy consumed. The
lift's energy consumption when on standby and in motor mode was greater with this unit connected
to the lift. During the time the unit does not produce electrical energy - it gets constantly 40 W.
Also, in standby mode, it needs to get 10 W constantly. So, the tested lift passed from 33W to 43 W
in this mode.
The electric energy generated by the lift was fed into the building's power grid, which begs the
question: where exactly did that energy go? The answer to that is not at all clear. It is dependent on
various factors.
A small amount of the energy was probably consumed by the lift system itself - by the lift car
lighting, the control system and the brakes, for example. Whatever remained was fed into the
building's grid and, depending on the specific characteristics of its wiring network and of the
electrical devices connected to it at the time and their particular impedance, the energy may have
been consumed within the building or otherwise used outside it.
In terms of energy consumption within the building, depending on exactly where the electricity
meter is located, it is entirely possible that the electricity produced by the lift was actually charged
for by the power company, without taking into account that that energy was produced by the lift,
and not by the power company. Nowadays, it is very rare that a bi-directional meter is installed.
While ignoring for the moment the high costs of a regenerative solution, our proposed innovation
(the dynamic control) poses the question as to whether or not working on “CONSUMING LESS
ENERGY” makes more sense than consuming energy but “PRODUCING A SMALL AMOUNT
OF ENERGY” (without assuring how and who will use it).
3.3 Implications of compliance with ISO 25745-2

We decided to run comparative tests using the criteria laid out in ISO 25745-2, now widely
accepted in the lift industry. Its chapter 4, which deals with data collection and analysis tools,
explains how to use the reference cycle set out in ISO 25745-1. So the measurements were taken
with the car empty, taking the profile seen as the best energy-saving option from the suggested
dynamic control previously identified. Also, at no point did the travel speed exceed 1m/s - the
motor's rated speed.
In this paper we have shown energy savings achieved in all journey types. These results improves
the efficiency depending on the car loading and distance travelled. Given that ISO classification is
based on the reference cycle, here we can only show energy-saving improvements in one particular
instance - when the car is empty.As a result, we can only regard the data which appears in the
corresponding table as indicative, rather than definitive.
Table 4. Savings and improvements produced by the suggested solution are shown.Terms
defined by ISO 25745-2: Erd: daily running energy consumption [Wh], Enr: daily non running
(idle/standby) energy consumption [Wh], Ed: total daily energy consumption [Wh], Ey: annual
energy consumption [kWh].
Data inputs to the ISO calculator

Approac. 700 journeys, Cat. Use 4 110 journeys. Cat. Use 2
Speed DTF Dynamic
0,05 m/s Results
Approac. Approac.
Number
6 Speed DTF Dynamic Speed DTF Dynamic
stops 0,05 m/s 0,05 m/s
Journey
E. cycle
34.558 32.546 32.062 energy, 5852.8 5235.7 5085.9 1017.6 917.77 893.55
ref.
Erd[Wh]
E. short
17.327 15.376 14.902 Class 3 3 3 3 3 3
cycle
Stop energy,
Load 1000 [kg] 832.39 832.39 833.6 864.55 864.55 864.72
Enr[Wh]
Speed 1[ m/s] Class standby 1 1 1 1 1 1
Power Energy day,
50[W] 6685.1 6068.1 5919.5 1882.2 1782.3 1758.3
idle Ed[Wh]
Power 5 Energy year
33[W] 2440.1 2214.9 2160.6 687 650.5 641.8
min Ey,[kWh]
Power
33[W] Class C C B B B A
30 min
In both cases studied, and working in the same conditions, a lift could achieve a better energy
classification if the dynamic solution is implemented.
4 CONCLUSIONS
On the basis of the improvements in the lift's energy efficiency in each type of journey which we
have identified in this paper, we now propose the development of an intelligent control system
which can take effective energy-saving decisions (adjusting the acceleration and speed curve
profiles and motor control variables) immediately prior to start-up, while always keeping passenger
comfort and travel time as a clear priority.
The Direct To Floor curve is clearly a fundamental starting point, and real-time communication
between control system and inverter is demonstrably essential to the achievement of the purposed
goal.
With these two keys available (and the use of PMSM motors), the solution shown in this paper is
developed as a software programme that adds intelligence to the system in order to control in an
energy efficient way the speed curve of the journey as well as motor control variables. Thus, no
additional hardware components are required (taking into account a good capacity of the
microprocessors and big enough storage spaces in the electronic boards).
We raise concerns about the use of the electrical energy produced by a regenerative drive system.
We pose the question as to whether or not consuming less energy makes more sense than producing
a small amount of energy. In certain buildings, the installation of lifts which are genuinely efficient
in their energy use may well be a more interesting energy-saving measure than installing a
regenerative lift system.
It has become more and more common in various markets to see regenerative lifts with small cars
and a rated speed of 1m/s installed in buildings with low traffic intensity. Under such conditions, it
may well be the case that a lift as the one we propose would constitute a more interesting and
economic solution.
5 LITERATURE REFERENCES.
[1] Lutfi Al-Sharif, Richard Peters, Rory Smith. “Elevator Energy simulation model”, IAEE,
Elevator Technology 14, April 2004.
[2] Gina Barney “Energy efficiency of lifts – Measurement, conformance, modeling, prediction
and simulation”. CIBSE LIFTS GROUP. Archive. 2007.
[3] Lorente, A-M, Núñez, J.L, Barney G.C. “Energy models for lifts – Determination of average
car load, average travel distance and standby/running time ratios” .2nd Symposium on Lift and
Escalator Technologies.2012
[4] Richard Peters. “Green Lifts?”.Elevation Autumn 1995
[5] Gina Barney “Energy models for lifts”. 1st Symposium on Lift and Escalator Technologies.
2011
[6] Richard Peters “Green lift control strategies” International Journal of Elevator Engineers.
Volume 2. 1998
[7] Gina Barney “Energy efficiency of lifts – Measurement, conformance, modeling, prediction
and simulation”. CIBSE LIFTS GROUP. Archive. 2007.
[8] Richard Peters. “Lift Performance Time”.2nd Symposium on Lift and Escalator
Technologies. 2012
[9] CIBSE Guide D: Transportation Systems in Buildings 2015. Appendix A2.CIBSE 2015.
[10] Vicente Pacheco, Pilar Molina-Gaudo, Luis Jiménez, Estanis Oyarbide. “Improving the
Energy Efficiency of Lifts”. 5th Symposium of Lift and Escalator Technologies. 2015.
[11] NC-Drive software application. http://www.vacon.com/es-ES/servicio/Software-
tools/Configure-and-monitor-drives/NCDrive/
[12] Strakosch, “The vertical transportation handbook”, John Wiley &Sons, 1998
[13] Gina Barney, “Elevator traffic handbook. Theory and practice”, Routledge 2002.
[14] CIBSE Guide D: Transportation Systems in Buildings 2015.CIBSE 2015.
Mr. Vicente Pacheco de las Cuevas, Ms.Sc. in Physics (Specialising in Electronics) (2000),
University of Cantabria (Spain) started work in IMEM Lifts in the year 2000. Nowadays, he
manages the Electrical, Electronic and Automation area of the R&D Dept.
Fire Lifts, Escalators & Moving Walks Management System (FEMS) in an
Airport
Giovanni Pappalardo
ANACAM LIFT ASSOCIATION – Via Emilia, 47 – Rome – Italy
email: [email protected]
Keywords: FEMS system (Fire Lifts Escalators & Moving Walks Management System), EN 81 –
73: 2005, fire detectors, fire central unit, airport terminal, open hardware architecture, open
software architecture, false fire alarms
Abstract. In December 2013 the management of the International Airports of Rome (ADR) decided
to assign me to project a fire safety system similar to EN 81-73 : 2005[1], for the lifts installed in
Leonardo da Vinci Airport of Rome Fiumicino. The specific requirement of the Direction of the
airport was to implement a fully automatic system, with no human supervision, to prevent
passengers in a lift to be stranded on a floor, or trapped in a lift, where a fire has broken out.
The Direction of ADR added some further specific project requirements:
- The hardware and software of the system was to have “open” architecture
- The system had to be reliable and with a high level of safety;
- Reduce to a minimum the probability of false alarms and safeguard the capability of the
system to start with a real fire alarm.
The Satellite terminal G gates were chosen to start the project. Escalators and moving walks
(EMWs) have also been included in the system. To reduce the probability of false alarms a 3D
simulation model of a map of fire sensors, based on about 1400 sensors in the Terminal and
statistical data coming from about 18,000 fire sensors, has been prepared and verified.
The requirements of open hardware (PLCs (Programmable Logic Computers) and electronic
equipment) and open software (PLC and SCADA (Supervisory Control And Data Acquisition)
software) have been fully accomplished and the system is now operative in the Terminal G gates.
ADR management has recently decided to extend the system to the rest of Leonardo da Vinci
Airport.
The FEMS project has been recently approved for lifts, escalators and moving walks in public
transport by the Italian Ministero delle Infrastrutture e Trasporti (Italian Infrastructures and
Transports Authority)
1 INTRODUCTION
Everything started when an external safety audit, conducted at the International Airport of
Fiumicino, required a fire safety system that was capable of avoiding an accident similar to that one
occurred at Düsseldorf Airport, Germany, in 1996 [2] to be applied to the lifts.
On Thursday 11th April 1996 a massive fire spread in a Terminal of the International Airport of
Düsseldorf, Germany, with 17 casualties, 72 injured and hundreds of people with light wounds or
symptoms of smoke intoxication.
Seven people died from toxic smoke inhalation in two lifts when they decided to escape from the
fire by using the lifts and landed at the ground floor, where the fire had spread.
After this massive fire event several countermeasures were taken to avoid further future similar
accidents and the EN 81-73: 2005 harmonised standard was issued to indicate the behaviour of lift
units in case of a fire.
The procedure of FEMS for lifts is almost fully conforming to EN 81-73: 2005 and, as a brief
reminder, the procedure is as follows:
- if the lift is at a floor with a fire, or it is at a different floor by a safe designated floor (to be
specified later), its doors must close and it must reach the designated floor. Once it has
arrived at the safe designated floor it opens the doors and stops;
- if the lift is already at the safe designated floor it opens the doors and stops;
- if the lift is travelling to a floor with a fire, or travelling to a different floor by the safe
designated floor, it must stop at the first available floor along the path, not open the doors
and then travel to the designated floor where it opens the doors and stops.
In all the cases, if the fire alarm has started, the lift does not respond to any call, internal or external,
and once stopped at the designated floor it requires a manual reset, by a technician, to resume
normal service.
In addition the photocells, or light barriers, that could be affected by smoke are excluded, without
excluding the safety contact of the doors or the protective device of the door operator to re-open
doors in case of an obstacle.
Practically the only real differences between EN 81-73 : 2005 and the FEMS system are: a special
procedure of manual reset, to be performed directly on the controller of the unit by an authorised
person, that it is not foreseen in the harmonised standard, and the extension of the system to
escalators and moving walks.
The FEMS procedure for escalators and moving walks has been simply defined as follows:
- if the direction of motion of the escalator (moving walk) is in the same direction of the
evacuation, in case of a fire, the escalator (moving walk) will continue to work;
- if the direction of motion of the escalator (moving walk) is in the opposite direction of
the evacuation, in case of a fire, the escalator (moving walk) will stop.
This could be useful, for example, in case of an underground passage with two parallel escalators
running respectively upward and downward. In case of a fire in the lower area of the underground
passage the escalator running upward will continue to work while the other escalator will stop,
increasing the evacuation route.
The possibility of reversing the direction of motion of EMWs in case of a fire has been carefully
examined, but the procedure of an automatic restart in the opposite direction of motion could lead to
a potentially dangerous situation for people not conforming to point 1.2.3. of 2006/42/EC Machine
Directive.
The safe designated floor is a variable floor that is determined automatically by the software in
function of floors with a fire alarm.
A priority list of designated floors has been prepared with the cooperation and under the direct
control of the Fire and Security Department of the Airport of Rome Fiumicino. Each floor has been
assigned a ranking based on a priority list to determine the order of evacuation in case of a fire.
In case of a fire alarm, the FEMS software verifies if the first floor of the priority list is threatened
by a fire.
Fire Lifts, Escalators & Moving Walks Management System (FEMS) in an Airport 17-3
If the first floor of the priority list is free from fire the software automatically assigns it the status of
the safe designated floor; otherwise the software verifies the status of the next floor in the ranking,
with the same decision procedure.
The procedure stops when the software finds the first floor in the ranking of the priority list that is
free of fire and assigns it the status of designated floor.
I have to underline that the above reported procedure could lead to a priority list where the safest
floor is not necessarily the lowest floor.
For example in a multi-level car park, in some particular cases, the safe designated floor could be
the top level floor where we have favourable conditions such as an open-air space, with reduced
danger of smoke intoxication, and fire-proof evacuation stairs leading to safe areas outside of the
building.
Note: for security reasons some information and constructive details, concerning the airport and
equipment installed in the airport of Rome Fiumicino, cannot be disclosed.
2 PROJECT FIRST PHASE

To start this new project the Terminal G Gates (also called Satellite) were chosen for the following
simple reasons.
1) The terminal is the newest at the Airport, with a relatively simple architecture and only four
levels, defined by altitude above sea level, consisting of one underground level and three different
floors as described in table 1.
Table 1 Description of levels of Terminal G Gates
Level Altitude above Note on the area

sea level [m]
Lower - 1.90 Restricted area – no landing of lifts or escalators
Airfield + 1.80 Restricted area to passengers – machine and power rooms
Intermediate + 6.50 Passengers arrival area – shuttle station to Terminal 3
Top + 11.00 Departure gates area, duty free shops, shopping centres,
restaurants, coffee bars, etc.
2) The fire detection system of the terminal is controlled by one fire central unit (FC 2080 Siemens)
connected to about 1400 fire detectors and reporting alarms, failures, etc. to a remote control room,
supervised 24 hours a day by an emergency team.
All the fire central units of Fiumicino Airport, included the terminal G gates fire central unit, are
connected to a DESIGO™ INSIGHT fire supervision software system provided by Siemens.
In the next two tables, 2 and 3, are details about the distribution and typology of fire detectors.
Table 2 Distribution of fire detectors in Terminal G Gates
Level Altitude above sea level Total number of fire

[m] detectors per level
Lower - 1.90 73
Airfield + 1.80 391
Intermediate + 6.50 548
Top + 11.00 388
Total number of fire detectors 1400
Table 3 Typology of fire detectors in Terminal G Gates
Fire detector code Number of fire detectors Typology

per typology
FDO241 1024 Smoke detector
FDM223 138 Alarm button (type A)
FDT214 40 Thermal detector
FDO221 17 Smoke detector
FDCIO221 112 Control module
FDM221 1 Alarm button (type B)
FDOOT241/8 64 Smoke / Thermal detector
FDOOT241/0 4 Smoke / Thermal detector
Total number of fire detectors 1400
3) There are only 28 hydraulic lifts and 17 escalators, for a total number of 45 units mainly installed
in the terminal by Schindler (in the other airport terminals there are about 270 units), with the
following status:
- 12 lifts installed by Schindler with Elettroquadri controller currently connected to
FEMS;
- 14 lifts installed by Schindler with Hydroware controller currently connected to FEMS;
- 2 lifts, currently not connected to FEMS, installed by Schindler with controllers to be
replaced before the end of 2016 by Elettroquadri controllers;
- 17 escalators (16 Schindler / 1 Paravia), 30° inclination, about 4.50 m rise, connected to
FEMS;
for a total of 26 lifts and 17 escalators currently connected to FEMS.
3 OPEN HARDWARE AND SOFTWARE ARCHITECTURE

To accomplish the management requirement, of an open hardware and software architecture, the
status of the fire detectors (i.e. : failure, out of order, pre alarm, alarm, etc..) is received directly
from the fire central unit with a LAN port, through an Ethernet cable, by-passing DESIGO™
INSIGHT system, based on the Windows™ NT operating system.
A Siemens NK 8237 [3] gateway has been interposed between the fire central unit and the FEMS
system to translate the BACnet [4] Siemens protocol of the fire central unit to a standard protocol
(in our case MODBUS™ TCP/IP [5, 6]).
The gateway also has the function of a safety firewall against uncontrolled revisions of the
configuration of the fire central unit.
With this system it has been possible to receive the status of the fire detectors, in a standard
protocol, that can be read by industrial PLCs. In the following table (table 4) it is possible to see an
example of some fields contained in one MODBUS™ register of the file, corresponding to a fire
detector (Loop 16 / Element 44 / Smoke detector) installed at 6.50m above sea level, translated by
the gateway in a CSV (Comma Separated Value) file.
Table 4 Example of a MODBUS™ register of a fire detector
MODBUS™ MODBUS MODBUS™ MODBUS

NodeId SlaveAddress ParentDescription ObjectName ™Table BaseAddress ™Address
8373022 2 L16 E44 6D77 FIDEDEGE96453406 WT_LogCh 3000 1398
The MODBUS™ Table field contains a word of 16 bits (0 – 15 bit): WT_LogCh, where if the
digital value of bit number 15 is 1, the fire detector is in a fire alarm status.
The master PLC of the FEMS system receives the alarm status from the fire central unit and starts a
software procedure that will be detailed later, to determine if it is a real alarm or a false alarm.
If the software procedure determines that we have a real fire alarm it assigns, based on the priority
list of the evacuation floors, the safe designated floor and starts the procedure for lifts and escalators
that has been described in the previous paragraph.
The master PLC, a model of the Schneider Electric TSXH57XX family, has been programmed with
a standard language Unity Pro™ [7] compliant to IEC 61131-3 standard, and connected to lifts and
escalators through seven Input / Output modules (I/O) installed in seven different machine rooms at
the 1.80 m airfield level.
Each I/O module controls a different cluster of about 6 – 7 units (lifts and escalators) in order to
divide the area of the Terminal into seven smaller areas to control all the 26 lifts and 17 escalators.
The controller of each lift, or escalator, is connected to a FEMS customized module (that we define
as the FEMS interface) that works as an interface between the unit and the I/O module.
The FEMS interface has been made by Elettroquadri, Hydroware and Schindler on my technical
specification and can be reproduced for each unit, (lift, escalator or moving walk) if the complete
and updated wiring diagram is available.
In the next figure (Figure 1) it is possible to see an example of a FEMS interface module made by
Hydroware for 14 lifts installed in the Terminal.
Figure 1 FEMS interface for Hydroware controllers
The master PLC receives all the signals concerning the status of the unit: moving up, moving down,
doors open, door closed, failures, etc., from a specified I/O module through the FEMS interface and
sends the commands to the unit, in case of a fire alarm, to the I/O module and then to the FEMS
interface, with a two way data communication flow.
Furthermore, the master PLC is reporting all the data (alarms, failures, status of the units) to a
SCADA system with software provided by Wonderware Inc. by Schneider Electric.
4 HIGH LEVEL OF SAFETY AND RELIABILITY

The EN 81-73 : 2005 does not require a specified level of safety and reliability but, anyway, the
FEMS has been projected and designed with redundancy and reliable components in order to be
prepared to meet, for other future projects in the Fiumicino Airport, at least a SIL (Safety Integrated
Level ) 2 [8] level1 in some sub-systems.
The current architecture of the FEMS system is:
- two MODBUS™ gateways (master and slave) with the slave gateway in hot backup;
- two industrial PLCs (master and slave in hot backup in order to have a HFT (Hardware
Failure Tolerance) ≥ 2), installed in two machine rooms at about 90 m of distance,
connected with fiber optic cable, with the following performance levels [9]:
- MTTFd ( Mean Time to Dangerous Failure ) ≥ 30 years working;
- PFHd < 10-7 ( Probability of dangerous failure per hour)
- EN ISO Performance Level e grade (PLe) < 10-7;
1
SIL 2 level is equivalent to probability of failure in the next hour (PFH) < 10-6
- seven I/O modules installed in seven different machine rooms, in order to have a
distributed logic network, connected in a closed loop. The reason of a distributed logic
network in closed loop is to reduce the loss of control on the units if one I/O module, or
the connection cable, is damaged by a fire or an accident;
- command to stop the unit given with a double contact command (for the lift the
command is given only if the PLC receives the confirmation signal, from the unit, that
the lift has reached the safe designated floor and has opened the doors);
- a firmware for PLCs developed for industrial environment certified IEC 61131;
- a standard protocol ( MODBUS™ TCP/IP in this case ) with data coming directly from
fire unit, bypassing operating systems Windows NT based;
- two servers for data recording, connected in hot backup, installed in two different data
centers at a distance of about 5.0 km, for a crash recovery;
- a UPS (Uninterruptible Power System) installed in a machine room with a minimum of
60 minutes of certified power supply, to FEMS apparatus, in case of a black-out.
In addition there is already a project to improve the reliability of the FEMS interface with lifts or
escalators with redundant electronic relays.
5 REDUCE FALSE ALARMS

There are about 18,000 fire sensors (thermal, smoke, alarm button, thermal/smoke) installed in
Fiumicino Airport in terminals, office buildings and airport facilities.
Our records of about 3 years (period 2011 – 2014) report an average of about 200 false alarms per
year, with about 1 false alarm every two days.
Unfortunately we do not have detailed failure analysis to separate data due to internal causes
(failures) or external causes (for example: smoke from kitchens, cigarettes, etc.). Therefore the
statistics is (1):
MTBF (parent population) = 365 / 200 = 1.825 days. (1)
From this we can assume, in first approximation, the average life time of a sensor with the
following formula (2):
Average life time = (50% Parent population)/false alarms per year = 9000/200 = 45 years. (2)
The above number is in accordance with an average life time of about 50 years of a fire detector
(life time derived from the technical data sheets available on the web sites of the main
manufacturers of fire sensors).
For Terminal G gates population of fire sensors we can estimate the average number of false alarms
with formula (3):
Number of false alarms / years = 700 / 45 = 15.56 false alarms per year. (3)
Hence the probability of a failure in the next hour (PFH) for a fire sensor can be computed with the
following formula (4):
PFH = 1 / (45 x 365 x 8) = 2.54 x 10-6. (4)
The estimate is in accordance with the statistics of the last 4 years with an average of about 12 false
fire alarms per year in Terminal G gates.
This means that if we had set the FEMS system to start with only one sensor, in alarm condition, we
should expect at least one false alarm per month, and this is absolutely unacceptable for an airport
management.
For this reason a matrix, based on the database of fire sensors of the Terminal G gates, has been
prepared with the following fields (table 5):
Table 5 Example of records of Terminal G gates fire sensors matrix
ParentDescription ObjectName MODBUS™Address X [m] Y [m] z [m] Typology

L16 E44 6D77 FIDEDEGE96453406 1398 113,26 157,01 6,50 Smoke
L14 E5 2057 FIDEDEGE96454793 1230 82,54 24,09 1,80 Smoke / Thermal
.... ….. …. … … … …
Furthermore a weight (wi) for each typology of fire detector has been determined in function of the
criticality of the detector as reported in the following table (table 6).
Table 6 Criticality weights of fire sensors
Fire sensor Weight (wi)

Smoke 0,3
Thermal 0,5
Button 0,2
Smoke / Thermal 0,8
With a computer simulation a certain number of sensors at the same time have been randomly
shifted in alarm state and the software has computed the number of times when a determined
number (x) of sensors were in alarm inside a circle of a pre-set radius R.
After about 107 simulations the following rule, for PLC programming, has been determined to start
a FEMS procedure:
IF ∑i (wi) > 1.00 (threshold value) of fire sensors in alarm status

AND all sensors are inside a 10.00 m radius circle
AND all sensors are at the same z coordinate (same level)
THEN start FEMS procedure
With this rule the probability of a false alarm in the next hour has been computed to be less than
10-15 for the distribution of fire sensors of Terminal G gates.
The FEMS system has been recording data from July 2015 and so far no false alarms have been
reported.
Furthermore data recorded from the system are used to compute Key Performance Indicators (KPI)
of maintenance such as MTBF, MTTR, MTTA, Machine Availability, etc. of the lifts and
escalators.
REFERENCES
[1] EN 81 – 73 – Safety rules for the construction and installation of lifts - Particular
applications for passengers and goods passenger lifts - Part 73: Behaviour of lifts in the event of fire
[2] Fire Investigation Summary – Dusseldorf, Germany – April 11th, 1996 – National Fire
Protection Association (NFPA)
[3] NK8237 Technical Data Sheet - MP4.60 MODBUS™ Gateway for Sinteso(TM) and
Cerberus(R) PRO Fire Detection Systems
[4] ASHRAE STANDARD SSPC 135 – BACnet A Data Communication Protocol for Building
Automation and Control Network – Approved by the ANSI (American National Standard Institute)
on February 25, 2004
[5] MODBUS™ - Modicon MODBUS™ Protocol Reference Guide - PI–MBUS–300 Rev. J
[6] TCP/IP – Transmission Control Protocol – DARPA Internet Protocol – September 1981
[7] Unity Pro – Program Languages and Structure Reference Manual – 12/2015 – by Schneider
Electric
[8] IEC EN 61508 – Functional safety of electrical / electronic / programmable electronic safety
related systems
[9] ISO EN 13849-1 – Safety of machinery – Safety related parts of control systems – Part 1
General principles for design
Giovanni Pappalardo has graduated in Mechanical Engineering in Italy and post-graduated in
University of Milano in Production. He has worked for Aerospace and Electronic Industry in Italy
before entering Otis Italy in 1991 as Quality Product Manager. Since 2000 he has worked as a
project engineer and consultant for Italian Real Estate companies. He is also an external consultant
for the Italian National Committee for Maintenance (CNIM), for ANACAM (Italian Association of
Lift Companies) and for Aeroporti di Roma S.p.a. (ADR).
Global Dispatcher Interface
Richard Peters
Peters Research Ltd,Bridge House, Station Approach, Great Missenden, Bucks HP16 9AZ, UK
[email protected]
School of Science and Technology, The University of Northampton, UK
Keywords: dispatcher, group traffic control, interface, conventional control, destination control
Abstract. The efficiency of a lift group depends heavily on its dispatcher (also known as the group
traffic control). A dispatcher decides how a group of lifts serve the passenger demand, normally
based on calls placed on the system by the passengers. Defining a common, global dispatcher
interface makes it easier for simulation and real word systems to talk to each other. The author
draws on practical experience to consider if the next generation of dispatchers should be centralized
or decentralised, and to suggest a dividing line between lift controller and lift dispatcher functions.
Having addressed dispatcher architecture and scope, the requirements of a global dispatcher
interface are considered. These include, but are not limited to single deck cars, double deck cars,
and multiple independent cars in a shaft. The dispatcher interface also needs to consider different
user interface options including landing call buttons, car call buttons, destination based input,
together with associated indicators and displays.
1 INTRODUCTION
The early group traffic control systems were human dispatchers who stood in the main lobby during
the morning uppeak directing passengers and in-car attendants who controlled individual cars.
Humans gave way to systems utilising relay logic, which in turn gave way to hybrid relay/electronic
controllers, programmable logic controllers and microprocessor based systems. Many modern
dispatchers apply artificial intelligence, mimicking the intelligence and insight of the original
human dispatchers [1].
The responsibilities of the dispatcher and controllers in most current systems mimic the division of
labour between the early human dispatchers and in-car attendants. The dispatcher allocates the
landing or destination calls to individual cars. The controller (in-car attendant) dictates how the
allocation landing calls and car calls are served, also managing door functions.
A dispatcher needs to communicate with passengers, typically through buttons and indicators, and
with the lifts to allocate calls. A proprietary dispatcher interface has been available since 1998 [2].
It has been widely applied in simulation, and with modifications in actual installations. However,
the interface was built for simulation without consideration of real time systems and has evolved to
support new technology rather than been designed for it.
This paper suggests moving the dividing line between dispatcher and controller functions. It
proposes a second generation global dispatcher interface as a basis for developing dispatchers
which are easily interchangeable and that apply exactly the same code in simulation and real world
systems.
2 SYSTEM ARCHITECTURE
2.1 Centralised and distributed control
There is a vast amount of data to exchange within a modern lift installation and a number of
possible methods by which the data can be collected and processed [3].
A common approach is distributed control, see Figure 1. Each lift controller receives all
information about new landing (or destination) calls over a network. Each lift controller performs
its own calculations, providing a bid for the call according to the traffic control algorithm. The
master lift control compares the bids and awards the call.
Landing and/or
destination calls
Lift Lift Lift Lift

controller controller controller controller
Figure 1 Distributed control (from Figure 9.6 CIBSE Guide D: 2015 [3])
Alternatively, a dedicated group controller, sometimes an industrial computer, collects the data and
allocates landing calls to a lift according to the traffic control algorithm, see Figure 2. As with
distributed control, once a lift has been allocated a call, it is normally the lift controller that
manages how calls are served.
Landing and/or Group

destination calls controller
Lift Lift Lift Lift

controller controller controller controller
Figure 2 Centralised control (from Figure 9.5 CIBSE Guide D: 2015 [3])
With distributed control, if the master lift fails, one of the other lifts automatically takes over the
group control functions. With centralised control, a backup group controller is often included in
case of failure. Both approaches can work. In new installations, distributed control tends to be
favoured. In modernisation, centralised control allows for the dispatcher to be upgraded while
keeping the existing controllers. Centralised control can also allow for a range of new controllers
from different sources to use the same dispatcher.
2.2 Controller design issues impacting the dispatcher
For the dispatcher designer, centralised control is simpler and more flexible. However, there are
other controller design issues which impact dispatcher optimisation.
Once a lift has been allocated a call, it is normally the lift controller that manages how its calls are
served. To provide good performance, it necessary for the dispatcher to make assumptions as to
how the lift controller is going behave once a call is allocated. Although collective control is a
prerequisite for almost all modern lift groups, when the fine details are considered, its
implementation varies.
For example, consider the scenario given in Figure 3. Lift A is travelling to pick up a down landing
call at level 7. Once lift A arrives at level 7 it will reverse direction. As it travels towards level 7,
before it reaches its slow down point, a new up call is registered at level 8. The dispatcher
determines the best solution is for lift A to change its target floor from 7 to 8 so that it can serve the
Global Dispatcher Interface 18-3
up call at level 8 (and its subsequent car call) before reversing and serving the down landing call at
level 7.
At what point is it too late to change lift A’s target floor from level 7 to level 8? Some lift
controllers will commit to the reversal at level 7 as soon as it becomes the target floor. Other
controllers may allow the change, but the point at which the target is fixed is not consistent.
In the context of conventional control, a mistake can be corrected; if the landing call at level 8 is
allocated to lift A, but the lift reverses at level 7, the level 8 call can be re-allocated to lift B.
However, with a destination control system, the passenger has been told to wait for lift A, and must
wait for a complete round trip of the lift before his or her call is answered. Other scenarios yield
similar issues. Without a time-consuming dispatcher design and test process involving analysis of
dispatcher logs post installation, some dispatcher errors only manifest themselves in long waiting
times. Sometimes these errors go uncorrected for the lifetime of the installation.
new
Level
Level 7 A
Figure 3 Scenario leading to potential dispatching error

In other scenarios, the dispatcher designer may want the collective control rules followed by most
lift controllers to be broken, for example where a reverse journey (initially taking a passenger in the
opposite direction to their destination) is the best compromise option in a destination control system
[4].
2.3 Intelligent door control

Intelligent door control provides one of the best opportunities to improve current dispatcher design.
For example, destination control systems do not take advantage of the information they have about
the number of passengers expected to alight and load a lift when considering when to start closing
the doors. This can be observed if travelling to a floor when it is known that only one passenger is
alighting. In this instance the dispatcher could tell the lift doors to start closing immediately that the
door beams are re-established. In all systems observed, the normal lift controller dwell times are
left to expire before the doors close, often wasting between two and four seconds.
2.4 New requirements
In the case of two or more independent cars in a single shaft, dispatching become increasingly
complex with the choices made for the operation of each car being impacted by the status and
position of all the other cars. For example, with three cars (A, B and C) in the same shaft, if a
dispatching solution involves car A being held because car B is being held to avoid collision with
car C, the ability of car A to make optimum dispatching decisions becomes increasing difficult. Car
A needs a lot of information to make an optimum dispatching bid for a call; a centralised solution
with a car A simply accepting travel commands will require significantly less network traffic.
2.5 Second generation global dispatcher interface
The global dispatcher interface proposed in the following sections assumes centralised control.
Furthermore, operation logic commonly managed by the lift controller is assigned to the dispatcher.
The lift does not manage calls, it goes where it is told to, and accepts door open and close
commands directly from the dispatcher. Yielding the minimum decision making to the lift
controller and other devices minimises the opportunity for the systems to conflict and for
inconsistency in dispatcher performance with different controllers. This allows the dispatcher
designer the best opportunity to optimise performance and maximise portability between lift
controllers. It also simplifies the task of implementing a lift controller and the components needed
to create user interfaces.
3 THE INTERFACE
The interface allows for both conventional and destination calls within the same lift group. It can
accept destination and conventional calls or operate as a hybrid accepting both conventional and
destination calls.
For brevity, this paper describes the interface for single deck lifts only. Future publications will
account in more detail for double deck lifts, multiple cars operating in the same shaft, and for
movement in three dimensions.
The paper describes the open loop version of the dispatcher; closed loop options will be described
in future publications. Close loop operation allows the dispatcher to confirm messages have been
received successfully and to test if theoretical possibilities are practically achievable, e.g. can a lift
stop in time for a new call placed while the lift is travelling.
The interface could be implemented with different mechanisms. Proof of concept tests have been
completed using messages communicated over TCP/IP applying Protocol buffers. This is a
language-neutral, platform-neutral, extensible mechanism for serializing structured data [5]. For
memory and speed, all variables are integers, hence the use of grams rather than kilograms and
millimetres rather than meters.
To reduce network traffic, data is divided into static and dynamic data. Static data does not change
and only needs to be communicated during initialisation.
4 STATIC DATA
A summary of the data required is given in Table 1. MAX_LIFTS and MAX_FLOORS define the
limit of number of floors and number of lifts that the dispatcher will manage. MAX_RISERS
corresponds to the number of destination input device risers; destination input devices with the
same riser number are in the same [x, y] position on different floors.
Table 1 Static data
Variable Description
Velocity[MAX_LIFTS] Rated lift velocity (mm/s)
Acceleration [MAX_LIFTS] Rated lift acceleration (mm/s/s)
Jerk[MAX_LIFTS] Rated lift jerk (mm/s/s/s)
DoorPreOpen[MAX_LIFTS] Door pre-opening (ms)

DoorOpen[MAX_LIFTS] Door open time (ms)
DoorClose[MAX_LIFTS] Door closing time (ms)
MotorStartDelay[MAX_LIFTS] Motor start delay (ms)
LevellingDelay[MAX_LIFTS] Levelling delay (ms)
Home[MAX_LIFTS] Default parking floor (normally ground).
Refers to lower deck if lift has more than
one deck.
Capacity[MAX_LIFTS] Nominal lift capacity (g)
FloorArea[MAX_LIFTS] Floor area of car (mm²)
NoFloors Number of floors served
NoLifts Number of lifts
FloorPositions[MAX_FLOORS] Positions of floors (mm above reference)
FrontDoors[MAX_FLOORS] To indicate if front doors on this landing
RearDoors[MAX_FLOORS] To indicate if rear doors on this landing
NoDestinationInputRisers Number of destination input riser
positions
WalkingDistance[MAX_LIFTS] Walking distance from destination input
[MAX_RISERS] risers to lifts (mm)
WalkingSpeed Passenger walking speed (mm/s)
PassengerLoadingTime Passenger loading time (ms/passenger)
PassengerUnloadingTime Passenger unloading time (ms/passenger)
MinPhotocellDelay Minimum photocell delay (ms)
MinDwellCarCall Minimum dwell time for car call (ms)
MinDwellLandingCall Minimum dwell time for landing call (ms)
DestinationIndicators[MAX_FLOORS] 1 or 0 to indicate if messages are required
for destination indicators on each floor
DirectionIndicators[MAX_FLOORS] 1 or 0 to indicate if messages are required
for directional indicators are on each floor
CarIndicators[MAX_ LIFTS] 1 or 0 to indicate if messages are required
for destination indicators on each floor
5 DYNAMIC DATA
5.1 Adding a call to the dispatcher
Calls can be added to the dispatcher with messages according to Table 2.
Table 2 Call message
Index Unique index created by system placing call
CallType Car call, up landing call, down landing call, destination call
Origin Floor index of origin of call (required for landing calls and
destination calls)
Destination Floor index of destination of call (required for destination
calls and car calls)
OriginSide Front or rear
DestinationSide Front or rear
Riser Destination input riser position (required for destination
calls)
ExclusiveGroup Exclusive group index where groups of passengers are to be
separated (optional and only available for destination calls)
PersonID Person ID (optional and only available for destination
calls). This will be obtained from a card reader or similar
on the destination input device.
Special functions to be added
The dispatcher should respond with a message in the following format, see Table 3.
Table 3 Call message response
Index Index provided in call message
Allocation Allocated car (zero for no allocation)
Error code Error code to indicate why no allocation made
5.2 Security
For systems requiring the dispatcher to manage security, the dispatcher needs to know if a person is
allowed to travel to the floor requested. In this instance the dispatcher will send a message to the
security system to request authorisation, see Table 4.
Table 4 Security message
PersonID ID of person requesting call
Destination Requested destination
The response from the dispatcher is described by Table 5.
Table 5 Security call message
PersonID ID of person requesting call
Authorisation 1 or 0 for true or false
Note that there are there are other ways to manage security. For example, on presentation of the
security card (or other ID device), the destination input device may only present the floor available
to that person. In this instance, the dispatcher does not need to be involved in the management of
security.
5.3 Lift status

Lift status messages are sent to the dispatcher on initialisation and subsequently when any variable
changes. Messages may contain one or more variable updates.
Table 6 Lift status messages
LiftNo Lift car this message corresponds to
CarService Indicating if lift is in automatic, manual or out of service
CurrentLoad Current car load (g). During loading and unloading of the
car, there should be damping of this variable. When the lift
is moving, no status updates for this variable are required.
CurrentFloorNo Current floor number, updated when the lift reaches its
destination. Intermediate floor numbers are not required
and will be interpolated from lift dynamics.
DoorBeams Flag indicating if door beams are interrupted
DoorBeamsRear Flag indicating if rear door beams are interrupted
TravelStatus Flag to indicate if lift is travelling
DoorStatus Current status of the front doors (1 fully open, 2 closing, 3
fully closed, 4 opening, 5 nudging)
DoorStatusRear Current status of the rear doors (1 fully open, 2 closing, 3
fully closed, 4 opening, 5 nudging)
Table 7 Lift command messages
CloseDoors Close lift doors
NudgeDoors Close the lift doors applying nudging operation
SetDestinationFloor Start journey to floor when doors have closed.
OpenDoors Open doors on arrival, or immediately if lift is not
travelling. If doors are closing, reverse.
5.4 Lift dynamic configuration

The purpose of the messages in Table 8 are to allow lifts to be locked off from selected floors. This
is sometimes required if a floor is unoccupied.
Table 8 Configuration message
FloorNo Index of floor this message corresponds to
FrontLocks 1 or 0 to allow front lift doors to open on this floor
RearLocks 1 or 0 to allow rear lift doors to open on this floor
5.5 Indicator status messages

The following messages in Table 9 are to address indicators on the landings and in the cars.
Table 9 Status messages
FloorNo Floor this indicator is on
Direction 1 or 0 corresponding to up or down
DestinationFloor Destination floor to add or remove
Status 1 or 0 corresponding to on or off
6 CONCLUSIONS
To design reliable and portable dispatchers, a well-defined dispatcher interface is required. This
should work with both simulation and in real systems.
This paper gives a high level description of a global dispatcher interface proposed for the next
generation of dispatchers. The author invites constructive comments and suggestions to help
improve and develop what it is anticipated will become a de facto standard. A more detailed
specification and examples will be available to those who wish to contribute.
REFERENCES
[1] G. Barney and L. Al Sharif, Elevator Traffic Handbook: Theory and Practice, Routledge, 2015.
[2] Peters Research Ltd, “Elevate,” [Online]. Available: https://www.peters-

research.com/index.php/elevate. [Accessed 1 July 2016].
[3] R. Peters, “Lift traffic control,” in CIBSE Guide D:2015 Transportation systems in Buildings,
The Chartered Institution of Building Services Engineers, 2015.
[4] S. Gerstenmeyer and R. Peters, “Reverse Journeys and Destination Control,” in 4th Symposium
on Lift & Escalator Technologies, Northampton, 2014.
[5] Google, “Protocol Buffers,” [Online]. Available: https://developers.google.com/protocol-

buffers/. [Accessed 1 July 2016].
ACKNOWLEDGEMENTS
The author is grateful to Dr Albert So who first persuaded him to implement a dispatcher interface
in Elevate, Dr Mike Pentney for designing the original Elevate Windows DLL interface, Dr
Jonathan Beebe for his informative work with open standard information models, and for Mr Jim
Nickerson for his expertise in developing the Elevate interface in a real time environment with
modern software technology.
Map-Based Active Compensation of Lateral Vibrations in Elevators
Raúl Monge1, Javier Saumell2 and José Manuel Rodríguez3
1
Instituto Tecnológico de Aragón (ITAINNOVA), Zaragoza, (Spain), [email protected]
2
3
Keywords: active vibration control, mapping, guide rails, rolling guides.
Abstract. Lateral vibration in elevators has an important effect in the comfort levels perceived by
the passengers. This phenomenon is highly affected by the geometry of the guide rails and the load
distribution of the car. In this connection, irregularities in the former behave as perturbations that
excite the oscillation of the vehicle. The effect is more and more important as the speed of the
elevators increase, which is the current trend in the industry. In order to improve the performance of
medium and high speed elevators, the present paper describes a method for compensating the lateral
oscillations appearing in an elevator due to the irregularities of the guide rails. The proposed
approach makes use of a mapping algorithm developed by the authors for identifying, learning and
efficiently storing the geometrical configuration of the rails as a combination of straight line
segments. The system is conceived for active roller guides, whose position can be continuously
controlled in order to dampen the oscillations of the vehicle and to compensate the perturbations
caused by the geometry of the guide rails.
In order to develop the system and validate its performance, a 2D virtual environment in Matlab
Simulink © is used. This environment includes the geometry of the guides and the main elements of
the elevator affecting the horizontal oscillation: inertial parameters (mass, inertia), stiffness of the
roller guides, among others. The present analysis does not take into account the oscillations caused
by the traction rope or the movements of the load inside the cabin.
The results of the proposed method show the improvement that can be obtained in the ride quality
of the elevator by mapping the geometry of the guide rails and properly using this information for
compensating the identified irregularities by active roller guides.
1 INTRODUCTION
Lateral vibration is an important source of discomfort in lifts. As described in [1] and [2], these
oscillations are mainly caused by low frequency oscillations from the suspension cables or
asymmetric load placing; high frequency oscillations due to the guide rails, aerodynamic turbulence
around the car or the movement of the passengers inside the cabin. The present work focuses on
reducing the effect of irregularities in the geometry of the guide rails. To do that, a combination of
three technologies is proposed: perturbation observers for recognizing the geometry of the guide
rails, mapping algorithms for storing it in an efficient way, and active vibration controllers using
that information for damping the oscillations.
The estimation of the guide rail geometry is not a new topic and several approaches can be found in
the bibliography. [3] describes a method for characterizing the profile of a guide by integrating the
information from accelerometers placed on the lift and merging that information with the relative
displacements measured at the rolling guides. The further work [4] uses the obtained profile for
active vibration control. [5] describes specific procedures for production validation of guide rails,
although these techniques are not intended for elevators in normal operation. The proposal in the
present paper utilizes the approach described in [6]. It uses stochastic perturbation observers for
identifying the irregularities in the guide rail profiles and merges the information in a map. This
method allows for a reduction in the amount of data to be stored, and minimizes the number and
quality of the sensors as the identification is improved in consecutive measurements.
The information stored in the map is used for compensating the irregularities of the guide rails
therefore reducing the lateral oscillations. Other damping techniques in the literature normally lay in
one of the following categories: passive, active and semi-active vibration control. The first one
relies on the use of passive elements like springs or dampers for modifying the dynamic response of
the lift. In contrast to that, active vibration control systems use actuators that can both dissipate and
enter power into the system. In the literature interesting active approaches can be found using
different technologies: roller guides isolation systems based on magnetic actuators are described in
[7], linear actuators are selected for the same sort of application in [8]. Due to their higher cost, the
active vibration control systems are currently restricted to high price elevators. Semi-active
vibration control technology is an intermediate solution and it is based on modifying in real time the
dynamic parameters of the lift, like the effective stiffness or damping [9]. The present paper uses an
active vibration control with a feedforward compensation based on a map of the guide rails in
combination with a sky-hook feedback.
The present paper is organized as follows: the description of the installation and its mathematical
representation is done in the section 2; the proposed architecture appears in section 3; and finally,
sections 4 and 5 summarize the results of the virtual validation and the conclusions.
2 DESCRIPTION OF THE REFERENCE INSTALLATION

The algorithms described in the present paper are validated in a virtual model based on the
installation available at ITAINNOVA (figure 1) and described in [10].
Figure 1. Installation at ITAINNOVA for elevator tests
2.1 Mathematical representation

The following work uses a 2D state-space representation of the installation (figure 2):
x& = Ax + Bu (1)
Map-Based Active Compensation of Lateral Vibrations in Elevators 19-3
xul xur
θ
k, c k, c
h1
M,I
h2
xdl xdr
k, c k, c
x Global reference
Figure 2. Main parameters of the elevator representation
Where,
 x  0 0 0 0
   
 x&  0 0 0 0
θ  0 0 0 0
   
 θ&  0 0 0 0
x  0
 dl   &x&dl   0 0 0 
 
 x& dl   &x&dr  1 0 0 0  A11 A12 
x =  ; u =  ; B =  ; A =  
 xdr  &x& 0 0 0 0  A21 A22 
 ul 
 x&dr   &x&  0 1 0 0
   ur 
 
 xul  0 0 0 0
 x&  0 0 1 0
 ul   
 xur  0 0 0 0
x  0 0 0 1 
 &ur  
 0 1 0 0 
 
 − 4k / M − 4c / M 2k (h1 − h2 ) / M 2c(h1 − h2 ) / M 
A11 =  
0 0 0 1
 
1 2 1 2 1 ( 2 )
 2k (h − h ) / I 2c(h − h ) / I − 2k h 2 + h 2 / I − 2c h 2 + h 2 / I 
 1 2 ( )
 0 0 0 0 0 0 0 0 
 
 k /M c/M −k /M −c/M k/M c/M −k /M −c/M 
A12 = 
0 0 0 0 0 0 0 0 
 
 kh / I ch / I − kh / I − ch / I − kh / I − ch / I kh / I ch / I 
 2 2 2 2 1 1 1 1 
0 1 0 0
0 0 0 0
 
0 0 0 0
0 0 0 0
0 0 0 0
1 0 0 0
 
0 0 0 0
0 0 0 0
A21 = 08 x 4 ; A22 =  (2)
0 0 0 0 
0 0 1 0
0 0 0 0
0 0 0 0
 
0 0 0 0
1 0 0 0
0 0 0 0 0 0 0 0 

Table 1 contains the main parameters of the representation in (1) and (2).
Table 1. Summary of parameters
M (kg) I (km.m2) h1 (m) h2 (m) k (N/mm) c (N.s/m)

1220 200 1.9 0.6 416 1e4
3 ARCHITECTURE
The proposed architecture appears in figure 3. The control system recovers information from the
sensors in the elevator and estimates the geometry of the guides storing it in a stochastic map. This
information is used by the Active Vibration Control (AVC) for commanding the active rolling
guides and damping the oscillations.
h(Height)
ELEVATOR CONTROL
< xdl , xdr , xul , xur > y
Perturbation
observer
< xˆdl , xˆdr , xûl , xûr >

ROLLING
ACV Mapping
GUIDES Maps
Figure 3. Architecture of the system
The following subsections describe the different modules in the controller.
3.1 Perturbation observer

The perturbation observer identifies the geometry of the guide rails by the sensors installed on the
lift (the compression of the rolling guides and the speed at the top and the bottom of the lift). The
observer uses a Kalman filter with a discrete representation of the system for predicting the state x̂ p
at each instant k+1 (sample time τ ):
xˆ p ,k +1 = (I12 x12 + Aτ )xˆ k + pu

(3)
C p,k +1 = ACe,k At + C u
Where I12x12 is the identity matrix of size 12, and pu is the perturbation vector of mean zero and
covariance Cu. Ce refers to the uncertainty of the previously estimated state. As observed, the
equation (3) does not contemplate the inputs u (second derivative of the guide rail profile in (1))
and therefore assumes a high uncertainty Cu in the prediction. The subindex p makes reference to
predicted value, in contrast to the estimated one x̂e done with the sensor measurements:
(
K k +1 = C p ,k +1C t CC p ,k +1C t + C s ) −1
xˆ e,k +1 = xˆ p ,k +1 + K k +1 ( y − Cxˆ p ,k +1)

C e,k +1 = (I − K k +1C )C p ,k +1
(4)
The estimated state xˆ e, k +1 contains the profile values of the guides as it appears in the equation (2).
The matrix C makes reference to the measurement equation associated to the sensors.
• Speeds measured by two accelerometers at the level of the lower and higher rolling guides:
 x&l   0 1 0 h2 0 0 0 0 0 0 0 0 
  = y1 =   x = C1 x
x
 h
&  0 1 0 − h 0 0 0 0 0 0 0 0 
1
(5)
• Compression in the four rolling guides:
 ∆xdl   −1 0 − h2 0 1 0 0 0 0 0 0 0
   
 ∆xdr  1 0 h2 0 0 0 1 0 0 0 0 0
 ∆x  = y2 =  − 1 0 h1 0 0 0 0 0 1 0 0 0
x = C2 x
 ul   
 ∆x  1 0 − h1 0 0 0 0 0 0 0 1 0 
 ur   (6)
3.2 Mapping the guide rail geometry
Figure 4 shows the algorithm used for obtaining the map. It is a list of segments characterized by
the starting height (h0), the initial position of the guide (x0), the slop (m) and the length (l).
Figure 4. Mapping algorithm
In the following, an example of the mapping process is given. So, given a segment i in the map:
xi = x0,1 + mi (h − h0i )
(7)
When a new point of the guide rail profile xdi (j=l,r) is obtained from the perturbation observer it is
decided if it is part of a segment i by using the Mahalanobis distance:
xdj − Fsi
< Th
Cdj + FCsi F t
(8)
And by comparing the current height with the length of the segment in the map:
((h0i − h > 0)AND(h0i − h < ThL ))OR
((h − h0i − li > 0)AND(h − h0i − li < ThL )) (9)
ThL and Th are positive thresholds. Cdj is the uncertainty of the estimation from the perturbation
observer (extracted from the covariance Ce in (4)), and Csi is the uncertainty of the segment
parameters in the map (x0, m). The measurement function is defined as:
F = (1 h − h0i )
x 
si =  0i 
 mi  (10)
If the distance is too large, then it is considered as a new segment and it is included in the map list.
Otherwise, the segment i is updated with the new point by using a Kalman filter. In this case the
estimation at instant k is done with the parameters stored in the map for the segment i < si ,k ,C si,k > :
xˆ dj ,map = Fsi ,k
(
K s ,k +1 = C si ,k F t FC si ,k F t + C dj ) −1
si ,k +1 = si ,k + K s ,k +1 (xˆ dj , po − Fsi ,k)

C si,k +1 = (I − K s ,k +1F )C si ,k
(11)
Where, x̂dj, map is the predicted profile from the map.
3.3 Active vibration control

The proposed algorithm for active vibration control appears in figure 5. It has two components:
• The lift position and vertical speed is used for estimating the geometry of the guide rails
with the built map (section 3.2). The irregularities are compensated by modifying the
preload level at the rolling guides.
• A sky-hook term is added for increasing the damping of the system. It adds a force
proportional to the negative value of the elevator lateral speed at each rolling guide.
ACV ELEVATOR
Sensors Perturbation Lateral sp eed
observer
h (Height )
Signal MAP Sky-hook
v ( speed ) Advance
Filter
ROLLING
GUIDES
Preload
+
control
Figure 5. AVC algorithm
The signal advance is used for compensating possible delays in the map representation, which is
less and less important as the sampling time of the controller (1kHz in the tests below) or the
response time of the elevator to the perturbation increases. The filter is used for avoiding too fast
compensation commands that could damage the actuation system or cause impacts.
4 VIRTUAL VALIDATION OF THE SYSTEM

The following test is done at 1 m/s speed in an installation of 18 m height. The lift identifies the
map by arranging a cycle upwards and downwards (figure 6). After that, the AVC is activated and
the lift arranges other two cycles of lower range.
20
18
16
14
Height (m)
12
10
0
0 5 10 15 20 25 30 35
Time (sec)
Figure 6. Movement of the lift during the mapping phase
Figure 7 shows the displacement of the lift and the identified geometry of the guide rails.
-3
x 10 -3
Lateral position of the left guide rail (m)
x 10
Lateral position of the right guide rail (m)
6
6
4 5
4
2
3
0
2
-2
1
-4
0
-6 -1
0 5 10 15 20 25 30 35 0 5 10 15 20 25 30 35
Time (sec) Time (s)
Figure 7. Identified geometry of the guides
At the sight of the results, the perturbation observer obtains a good approximation of the geometry.
As the number of points to store is high, the mapping module merges them into a linear
representation. Figure 8 shows the result:
20
20
15
15
Height (m)
10
Height (m)
10
5 5
0 0
-5 -5
-0.02 -0.015 -0.01 -0.005 0 0.005 0.01 0.015 0.02 -0.02 -0.015 -0.01 -0.005 0 0.005 0.01 0.015 0.02
Lateral position of the left guide rail (m) Lateral position of the right guide rail (m)
Figure 8. Estimated map (the dashed lines with circles shows the map)
The result shows that the algorithm can merge the geometry in segments with a clear
correspondence with the real geometry. As observed in the figures, the map contains some segments
without a clear correspondence with the real geometry. This is caused by bad matchings due to
noisy measurements. Nevertheless, as there are more segments correctly identified in the region this
is not really a problem. Given the high uncertainty of these segments, it is easy to distinguish the
good ones from the bad ones. In order to choose the best estimation for the AVC two main
approaches have been evaluated:
• The estimation with lowest uncertainty from the different segments in the map is used.
• A weighted estimation considering the uncertainty obtained with the different segments:
1
segments
∑
k =1 Cij ,k
xîj ,k
xîj = segments (12)
1
∑
k =1 Cij ,k
In the final implementation, the first approach is used. Figure 9 shows the matching of the
estimation with the real value.
-3 -3
x 10 x 10
5 5
DR (m)
DL (m)
0 0
Real one
-5 Predicted one -5
0 10 20 30 0 10 20 30
Time (s) Time (s)
-3 -3
x 10 x 10
5 5
UR (m)
UL (m)
0 0
-5 -5
0 10 20 30 0 10 20 30
Time (s) Time (s)
Figure 9. Estimated perturbation in real time using the previously identified map
Figure 10 shows the perturbations at the elevator due to the geometry of the guides in two
consecutive cycles of 15 m at 1m/s, first one without AVC and the second one with AVC. As it can
be observed, the movement of the active guides compensates the irregularity of the geometry and
reduces the perturbation magnitude that arrives at the elevator.
-3
x 10
6
LL
LR
4 UL
UR
Perturbations (m)
2
-2
-4 No AVC AVC
-6
0 5 10 15 20 25 30 35 40 45
Time (s)
Figure 10. Perturbation entering the system
The resultant damped and short perturbations are well below the response time of the system (figure
11). The use of the proposed AVC results in a clear reduction in the speed levels of oscillation
(right), and almost no lateral movement (left):
-3
x 10
4
0.15
2
0.1
Lateral speed (m/s)
Lateral position (m)
0
0.05
-2 0
-4 -0.05
No AVC AVC -0.1 No AVC AVC

-6
-0.15
-8
0 5 10 15 20 25 30 35 40 45 5 10 15 20 25 30 35 40 45
Time (sec) Time (sec)
Figure 11. Response of the system when using the AVC
5 CONCLUSIONS
The present paper describes an Active Vibration Controller for compensating the horizontal
oscillations caused by irregularities in the guide rails. It continues the work in [6] for mapping the
geometry of the guide rails and it uses that algorithm for developing a feedforward command to the
preload level at the rolling guides so that it compensates the perturbations that reach the lift. In
order to damp the remaining oscillations the controller also includes a sky-hook compensator. The
algorithm has been tested by simulation and shows reductions in the maximum speed of around
70%.
The description in the present report represents a proof of concept based on simulation. The results
show promising capabilities of the technology and it is expected to implement it in real installations.
In order to do that, there are improvements that can still be evaluated, like arranging mapping and
compensation at the same time.
6 ACKNOWLEDGEMENTS
The work has been funded by the European Social Fund, the Department of Innovation, Research
and University (Spain), and the Ministry of Economy and Competitiveness (Spain).
REFERENCES
[1] A. S. Hamdy, Active damping of vibrations in elevator cars. PhD from the Swiss Federal In-
stitute of Technology Zurich (1997).
[2] M. Otsuki, K. Yoshida, K. Nagata, S. Fujimoto, T. Nakagawa, Experimental study of
vibration control for rope-sway of elevator of high-rise buildings. Proceedings of the American
Control Conference (2002).
[3] R.K. Roberts, Methods and apparatus for storing sensed elevator horizontal displacement
and acceleration signals. US Patent 5,524,730 (1996).
[4] R.K. Roberts, T.M. Remmers, C.A. Skalski, Elevator active guidance system having a
coordinated controller. US Patent 5,652,414 (1997).
[5] X. Yang, G. Zhang, J. Wang, L. Zhen, Instrument for measuring geometric errors of guide
rail of elevator. Transactions of Tianjin University, vol.5, n.2 (1999).
[6] J.M. Rodriguez-Fortun, R. Monge, J. Saumell, A novel state estimation algorithm for
identifying lateral perturbations in an elevator. ELEVCON (2016).
[7] K. Utsunomiya, K.-I. Okamoto, T. Yumura, K.Funai, H.Kuraoka, Active roller guide system
for high-speed elevators. Elevator World (2002).
[8] S. M. Rivas, E. A. Perondi, Development of a controller for an active suspension system for
high performance elevators. ABCM Symposium Series in Mechatronics, vol. 3, pp. 46-55 (2008).
[9] R. Monge, J.M. Rodríguez-Fortún, A. Gómez, J.A. Roig, P. González, Design of semi-active
roller guides for high speed elevators. Applied Mechanics and Materials, 706, 108-116 (2015).
[10] R. Monge, J.M. Rodriguez-Fortun, A. Gomez, J.A. Roig, P. Gonzalez, Design of semi-active
roller guides for high speed elevators. 11th International Conference on Vibration Problems
(2013).
Raúl Monge took his degree in Industrial Engineering in 2001 at the University of Zaragoza. He
joined ITAINNOVA in 2005 where, among other activities, he has worked on design and
experimental validation of different parts for the elevator industry. He is currently project
responsible.
Javier Saumell obtained his degree in Industrial Engineering from the University of Zaragoza,
Spain, in 2011. He is currently working at ITAINNOVA as R&D engineer. His research interests
are: system simulation and control methodologies.
José Manuel Rodríguez obtained his PhD in Informatics and System Engineering from the
University of Zaragoza, Spain, in 2012. Previously, he received the MEng and MS degrees in
industrial engineering in 2002 and 2007 respectively. He is currently working at ITAINNOVA as
R&D engineer and project responsible. His research interests are: nonlinear control algorithms and
autonomous navigation.
The Role of Economic Factors in Traffic Planning and Selection of
Lift Equipment
Rory S. Smith
University of Northampton, Northampton UK
[email protected]
Keywords: Traffic Planning, Equipment Selection, Net Present Value, Economics
Abstract. Traffic planning and lift equipment selection normally focuses on the quality and
quantity of lift service. While these metrics continue to be of great importance, economic factors
should also be considered when selecting lift equipment. The financial consequences of both over
lifted and under lifted buildings are explored. Low cost low performance solutions are contrasted
with high cost high performance systems. Simple financial engineering methods to evaluate
equipment selection, such as Net Present Value analysis, are presented. The financial aspects of
complex lift systems such as double deck, destination dispatch, and multiple cars in a single
hoistway are explained.
1 INTRODUCTION
Traffic analysis focuses on Quantity of Service, usually expressed as Handling Capacity, and
Quality of Service which is usually expressed as either Waiting Time or Interval. What is often
assumed is a lift system that meets the design requirements for quantity and quality of service is
also an economical solution. However, this may not always be the case.
2 ECONOMIC ANALYSIS
To understand how to define an economic solution one needs to review the economic factors that
affect buildings.
2.1 Return on Investment
Buildings are assets and represent an investment in real estate. Investors are always interested in
the return on their investment. Return On Investment (ROI) can be expressed as follows [1]:
= (1)
In the case of buildings, income represents the rents collected from tenants minus operating
expenses and asset value is the value of the lifts and all other building components.
If an investor places €1,000 in a bank account and the bank pays the investor €30 per year in
interest, then the investment has an ROI of 3%. If an investor builds a building for €100,000,000
and receives an income of €3,000,000 after all expenses, then the building investor also has an ROI
of 3%. Applying Equation 1 to the building example yields the following results:
, ,
= = = 3%
, ,
Three important factors that affect ROI in real estate investments are building cost, building lease
rates, and building efficiency. Building cost establishes the asset value for determining ROI.
Building lease rates are a major factor in determining income. Building efficiency, in this example
is the percentage of gross building area that can be leased.
2.1.1 Under lifted buildings

An under lifted building will have a poor lift service. Chapter 3 of CIBSE Guide D provides
guidance in the area of quality and quantity of service and what constitutes above and below
average quality of service in various types of buildings [2]. An under lifted building will have
below average service quality indicated by long waiting times and large queues.
In time, tenants will complain and eventually not renew their lease contracts. The building will get
a bad reputation and the only way to attract tenants will be to offer reduced rents. Reduced rents
and vacancies reduce income which in turn reduces ROI [4].
2.1.2 Over lifted buildings
Buildings with too many lifts will have great lift service. The asset value in the ROI equation will
increase but the income will not increase. Tenants will not pay a premium for service that is better
than needed. The higher asset value without an increase in income will reduce ROI. Additionally,
the extra hoistway space reduces building efficiency.
2.1.3 Over lifted buildings

A building owner would like 100% of the building area to be leasable. This would equate to 100%
efficiency. However, such things as electrical spaces, vertical Heating Ventilating and Air
Conditioning (HVAC) ducts and lift hoistways cannot be occupied and therefore cannot be leased.
It is for this reason that building owners want the lifts to occupy the least amount of floor space
possible.
Example 1: A lift manufacturer offers 1000 kg, 1 m/s lifts in two sizes; 1100 mm wide by 2100mm
deep and 1600mm wide by 1400mm deep [3]. The Hoistway area for the first size is 5.16 m² and
3.94 m² for the second size. The difference is 1.22 m² per lift per floor.
If 2 lifts with the smaller hoistways were installed in a 6 floor office building in Birmingham, UK
where office rents are €463 per square meter per year [5], the building would receive €6,778 in
additional income per year with no increase in costs.
If the building had 4 lifts and served 12 floors, then the increase in annual income would be
€27,112.
In these examples, the asset cost did not increase but the income did increase. Therefore, ROI is
increased. In evaluating competing lift designs that are otherwise equal, the one with the highest
ROI should be selected.
The Role of Economic Factors in Traffic Planning and Selection of Lift Equipment 20-3
2.2 Payback Period

Another simple and conservative financial tool used to evaluate investment decisions is Payback
Period [1]. This is a rule of thumb financial tool rather than a financial engineering tool and can
easily be misapplied. The Payback period is the number of years required to recover the cost of the
investment. Payback Period can be expressed as follows [1]:
#
!" = # (2)
#$ % & ' ( )* +
Example 2: Simulation has shown that a regenerative drive for a proposed 6 story building in
Birmingham, UK will save €120.90 per lift per year based on a cost of €0.18/kWh [6]. The lift
salesperson offers to provide regenerative drives for €350 each and advises the owner he will have a
payback period of less than 3 years.
What the salesperson did not understand is the Payback Period method is based on after tax cash
flow. If the building must pay a tax on profits of 20% [7], then the after tax cash flow is €120.90 x
0.8 or €96.72. Applying these values to Equation 2 yields a payback period that is now 3.6 years.
The Payback Period method does not consider the cash flows after the payback period. A better
method for evaluating the desirability of an investment is the Net Present Value (NPV) method [1].
2.3 Net Present Value

The NPV method considers the initial cash outlay, the desired return on investment, and the annual
after tax cash flow over the life of the investment. NPV is calculated as follows [1]:
#(*/
, -=∑ 4 ( 12)/ − (3)
Where: , - represents Net Present Value
6 represents time period
7 represents the project’s life
89: represents After tax Cash Flow
represents required rate of return
represents Initial cash Outlay

The Time Value of Money describes the benefit of receiving money now rather than in the future
[1]. A Euro today is worth more than a Euro a year from now because in normal times money
deposited in a bank earns interest. Ninety three cents (€0.93) deposited in an account that pays 8%
interest will be worth 1 Euro in a year. Therefore, the present value of a Euro collected one year
from now is 93 cents.
The Net Present Value calculation converts future cash flows to today’s value. NPV is used to
accept or reject an investment proposal. Since the NPV equation includes the minimum required
rate of return, , an NPV value of 0 or greater indicates that the investment delivers the required
rate of return [1].
NPV can be applied to the decision to accept or reject the offer to pay an additional €350 for
regenerative drives. The building developer requires a rate of return of 8%. Therefore, will have
a value of 0.08. The developer has determined that the economic life of the lifts is 20 years, giving
7 a value of 20.
Applying these values to equation 3 yields the following:

E
96.72
, -=; − 350
(1 + .08)
4
, - = 599.61
Since €599.61 is greater than 0, the offer should be accepted. Another way to look at this result is
to add the NPV to the IO; this yields a value of €949.61. If the regenerative drives were offered for
€949.61, the energy savings of €96.72 would have returned a NPV of 0, indicating that an
investment of €949.61 would have yielded the required ROI of 8%.
Today there are several lift solutions that can reduce the building area required for lift shafts at
additional initial cost. Many of these include more than one cabin per shaft such as double deck
lifts and systems with 2 or more lifts per shaft. To evaluate the economic viability of one of these
solutions one must calculate the increased rent income over the life of the building and use this to
income to calculate the NPV of the increased IO.
Example 3, High Tech Lift:

A proposed building in Birmingham, UK will be 18 storeys tall and will have 6 lifts rated at 1275
kg that will operate at 3 m/s. One lift provider has proposed a lift system that will only need the
hoistway and lobby area of a 4 car group of lifts. However, this lift system will cost €600,000 more
than a conventional 6 car group.
The developer likes the idea of the additional income but doesn’t like the additional cost. The lift
company suggests a financial analysis of the proposed system using the NPV method.
Each of the hoistways is 2800mm x 2000mm. The original layout called for 3 lifts to be located on
either side of a 3 meter wide lobby. Therefore, the area of the 2 lift shafts and the lobby that can be
eliminated is 2800mm x 7000mm or 19.6 m² per floor. As there are 18 floors, the high tech system
will increase the net leasable space by 352.8 m². Since rental rates in Birmingham are €463 per
square meter per year, rental income will increase by €163,346.
Income and after tax cash flow are not the same. The developer estimates there will be
administrative costs of 15% and corporate tax of 20%, which make the after tax cash flow
€106,175.
The €600,000 cost of the high tech system was not budgeted. The developer must borrow the
additional funds. For this reason he requires a rate of return of 15%. A portion of this 15% will be
interest paid to the lender and the remaining portion will be his rate of return after interest expense.
The lift system is planned for a 20 year life.
The NPV calculation based on these values looks as follows:
E
106175
, -=; − 600,000
(1 + .15)
4
, - = 64,585
The Role of Economic Factors in Traffic Planning and Selection of Lift Equipment 20-5
Since the NPV is greater than 0 the proposal can be accepted.
Example 4, Modernization
The management of an existing 12 storey building in Birmingham, UK is evaluating ways to
improve the profitability of their building. The building is located in a prime location, is well
maintained, but is 30 years old. The building is commanding rents that are 10% less than newer
buildings. One of the building’s major detractors is the performance of its 4 lifts.
The building management believes with modernized lifts, rents could be raised from the current
€417 per square meter to €430 per square meter. Rents for new buildings in the Birmingham area
are €463 per square meter. The management believes that modernized lifts would also halt further
erosion of their rental income.
The Building has 6,050 square meters of rental space above the lobby. The additional €13 per
square meter would increase gross income by €78,650 (13 x 6050) per year. There will be no
additional administrative costs associated with the increased income but there will be a 20% tax on
this income. Therefore, the revenue after tax will be €62,920 (0.8 x 78,650).
A lift company has offered to modernize the 4 lifts for €98,000 each or €392,000 total.
The building management and their lift consultant decide to evaluate the economic feasibility of the
lift modernization. They will use the NPV method and will assume a 20 year life and a rate of
return of 15%. The rate of return was based on financing the modernization. The NPV based on
these values appears as follows:
E
62,920
, -=; − 392,000
(1 + .15)
4
, - = 1,835
Since NPV is greater than 0, the modernization is economically feasible.
3 CONCLUSIONS
In the design of a lift system Quality and Quantity of Service are important. However, Economy is
also important. Alternative proposals that cost more than a traditional design may in fact be more
economical than a low cost design.
Understanding the time value of money, the concepts of Return On Investment and Payback Period,
as well as the use of financial engineering tools such as Net Present Value can help the lift
professional to present solutions that are more economical than those solutions developed without
considering economy.
4 REFERENCES
[1] Martin et al, Basic Financial Management. Prentice Hall, Englewood Cliffs, (1991).
[2] CIBSE, CIBSE Guide D: Transportation systems in buildings. The Chartered Institution of
Building Services Engineers (2015).
[3] Kone Design Collection Monospace 700 Available from:
http://cdn.kone.com/www.kone.co.uk/Images/MonoSpace_700_Design_book.pdf?v=4 Last
accessed: 26 May, 2016
[4] Rory Smith, “Underlifted Buildings in the Middle East” Symposium on Lift and Escalator
Technologies (2014)
[5] Cushman & Wakefield United Kingdom Office Market Snapshot Q1 2016 Available from:
http://www.cushmanwakefield.com/en/research-and-insight/uk/united-kingdom-office-snapshot/
Last accessed: 25 May, 2016
[6] Electricity Pricing Available from: https://en.wikipedia.org/wiki/Electricity_pricing Last
accessed: 24 May, 2016
[7] List of countries by tax rates Available from:
https://en.wikipedia.org/wiki/List_of_countries_by_tax_rates Last accessed 25 May, 2016
5 BIOGRAPHICAL DETAILS
EUR ING Dr. Rory Smith is Visiting Professor in Lift Technology at the University of
Northampton. He has over 47 years of lift industry experience and has been awarded numerous
patents.
Exploring the Concept of Using Lifts to Assist the Evacuation of
Very Tall Buildings
Peter Sumner
WSP Parsons Brinckerhoff
One Queens Drive, Birmingham, B5 4PJ, UK
Keywords: Evacuation, fire, lifts, tall buildings

Abstract: Evacuation times for very tall buildings, whether for planned evacuation, real fire or non-
fire emergencies, can be extreme. This paper explores the concept of using lifts to assist the
evacuation of tall buildings and discusses the major considerations for building designers.
As buildings are getting taller there is a need to consider the safety of occupants during evacuation.
The physical exertion that is required to walk down 100+ flights of stairs and in some cases for times
in excess of 2 hours can be very challenging for many people. This unexpected physical exertion
added to the stress of evacuating a building during an emergency can lead to tiredness, physical or
mental injury and fatality.
The design of buildings and complete lift systems to withstand the effects of fire, smoke, water and
loss of power can be very expensive in terms of equipment and in the potential loss of rentable area.
However, depending on the fire and life safety strategy of a given building, an emergency may not
require the simultaneous evacuation of all the occupants, therefore, evacuation by lift may not be
required from all levels of a building and may not require the use of all lifts.
1 INTRODUCTION
Evacuation times for very tall buildings, whether for planned evacuation, real fire or non-fire
emergencies, can be extreme. Whilst designers are required to consider safe evacuation of occupants
from all buildings, conventional stair evacuation of tall and very tall buildings can in itself be
hazardous.
Designers have been considering the use of lifts to assist the evacuation of tall buildings for some
time; firstly as a matter of code to enable safe egress of all persons including persons with disabilities
and secondly as a means of reducing the overall evacuation time and risk of injury to evacuees.
The premise for most buildings is that lifts shall not be used in case of fire and that there shall be
sufficient evacuation stairs to ensure a safe evacuation by all building occupants. The question is:
does this current design model best serve the needs of occupants of very tall buildings?
There are two main issues with stair evacuation; does the number of flights of stairs cause undue
physical stress to evacuees, considering their size, age and general ambulatory condition and; does
the time required to evacuate by stair lead to fatigue and cause undue physical and mental stress.
Evacuation stairs will always be an essential requirement for the Life Safety design of buildings either
as the sole means of evacuating the building or as a back-up to others means of evacuation. That said
there are obvious benefits of using lifts to assist the evacuation when it is safe to do so.
If lifts are to be used to assist the safe evacuation of buildings, cooperation must be achieved between
all persons responsible for the design of a building including, client, Architect and Engineers and
consent will be required from the local fire authority or building control department.
Incorporating lifts into the evacuation strategy of a building should therefore begin in the early
feasibility and concept design stages of the building. This paper considers the options available for
the evacuation of very tall buildings by use of lifts and stairs and discusses the design issues, technical
solutions and benefits, in terms of evacuation time and evacuee wellbeing.
2 BACKGROUND
As buildings have grown taller the need to consider efficient and assured access into and around
buildings and egress from buildings at all times and for all persons including persons with disabilities
has grown ever more important.
There have been technical discussions, specialist meetings, symposia and a vast number of papers
written over the years that give an understanding of the problems to encounter and solutions to be
found if lifts are ever going to be used to assist the general evacuation of buildings.
Throughout the 1990’s, American Society of Mechanical Engineers (ASME), National Fire
Protection Association (NFPA) and National Institute of Standards and Technology (NIST) all held
workshops where papers were submitted to aid discussion on fire evacuation using lifts (Elevators).
At that time the consensus was not generally in favour as there was a huge scepticism about the safety
of users, mainly due to a number of well documented disasters where people had died while using
lifts during building fires.
Most of the issues discussed were technical ones and included: machine failure, reliability of power
supplies, lifts passing through fire and smoke zones, exposure to water and inadequate operation; all
of which have since been addressed and can pose little or no problems for today’s design Engineers.
One other major concern remained and a study by So, Lo, Chan and Liu in 1997 [1], considering the
issue of human behaviour while evacuating from building fires, concluded that further research into
the subject should be undertaken.
The unprecedented attack on the World Trade Centre in 2001 which led to the collapse of WTC1,
WTC2 and WTC7 in less than 2 hours and to the death of 2,752 people has driven further studies into
Human Behaviour in fire emergencies. It is unlikely that the disaster could have been prevented by
enhanced design measures but the sheer length of time that it took evacuees to escape the building is
a matter for life-safety design and has been the subject of many studies since the 2001 disaster.
Egan [2] discovered that fatigue would be experienced in about 5 minutes and Pauls [3] that the
average speed of evacuation would be 1 floor per 16 seconds. Investigations into the evacuation of
WTC2 have shown times in excess of 60 seconds per floor. One of the problems is that as fatigue sets
in, evacuees will stop to rest and cause blockages in the escape stair thus causing increased evacuation
times for all.
The behaviour of human beings under the stressful activity of evacuating buildings in real fire
emergencies is something that it has been very difficult to model or to predict. However, the
evacuation of the World Trade Centre complex following the events of 11th September 2001 has
presented students and researchers with excellent insights into the factors that assist and hinder egress
within the high-rise building environment.
There have been many research papers on the subject of human behaviour in fires and many that were
commissioned following 9/11 and to address the issues raised by the evacuation of the WTC complex.
Since 1998, the annual International Symposium on Human Behaviour in Fire has given students and
researchers a platform to present their work and for delegates to debate the issues raised by research.
Exploring the Concept of Using Lifts to Assist the Evacuation of Very Tall Buildings 21-3
3 PROGRESS
Much progress has been made and more is now known on the behaviour of people in fire emergencies
and on the likely behaviour of people during an evacuation. As such, more people are beginning to
see the huge benefits that can be gained by designing lift systems to operate in fire emergencies and
to assist the evacuation process.
It is a fact that the design and management of lifts will cost more in terms of capital expenditure for
both the design and construction phases of a building but it can also lead to a reduction in income
return due to a likely reduction in the net lettable area through additional space requirements of lifts,
refuge spaces and the various other aspects of building design. As such, all parties involved with the
design will need consent from both client and Architect if the concept of improved life-safety through
reduced evacuation times is to become a reality.
Since 9/11, a number of buildings have been designed and constructed with the use of lifts to assist
the evacuation strategy and many more have undergone changes to their original life-safety strategy
to enable the use of lifts. One such development is Petronas Towers in Kuala Lumpur whose
evacuation strategy has changed since the building first went into service. [4]
This paper investigates the evacuation strategy of a number of very tall buildings, including Petronas
Towers and discusses the use of lifts to assist evacuation and life safety in those buildings. Finally,
the paper sets out the general principles of design and issues to be overcome when using lifts to assist
the evacuation of very tall buildings.
4 EXISTING BUILDING STUDY

Although evacuation by lift was not always a design priority, the use of lifts to assist the evacuation
of buildings in fire and non-fire emergencies has become increasingly more common place in recent
years.
There are a number of today’s tall and very tall buildings that use lifts in some way to assist the
evacuation process. A recent technical note by the National Institute of Standards and Technology
(NIST) in the US explored the evacuation strategy of twelve high-rise buildings and provided an in-
depth discussion of six of those buildings.
This study considers four of the buildings discussed in the NIST paper, three of which have held the
title of ‘world’s tallest building’. All 4 buildings are over 450m tall and as such are classed as very
tall buildings.
Table 1: Very tall buildings
Building Location Building World’s tallest Year

height (m) building opened
Petronas Towers Kuala Lumpur, 452 1997 – 2004 1998

Malaysia
Taipei 101 Taipei, Taiwan 509 2004 – 2009 2004
Shanghai World Shanghai, China 492 No 2008

Finance Centre
Burj Khalifa Dubai, UAE 828 2009 – Present 2010

4.1 Petronas Towers

The construction of Petronas Towers, Kuala Lumpur was completed
in 1998 and its height of 451.9m made it ‘the world’s tallest building’
at that time. The buildings are primarily office space with a single
tenant, Petronas chemical company occupying one tower and the other
being let to multiple tenants. One unique feature of the towers is the
adjoining bridge link at levels 41 & 42.
The fire and life safety strategy for the Petronas Towers was at the
time, designed to meet British Standard Code of Practice BS 5588:
Part 5. However, the Code of Practice that gave guidance on the means
of escape for disabled people was BS 5588: Part 8 which has since
been withdrawn in the UK but is often still referenced in other parts of
the world. The COP recognised the use of lifts for evacuation of
persons with disabilities and gave guidance on refuge spaces,
evacuation strategies and lifts [13, 14].
The evacuation strategy for persons with disabilities is to wait in a
designated refuge space adjacent to a firefighting lift (or other lift suitable for evacuation use) and/or
an escape stair and await assisted evacuation by building management or the fire service.
The evacuation strategy for persons within disabilities within Petronas Towers is as described in BS
5588: Part 8 and provides refuge space on all levels. The fire-fighting lifts are then used to evacuate
waiting persons with disabilities before the fire service arrive.
The evacuation strategy for able bodied people has changed since Petronas Towers first opened. At
that time, the strategy required occupants below the bridge link (level 41) to use the escape stairs to
safety and for those occupants above the bridge link to use the escape stairs to reach level 41 then to
cross the bridge link and use the lifts in the adjacent tower to safety. At that time an incident requiring
simultaneous evacuation of both towers was not considered.
Post 9/11, the strategy changed to account for the simultaneous evacuation of both towers. Today,
occupants above the link bridge use stairs to reach level 41 & 42 where double deck shuttle lifts are
available to transport evacuees to the ground and mezzanine floors and to safety. It is not known if
the shuttle lifts have special design features that allow their use when the fire is local to the lifts or
whether in this case the lifts shut down and evacuation reverts to either stairs or the other tower.
4.2 Taipei 101

Taipei 101 is an office building that also houses retail, a conference centre and
restaurants. Construction was completed in 2004 at which time Taipei 101
became the latest building to claim the title of ‘the world’s tallest building’.
The designers of Taipei 101 originally planned for traditional stair evacuation but
an evacuation which was conducted prior to completion took approximately two
hours to complete. Aware of the research undertaken after 9/11, the authorities
decided to try another evacuation but this time with the passenger lifts remaining
in service. The evacuation using lifts and stairs took 57 minutes as opposed to 2
hours using the stairs alone. [5]
The decision to include evacuation by lift in the Taipei 101 evacuation strategy
was made after the final design and construction stages and so only limited
modifications could be made to enhance the reliability and safe operation of lifts.
However, the enhanced features were only applied to special emergency/service lifts and to fire-
fighting lifts [5].
The building was designed with special refuge areas every 8 floors to allow persons who could not
use the stairs to wait in a fire-protected area to be evacuated by either the special emergency/service
lifts or the fire-fighting lifts [4].
The fire-fighting lifts and special evacuation/service lifts are the only lifts used in a fire emergency
and all other lifts, including the main passenger lifts are shut down. Although the full evacuation
strategy is unknown, it is stated that refuge areas and lifts are available to assist the evacuation of all
persons who cannot use the stairs which may be targeted at persons with disabilities but does not
discount other occupants.
It is a fact that the designers considered the evacuation by lift for persons who have difficulties using
the stairs although this strategy could not accurately quantify the number of persons who may need
to use the refuge spaces and lifts.
4.3 Shanghai World Finance Centre
The World Finance Centre in Shanghai is a mixed use development mainly
consisting of offices, a hotel and conference centre. Construction was complete
in 2007 and although the design intent had been to construct a 510m high
tower, due to restrictions on the height of the roof the building was constructed
to a final height of 492m [4].
The Shanghai World Finance Centre was designed to surpass the 1995 Chinese
code for the fire protection design of tall buildings (GB50045-95) which
required a refuge floor every 15 floors. Shanghai World Finance Centre was
designed with a refuge floor every 12 floors [6].
Two special lifts were originally designed to serve the observation deck at the top of the tower but
were modified to support evacuation from each of the refuge floors in an emergency [6]. Occupants
with disabilities and other occupants who cannot use the stairs to reach a refuge floor are required to
wait adjacent to one of the fire-fighting lifts for evacuation by building management or the fire service
[7].
The refuge areas serve two purposes, evacuees can wait and rest in a safe place before continuing
their journey on foot or they can wait for a lift to transport them direct to the ground floor. Evacuation
by lift is a managed strategy where priority is given to persons with disabilities and others who find
it difficult to manage conventional stair evacuation [7].
4.4 Burj Khalifa
The Burj Khalifa is a mixed use tower in downtown Dubai, United Arab
Emirates incorporating offices, a hotel and residential apartments.
Construction of the tower was complete in 2009 and to a height of 828m which
made it the world’s tallest building from that date. The building was opened to
the public in 2010.
The building was constructed to IBC: 2003 and to NFPA101 fire and life safety
code and was designed for the use of some lifts to assist the evacuation process.
A full building evacuation uses 10 of the 58 lifts installed in the building. [4]
Burj Khalifa has a total of 163 floors and has full fire protected and pressurised
refuge spaces on levels 43, 76 and 123. Occupants are expected to leave the
fire affected floors via the emergency stairs and walk to one of these refuge spaces where they will
be transported via lifts to the exit floor and safety [4].
Design information states that total estimated evacuation time using a mixture of stairs and lifts is 90
minutes with 55% of the 19,000 occupants using stairs and 45% using lifts. [7]
4.5 Summary of existing building evacuation strategy

All of the above buildings use lifts to assist the evacuation strategy but each in a slightly different
way.
The evacuation strategy for Petronas is different for occupants of the upper and lower zones of the
building. Occupants of the lower floors are expected to use the stairs to reach the building exit while
occupants of floors above level 42 use the stairs to reach level 42 before transferring to shuttle lifts.
Taipei 101 has special service / evacuation lifts to transport evacuees from refuge areas located every
8 floors to the main exit floor. The lifts were designed for the purpose of evacuation after the
construction stages and so only limited modifications to enhance reliability could be made. As such
it is unknown whether the available lifts provide sufficient capacity for the expected number of users.
Similar to Taipei 101, Shanghai World Finance Centre (SWFC) has special evacuation lifts to
transport evacuees from refuge areas to the main exit but in this case the refuge areas are every 12
floors. SWFC has only 2 lifts designed to assist the evacuation so the evacuation strategy is unlikely
to make provision for all occupants.
Burj Khalifa uses 10 lifts to assist the evacuation of the tower which operate between 3 specially
designed refuge floors and the main exit. Occupants use the stairs to reach the nearest refuge floors
where they wait to be evacuated by lift. The building design and evacuation lift configuration are
unknown but the referenced paper states that 45% of the occupancy can be evacuated by lift which
equates to 8,550 people.
5 DESIGN ISSUES
5.1 General
Irrespective of whether the building evacuation strategy makes provision for the evacuation of all
occupants or for disabled and injured persons only, lifts that are used to assist the evacuation will
have to be specially designed for the purpose and should be installed in a fire and smoke protected
core.
5.2 Safe and Reliable Operation
Many previous studies have considered the design issues relating to the safe use of lifts in a fire
emergency. One very early study in this regard was by So et al (1997) who listed a number of areas
of concern needing further research if lifts were ever to be used as part of an evacuation plan [1]. The
areas of concerned are discussed below:
The danger of machine failure can be brought about by: loss of power, non-fire related failure of
equipment or fire related damage to equipment and can occur whether the lifts are in normal service
or in firefighting or evacuation mode. With an unprotected lift there would certainly be an increased
risk of failure during a building fire; the main issue here is to try to minimise the risk of failure through
good design.
A building and lift installation that were designed and constructed in line with the requirements of
BS 9999 (2008) and BS EN81-72 (2015) should have a reduced risk of loss of power or machine
failure due to the effects of fire, smoke or water.
The above Code of Practice (COP) recommends that machine rooms be constructed within
firefighting shafts, defining a firefighting shaft as “a protected enclosure containing a firefighting
stair, firefighting lobbies and, if provided, a firefighting lift together with its machine room”. When
considering the possible failure of the main power supply the COP recommends the use of a back-up
power supply from an alternative source. Such a source could be either a separate substation or a
generator driven supply.
Research shows that lifts have a likely breakdown rate of 1 every 62 ½ days, equating to one
breakdown every 90,000 minutes [10]. The likelihood of a breakdown in a 10 minute evacuation
period would therefore be considered as 9,000:1. Since this is a case of balancing the possibility of
smoke breaking through to the firefighting shaft against that of the lift breaking down, two simple
control measures could be put into place that would reduce the risk.
Firstly, the breakdown rate could be improved by employing a more rigorous maintenance program
for lifts that may be used for evacuation and secondly, by monitoring for signs of smoke within the
firefighting shaft, the lift could be forced to the evacuation floor and out of service at the first signs
of danger [9]; in this case, evacuation would revert to stair only.
Obstruction of the fire service would only become an issue if the firefighting lifts were used as the
main evacuation lifts. As previously discussed, firefighting lifts can be used before the fire service
arrive on site to assist the evacuation of disabled persons. Once the fire service arrives, they would
take control of the lift and the operation of assisting injured and disabled persons out of the building.
It is recommended that in line with the current COP, the evacuation strategy only consider the use of
fire-fighting lifts for injured persons and for persons with disabilities. If the evacuation strategy
requires the use of lifts for the evacuation of other occupants, then lifts designed for the specific
purpose of evacuation should be used. In this case the risk of obstructing the fire service from going
about their duty is reduced.
The evacuation of a building may require large numbers of people to be transported from specific
floors of the building to an exit level in a very short space of time. Lift groups are not normally
configured for this type of traffic and may have inadequate lift configuration and operation for
this type of traffic.
For office buildings, the main passenger lifts are generally configured to provide acceptable
performance during the morning or lunchtime peaks, up peak and two-way peak respectively. Lift
systems for hotels and residential buildings may also be configured for acceptable performance during
two-way peaks but at different times of day. In all cases, the lift configuration will be designed for a
peak period of operation other than evacuation.
This does not mean that lifts cannot be configured with evacuation in mind or that the control system
cannot incorporate adequate evacuation software. Whatever configuration of lifts is eventually used
to assist the evacuation will require calculations to be performed to understand the likely evacuation
time when using lifts.
Lifts used during a fire emergency could be exposed to the effects of the fire while passing through
zones of danger. One solution for prevention of smoke entering the fire-protected core or lift shaft
would be to pressurise the core and/or lift shaft. The need for and extent of pressurisation would
depend on the evacuation strategy, building arrangement and lift configuration but in all cases, the
evacuation lifts, lift lobbies and refuge areas should be protected against smoke and the effects of fire
much the same as any other escape route or stair.
All lifts use electrical circuits, on the lift car, in the lift well and in the machine room and as such
should be protected from exposure to water. Water from sprinkler systems and direct from fire
service hose pipes could cause electrical failure if allowed to enter the lift well or machine room.
Firefighting lifts are designed to prevent water entering the lift shaft by ramps or gullies and to detect
and remove any water that finds its way into the lift shaft by sump pumps or drains in the lift pit. In
addition, wiring and equipment should be protected against the effects of water by being installed in
a minimum of IPX3 rated enclosures. [16]
Generally, passenger lifts are not designed to operate in the presence of water and additional features
should be installed to ensure that casual water from building fire prevention systems does not affect
the reliability of lifts that are to be used for evacuation.
Consideration should be given to the design of fire protection systems that do not require sprinklers
in lift shafts or lift lobbies and ramps and gullies should be installed at convenient locations to prevent
water entering the lift lobby and lift shaft.
Preventing water entering the lift lobby and lift shaft would be a better solution than providing the
water protection described above for firefighting lifts but it is unlikely that prevention methods can
be assured so a level of protection will also be required to equipment to ensure continued reliable
operation at all times.
One other non-technical point of concern was raised by So et al (1997), who foresaw problems
relating to the complex psychological reaction of the evacuees to a building fire and a forced
evacuation of the building. Evacuees may suffer an inability to understand and follow evacuation
guidelines in the stressful environment of a fire emergency. Apart from the stress, anxiety and
possible panic that evacuees may experience when the fire alarm is raised, they are likely to struggle
to carry through any pre-planned evacuation routine.
There is a recognised theorem that people require information in order to prevent the onset of panic.
Research in the field of human behaviour in fires has shown that panic is not inevitable and that clear
and precise information can help people to remain calm. [11]
So et al (1997) were concerned about lift operation in an evacuation and made a suggestion that lift
control systems with ‘computer vision’ would be better and that modern systems were more than
capable of this type of operation. From this approach it would seem that the authors were advocating
some type of crowd control by vision adjusted elevator control operation.
The above concept is not only possible but such equipment is available and adaptable for use on lift
control systems. It is recommended that all evacuation control systems use a type of Information Fire
Warning system (IFW’s) to pass lift and evacuation status information to evacuees waiting at upper
floors in an attempt to stop the onset of panic. [9]
This does not mean that evacuation operation should be by automatic control or any other type of
control, just that the progress of the evacuation and lift operation should be made visually and audibly
available for building occupants waiting to be evacuated.
6 LIFT CONFIGURATION
6.1 General
Each of the existing buildings presented in this report uses a different strategy for evacuation and
each strategy requires a different number of lifts to meet the expected demand.
However, some of those buildings had a different evacuation strategy in place at the design stage than
they have in place today and as such it is uncertain whether the lifts have sufficient design features to
ensure reliable operation or to ensure their use in all types of emergency.
It is important that the strategy is set early in the design life of the building and it can be met by the
existing lift configuration otherwise additional lifts may be required. Additional lifts mean less
rentable area and could affect the viability of the project.
The right lift configuration to assist the evacuation of any given building may be inappropriate for
another building and will depend on the type and use of the individual building and on the existing
lift arrangement.
Earlier we discussed the design issues to be overcome if lifts are to provide safe and reliable operation
during a building evacuation and touched on lift performance during evacuation mode. The right
solution is one that provides sufficient lift capacity to meet the needs of the evacuation strategy and
a robust design that ensures each of the design issues is met.
6.2 Theoretical Lift Performance

Diagram 1 depicts a lift arrangement for a typical tall building. High level calculations have been
performed to determine how many and what type of lifts are required to meet the expected demand
given the occupancy in the table.
The low, mid and high zones are served by Double Deck lifts, the sky lobby by Double Deck shuttle
lifts and super high rise zones by Single Deck lifts.
Table 1, contains high level results for the stated typical building arrangement and for an assumed
12% demand during an up peak period.
Diagram 1: Typical tall building arrangement
Table 1: Lift Arrangement
Building Lift Arrival Interval Capacity

Zone arrangement Rate Factor
(%) (s) (%)
Super high 6 x 1275kg SD 12 21.0 52.6

rise - Upper
Super high 6 x 1275kg SD 12 23.3 58.2

rise - Lower
Sky lobby 4 x 1600kg DD 12 29.9 60.9
High rise 8 x 1275kg DD 12 24.7 67.7
Mid rise 8 x 1600kg DD 12 26.1 66.5
Low rise 8 x 1600kg DD 12 25.8 73.8

Assuming the lifts for the above typical tall building meet the performance requirements for ‘Up’ and
‘Two-way’ peak traffic, it is almost certain that they will provide sufficient capacity for ‘Down’ peak
traffic; evacuation can be considered a form of down peak demand.
A potential evacuation strategy for the above building would be for all persons below the sky lobby
to use the stairs and for all occupants of the super high rise zone to use the stairs to the sky lobby at
levels 49 & 50, and from there use the shuttle lifts to exit the building.
If we assume a worst case of a total evacuation (1,920 persons) of the super high rise zone and that
the evacuation demand will be 100% down traffic, then the Round Trip Time (RTT) and Handling
Capacity of a given lift arrangement can be determined by calculation.
, 2003 = 2 + 2− +1 + + 2− (1)
The above RTT equation (1) presented by Barney [8] is for a Double Deck lift with multiple stops.
However, if we assume the shuttle lifts will travel between 2 set stops then, we can state that each
trip would include one stop only, with one period of loading, one period of unloading and two high
speed journeys between the ground floor and the sky lobby. The RTT equation can be simplified for
the proposed manual evacuation and would become as equation (2) below.
!"" ℎ$ % = 2 & + +1 +2 (2)
Where,
= Average number of stops
= Average number of passengers
& = Single journey travel time which can be calculated by kinematics for each journey to and from
the sky lobby.
= Time, associated with each stop: Door Open time + Door Close time + Start Delay
' = Period of time for a single passenger to enter or leave the car
Table 2: Kinematics
Lift Group Rated Speed Acceleration Jerk Travel distance Travel time
V (m/s) A (m/s2) J (m/s3) H (m) tT (s)
Shuttle Lifts 6 1.0 1.4 200 40.0
( ) * +, * + , )) = 24 * .! .
/0 1 $2" !3 * .. 1 . * + , = 24 × 0.8 = 19.2
.. 1 0 1 .3 ,2 , = 0.8 . +! (.
!"" = 2 & + +1 +2 (2)
!"" = 2 × 40.0 + 1 + 1 1.9 + 2.9 + 0.5 + 2 × 19.2 × 0.8

!"" = 121.32 . +! (.
9: !"" = ⁄:! ,3 . = 121.32⁄4 = 30.33 . +! (.
:!. <!$ . = $* ,1ℎ ,. !++$* + ⁄ .! . * ,* 2 = 1920⁄38.4 = 50

9: × :!. ,*. @
=0 +$ ,! ,2 $* ,1ℎ ,. >! = 60
=0 +$ ,! ,2 $* ,1ℎ ,. >! = 30.33 × 50 @
60 = 25.275 2, $ .
The above calculation is very simplistic but gives an idea of the possibilities of using lifts for
evacuation. The example lifts are shuttle lifts designed to meet performance targets for the morning
up peak and could theoretically evacuate the total super high rise zone in approx. 25 minutes.
The above example tall building may have an evacuation strategy that also requires the use of lifts
for persons in the mid or high rise zones for which there would be numerous options to execute the
evacuation. Each option would require consideration for the design of lifts, the building environment
in which the lifts operate and the performance of the group in evacuation mode.
7 LIFT DESIGN
7.1 Control
A decision needs to be made if the lifts are to operate on normal control, under management control
or with some special bespoke evacuation control.
The current Code of Practice for reference to means of escape for disabled people is BS 9999 (2008),
the Code of Practice for the design, management and use of buildings. The COP recommends to adopt
a management strategy for evacuation and suggests that lifts used to assist the evacuation of disabled
people should be operated under the direction and control of the fire safety manager. [15]
The previous, now withdrawn Code of Practice for means of escape for disabled people, BS 5588:
Part 8, also recommended to adopt a management strategy for evacuation and to avoid automatic
operation of lifts. [14]
7.2 Lift Lobbies
Lift lobbies and refuge areas should be considered fire protected cores with access to escape stairs
and with minimal risk of fire and smoke infiltration. Information Fire Warning systems should be
incorporated into the refuge areas and lift lobbies to provide up to date information on lift arrival and
departure status and to keep the evacuees informed as to the progress of the evacuation. [9]
7.3 Structural Implications
As discussed, the lift shafts should be designed with a minimal risk of smoke infiltration by
considering the fire loads of the lower ground and mezzanine floors and by avoiding the need for
pressure release holes in lift shafts.
Sky lobby lift lobbies should be designed without sprinkler systems or provided with a means to
prevent water entering the lift lobbies and lift shafts.
7.4 Building design implications

Every solution is different but in this case, the only implication on building design is the requirement
for a refuge area at the sky lobby level which will reduce the rentable area and increase the design
cost.
7.5 Lift design implications
The shuttle lifts should have additional features to enable safe and reliable use in fire or non-fire
emergency evacuation.
7.6 Building Services Implications

Normally only fire-fighting lifts require emergency power supplies but in this case an additional
supply would be required for the four double deck lifts serving the sky lobby.
In addition, there may be a need to pressurise the shuttle lift lobbies although it may be possible to
achieve this through natural means.
7.7 Evacuation strategy
The evacuation strategy for a design similar to the above example would require management control
of the evacuation, emergency telephones at the lift lobbies, refuge areas and emergency command
centre and a number of trained staff located in the refuge areas and lift lobbies.
8 DISCUSSION
The intent of this paper was to discuss the implications of using lifts to assist the evacuation of tall
and very tall buildings. The paper investigated the evacuation strategies of four existing tall buildings
including three that have held the title of ‘The World’s Tallest Building’.
With reference to previous papers on the subject, the main design issues were discussed and solutions
presented for design requirements that would ensure safe and reliable use of lifts during evacuation.
A typical building was presented as an example, showing that generally, lifts configured to meet
performance requirements during a main traffic peak would normally provide acceptable performance
during evacuation mode. In this case, shuttle lifts that were designed to transport 12% occupants of a
super high rise zone in 5 minutes during a morning up peak, were capable of evacuating the entire
super high rise zone in <25 minutes.
It is important that if lifts are to be used to assist the evacuation of a building that they are part of the
overall life safety strategy for the building. Many modern buildings have compartmentalised
construction and employ phased evacuation where only floors immediately adjacent to the fire floor
are evacuated. However, in cases where the fire spreads and the phased evacuation is escalated, it
may become necessary to evacuate a complete zone or even complete building. For this reason, a
total evacuation should always be modelled.
Every building is different but if consideration is given to the design of lift systems to assist the
evacuation strategy at the building concept stage then all parties to the design process can have an
input. The use of lifts to assist either a full or partial evacuation of any building is possible but depends
on early cooperation between client, architect and design engineers. There will always be sufficient
lifts in a building to evacuate the total occupancy in a reasonable time, the question is can the building
afford a design that would make it safe to use the lifts.
A number of buildings currently under design have accepted principles of design that will enable lifts
to be used to assist evacuation and it is hoped that presentation of case studies for these buildings will
be possible at future symposium.
REFERENCES
[1] So, A. T. P., Lo, S. M., Chan, W. L. & Liu, S. K., (1997) Fire Escape by Elevators – Elevatori,
Volume 26 Milan, Italy
[2] Egan, M. D. (1986) Concepts in Building Fire Safety: Robert Krieger Publishing Co. cited in So,
A. T. P. et al (1997) Fire Escape by Elevators – Elevatori, Volume 26 Milan, Italy
[3] Pauls (1987) cited in So, A. T. P. et al (1997) Fire Escape by Elevators – Elevatori, Volume 26
Milan, Italy
[4] Kinateder, M. T., Omori, H, Kuligowski, E. D. (2014) NIST Technical Note 1825 – The use of
Elevators for Evacuation in Fire Emergencies in International Buildings
http://dx.doi.org/10.6028/NIST.TN.1825
[5] Chien, S-W and Wen, W-J (2011) A research of the Elevator Evacuation Performance and
Strategies for Taipei 101 Financial Centre – Journal of Disaster Research 6, page 581-590
[6] ArcelorMittal (2013) Shanghai World Finance Centre – Constructalia. Cited in So, A. T. P. et al
(1997) Fire Escape by Elevators Elevatori Volume 26 Milan, Italy
[7] Sun, Yang (2013) MSc Thesis, Egress as part of fire safety in high-rise buildings – DELFT
University of Technology
[8] Barney, G. Dr. (2003) Elevator Traffic Handbook, Theory and Practice – Taylor & Francis,
London, UK
[9] Sumner, Peter 2003 – Fire-fighting and Evacuation Lifts, exploring the concept of using lifts to
escape building fires – MSc Lift Engineering, University of Northampton
[10] Howkins, R. E. (2002) Elevators for Emergency Evacuation and Egress – Elevator World
February 2002 Mobile, Alabama
[11] Canter D. (1980) Fires and human behaviour – London: John Wiley and Son
[12] Klote, J. H., Deal, S. P., Donoghue, E. A., Levin, B. M. & Groner, N. E. (1993) Fire Evacuation
by Elevators – Elevator World June 1993 – Mobile, Alabama
[13] British Standards Institute (1991) BS 5588: Fire Precautions in the Design, Construction and use
of Buildings: Part 5: Code of Practice for Firefighting stairs and lifts – London, UK
[14] British Standards Institute (1999) BS 5588: Fire Precautions in the Design, Construction and use
of Buildings: Part 8. Code of Practice for means of escape for Disabled Persons – London, UK
[15] British Standards Institute (2008) BS 9999: Fire safety code of practice for the design,
management and use of buildings – London, UK
[16] British Standard Institute (2015) BS EN 81-72 Safety rules for the construction and installation
of lifts – Particular applications for passenger and goods passenger lifts Part 72: Fire-fighters lifts –
London, UK
BIBLIOGRAPHY
Canter, D., Powell, J. & Booker, K. (1988) Psychological aspects of Information Fire Warning
Systems Watford: BRE
Tong, D. & Canter, D. (1985) Informative Warnings, in situ evaluations of Fire Alarms – Fire Safety
Journal cited in Canter, D. et al Psychological aspects of Information Fire Warning Systems – BRE,
Watford, UK
CIBSE (2015) Guide D: Transportation Systems in Buildings – Page Bros, Norwich, UK
Council on Tall Buildings and Urban Habitat (1992) Fire safety in tall buildings – McGraw-Hill, New
York, USA
Dupree, J. (2013) Skyscrapers, a history of the world’s most extraordinary buildings – Black Dog &
Leventhal Publishers, New York
Peter Sumner is currently an Associate Director with WSP | Parsons Brinkerhoff working as an
engineering consultant in the Vertical Transportation team. Peter has been in the lift industry for 34
years and before entering consultancy had previously worked in all sectors of the business from
maintenance engineer to International Technical support engineer with ThyssenKrupp Elevator.
In 2003, Peter earned an MSc in Lift Engineering at the University of Northampton, gaining a
distinction and a Professional Engineering Institute award for his dissertation on Fire-fighting and
evacuation lifts. He was appointed to the Board of Studies at University of Northampton in support
of Undergraduate Degree and MSc courses in Lift Technology and currently provides support to the
School of Science and Technology on lift traffic design.
Understanding the Requirements of the New EN-81 Standards with
Respect to Speed Monitoring, Speed Reducing and Prevention or
Stop Devices
Sam Tanno
Southmeads, Ballinger Road, South Heath, Great Missenden, Buckinghamshire, HP16 9QH.
[email protected]
Keywords: EN81-20:2014, EN81-50:2014, BSI, standard, documentation, construction,

installation, lift, overspeed governor, ascending car overspeed, unintended car movement,
protection, car doors, buffers, safety gear, rope brake, stop devices, detection means.
1 ABSTRACT
Standard documents are industry guidelines rather than legislation to be enforced; standard
publications do not act as industry police, but rather advice to be discussed by those involved in the
industry. Manufacturers and producers are not obliged to adhere to them, but can fully expect to
find themselves at the mercy of the law should consequences arise from any negligence, neglect or
malpractice. Designers, however, are free to have their design tested by another notified body,
should they desire. Thus, they are widely followed, and the source of much discussion and debate.
This piece examines sections of the latest set of standards, EN81-20 and 50, to be released by
British Standards Institute (BSI), governing the safety rules for the construction and installation of
lifts.
2 INTRODUCTION
In August 2014, BSI published their new standard documentation covering safety rules for the
construction and installation of lifts. The aim of this paper is to assist in the cause of understanding
a few key areas of the publication.
The first part of the paper examines what the new standard defines as the difference between
ascending car overspeed (ACO) and unintended car movement with doors open (UCM), in the
context of the speed monitoring, speed reducing and prevention or stop devices allowable.
It considers the nature of each situation and how the standard deals with them accordingly. It
ponders apparently contradictory sections of the standard before attempting to clarify them, and also
casts doubt over common ways of complying with the standard such as doubling or tripling up on
lift machine brakes. Having done this, the document then proposes standard-compliant systems for
protection against excessive speed in the up direction and unintended movement with car doors
open respectively.
The paper then moves on to the matter of the application of the speed monitoring, speed reducing
and prevention or stop devices and looks at what the requirements are for testing and certification,
examining certain sections of the standard. Once again, it picks out passages of the standard that
could be construed as ambiguous and unclear, considers the implications of what the standard says
and attempts to elucidate the abstruseness. This section is slightly more technically involved,
considering calculations to determine permissible mass and energy absorbed by safety gears as well
as looking at type testing procedures, before eventually drawing conclusions as to the author’s
interpretation of the standard as well as compliant safety systems.
3 UNDERSTANDING THE REQUIREMENTS OF THE NEW STANDARDS

3.1 What is the difference between excessive speed in the up direction and unintended car
movement with doors open in terms of the speed monitoring, speed reducing and
prevention or stop devices allowable by the standard?
Protection means against excessive ascending car speed and unintended movement are covered in
the BS EN81-20:2014 [1] standard publication, sections 5.6.6 and 5.6.7 respectively. To the casual
observer, protective means against excessive speed in the up direction and unintended movement
with doors open could be seen as needing, fundamentally, to do the same thing – stop or slow the
lift upon detection of an undesirable situation. Be it “the lift is moving when it should not be” or
“the lift is moving too quickly”, the objective is to then stop the lift as quickly as is safe. While this
is true to an extent, it does not take into account important differences between the two situations.
Although both are hazardous, unintended movement with car doors open is a more dangerous and
serious situation than excessive speed in the up direction – primarily because the car doors are open
and people may have been getting in and out of the lift when it started to move. The means for
protection against unintended car movement needs to be able to detect a certain amount of
movement with the car doors open, whereas ascending car overspeed protection needs to detect a
certain speed in the up direction. As such, the standard appears to place tighter constraints on
unintended movement protection means. It says the means must detect unintended movement of the
car, stop the car and keep it stopped. This is as opposed to the excessive speed detection means,
which is to detect excessive speed of the ascending car and stop or reduce its speed to that for which
the counterweight buffer is designed.
A potential area for debate pertains to the causes of unintended car movement. This is covered in
EN81-20, clause 5.6.7.1. It says:
“Lifts shall be provided with a means to prevent or stop unintended car movement away
from the landing with the landing door not in the locked position and the car door not in the
closed position, as a result of any single failure of the lift machine or drive control system
upon which the safe movement of the car depends.
Excluded are failures of the suspension ropes or chains and the traction sheave or drum or
sprockets of the machine, flexible hoses, steel piping and cylinder. A failure of the traction
sheave includes a sudden loss of friction.” [1]
The question is whether it is fair and responsible of the standard to discount failures of the
suspension ropes, chains and traction sheave, including a loss of friction. Granted, the standard
includes design measures for traction which should ensure traction over the range of 0% to an
overload, and uncontrolled slipping would not occur without steady deterioration to the ropes and/or
sheave, which should be detected by a suitable maintenance regime. Unfortunately, these design
principles are not always adhered to and maintenance practice not always followed, especially as
lift engineers currently have an average of 12 minutes to perform maintenance checks on lifts, and
uncontrolled slipping does occur, albeit seldom.
Section 5.6.7.3 says:
“The means shall be capable of performing as required without any assistance from any lift
component that, during normal operation, controls the speed or retardation, stops the car or
keeps it stopped, unless there is built-in redundancy and correct operation is self-
monitored.” [1]
Understanding the Requirements of the New EN-81 Standards with Respect to Speed Monitoring, Speed 22-3
Reducing and Prevention or Stop Devices
This is affected by the issue about neglecting uncontrolled slipping because, obviously, overcoming
of friction forces between the ropes and the traction sheave would lead to unintended car movement
regardless of whether the sheave is moving. If that was the case, the brakes would be clamped
closed on the brake drum and the traction sheave would be stationary, but the ropes would still be
moving through the grooves. Then it would not matter how many machine brakes you have, they
would all be useless – the ropes are going to continue to run through the grooves and the car would
not stop under this means.
The same problem also applies to sheave brakes. They act on the sheave itself so, once again, you
can have as many sheave brakes as you like, but they will not make a difference if the suspension
ropes have overcome the friction forces and are slipping through the grooves of the traction sheave
because the ropes will be moving regardless of whether the sheave is. It is worth noting as well, on
the subject of sheave brakes, that, while the standard states that the brakes are to be able to bring the
lift to a halt without the need for at least one brake so as to achieve built in redundancy, the reality
is that, in a lot of cases, it is unknown by those concerned with the safe operation of the lift whether
the lift is being stopped by all of the sheave brakes under normal operation. If all of the sheave
brakes are stopping the lift under normal operation, there is no longer any built-in redundancy, so a
sheave brake is no longer a compliant means against unintended car movement.
Overspeed governors in tandem with safety gears or rope brakes can be used as a protection means
against unintended movement with car doors open, but the governor must be able to detect 150-200
mm of movement with car doors open in order to perform as required and trigger the stopping
element of the means. Although standard friction type governors are unable to detect uncontrolled
movement, certain types of friction governors as well as drop-jaw and electronic governors are able
to detect unintended movement. Drop-jaw governors, for example, incorporate an electrical safety
control circuit where a contact is closed upon detection of a 200 mm movement of the governor
rope in one direction, triggering the stopping means. An overspeed governor that has been type-
tested for use as a detection means and can detect unintended movement in tandem with safety
gears or rope brakes is one example of a compliant means of protection against unintended
movement, however more often than not, this means consists simply of a safety circuit that de-
energises a solenoid when the car skate leaves the landing lock roller.
Another issue that should be confronted, on this matter, is whether the level of safety stipulated in
the standard for both UCM and ACO situations are reasonable; is it reasonable to expect a
protection system to bring a lift car to a halt in 1200 mm in an unintended car movement situation?
At each of the detection, activation and stopping stages of the response, there are systemic delays,
from sensor trigger, electrical signals and solenoid delays to overcoming governor rope inertia and
physical movement of the stopping means. These delays all add up as well as leading to a higher
maximum speed than what would be expected and mean that the detection distance and, as a result,
the braking distance are both increased. As such, the tripping speed of the governor could easily be
surpassed in the time taken for the safety gear to start braking, especially in a modern MRL system
with a gearless machine and very low system friction and high inertia, where the acceleration in the
up direction with an empty lift car could be significant. In the case of ascending car overspeed
protection, the tripping speed is the activation condition, meaning that there is no question that the
trip speed is exceeded by the time the stopping means start to act. Therefore, 1200 mm to stop the
lift car could be portrayed as slightly optimistic in a control circuit-solenoid activated system when
talked about in this light. The object of the standard is to provide pragmatic regulations that give
designers some leeway to achieve them based on the technology available, whilst still maintaining
an acceptable level of safety. However, of course, the standard needs to stipulate a stopping distance
which ensures injuries (or otherwise) do not occur as a result of UCM or ACO.
Having said this, it could also be argued that to call for 200 mm of movement as the trigger for
UCM with car doors open is a somewhat luxurious bearing in mind the accuracy of modern control
circuits and self-levelling capabilities of lift systems. Control systems are capable of detecting the
tiniest fraction of unexpected lift car movement, and the system is capable of self-levelling to within
a fraction of 200 mm. Calling for more tightly controlled trigger conditions, such as a reduced
amount of movement as the trigger for UCM, would result in reduced likelihood of exceeding the
governor tripping speed in a UCM situation.
The ascending car excessive speed protection means is, understandably, a more straightforward
standard to comply with than unintended movement, with less ambiguity. The standard recognises
that it is fairly standard practice to use an overspeed governor to detect excessive speed and then
trigger a rope brake or safety gear upon detection.
The standard for ascending car overspeed protection means states, in clause 5.6.7.3:
“In the case of using the machine brake, self-monitoring could include verification of correct
lifting or dropping of the mechanism or verification of the braking force.” [1]
It is not entirely clear what this passage means; it could be construed as referring to a machine brake
as the means for self-monitoring. Of course, the machine brake cannot be used for overspeed
detection. The machine brake does not contain a speed monitoring component as standard and
would not be able to detect overspeed. A reasonable conclusion to draw would be that it is referring
to an overspeed governor.
The standard permits a mechanical linkage to the car to assist the ascending overspeed protection
means in its performance. No mention is made of such a linkage in the standard on unintended car
movement. If this ‘mechanical linkage’ is assumed to mean the actuation means, this would, of
course, be analogous to the overspeed governor rope in an overspeed governor/safety gear system
used for protection means against ascending car overspeed. Whether this means that a mechanical
linkage is not permitted on the unintended movement protection means is another area of
ambiguity. Since overspeed governor and safety gear systems can be used for protection against
unintended car movement, it would seem fair to conclude that this linkage is permitted, or that
requirements elsewhere in the standard for the overspeed governor rope would be applied.
To sum up, devices that are definitely compliant for protection against excessive speed in the up
direction and unintended car movement with car doors open respectively include:
Table 1: Devices compliant with BS EN81-20:2014 for protection against excessive speed in
the up direction and unintended car movement with doors open.
Protection against excessive speed in the Protection against unintended car

up direction movement with car doors open
• Overspeed governor in tandem with: • Overspeed governor (such as drop-

• Rope brake; or jaw) that can detect car movement
• Safety gear with doors open in tandem with
• Other ascending car overspeed • Rope brake; or
protection means • Safety gear
• Counterweight buffers • Other unintended car movement
protection means
3.2 Depending on application of the speed monitoring, speed reducing and prevention or
stop devices, what are the requirements for testing and certification?
It should be reiterated, to commence this section, that unintended car movement and ascending car
overspeed are two entirely different situations with important differences as far as protection is
concerned. The testing and certification standards for all of these devices are laid out within BSI’s
EN81-50:2014 standard publication, sections 5.3 – 5.5 and 5.7 – 5.8.
Section 5.3.2.2.1 covers the method of type testing for instantaneous safety gears, the standard
stipulates that “the deformation of the safety gear block as a function of the force or as a function of
the distance travelled” [2]; what this statement refers to could be seen as slightly ambiguous.
Without knowing whether “block” refers to a specific component on the safety gear or the safety
gear itself, a reasonable guess would be that “deformation of the safety gear block” means the
change in distance between the load-bearing side elements of the safety gear throughout the course
of the test. Test methodology and instrumentation could be discussed with a notified testing body.
This use of the term “blocks” also creates doubt in section 5.3.2.2.2 for the test procedure; the
standard reads “reference marks shall be traced onto the blocks”.
On the same subject, it is noticeable that there is no mention of exactly how far the distance
travelled is, unless that is covered in “the arrangement and fixing details” which, as outlined in
5.3.2.1, are to be determined by the laboratory in accordance with the equipment that it uses.
Section 5.3.2.2.3.2 concerns the measurement of the capacity of the safety gears. It reads:
“The capacity of the safety gears shall be established by integration of the area of the
distance force chart”
Determination of the permissible mass encompasses energy absorbed by the safety gear, and section
5.3.2.3.2 outlines this. The standard gives a formula by which permissible mass,( + ) is worked
out based on the preceding theory around the total energy, K, one safety gear is capable of
absorbing. It gives this formula as:
2 ∙ =( + ) ∙ ∙ℎ
Although the standard does not explain why the K value is multiplied by 2 on the left hand side of
the formula, it would be a more than reasonable assumption that it is because there are (generally) 2
safety gear blocks acting in the same direction on any given lift car.
An area of possible controversy is the dividing safety coefficient on the bottom of the equation for
permissible mass. The formula is given by the standard as:
2∙
( + ) =
2∙ ∙ℎ
K is calculated by the integration of the area under the distance-force chart;
2 is taken as the dividing safety coefficient.
Traditionally, safety gear design pre-dates modern techniques such as FEA, and encompasses the
potential for a number of activations during the safety gear’s lifetime. This is why the safety factor
is stipulated as 2. Although taking two as the dividing safety coefficient seems a reasonable
approach, in reality there a number of factors that affect the appropriate safety margin in any given
case. These include material, stress, geometry, failure analysis and reliability issues. [3] When
calculated as a function of all these factors, the safety margin could be anything from around 1.1 to
8 or more. It can be possible to significantly under- or over-estimate the performance of the product
if an excessive safety factor is taken on related calculations. It could be argued that, because these
calculations are based on experimental findings, no safety factor need be taken at all due to the
calculations being based on real performance. Perhaps however, due to the attained K value being
put into another formula to find the P + Q mass and theoretical relationships not always being
entirely accurate in reality, this would not be advisable.
Section 5.3.3 of the standard covers the testing regulations for the progressive safety gear. The
opening section of this reads:
“If the safety gear shall be certified for various masses, the applicant shall specify them and
indicate in addition whether adjustment is by stages or continuous.” [2]
Surely, whether the adjustment of the safety gear is continuous or in stages does not make a
difference because, ultimately, the only way to carry out testing is to conduct a number of
individual tests at incremental masses with the safety gear adjusted according to each mass.
Therefore, however small those increments are, it is always going to be adjustment in stages as
opposed to continuous. It is not terribly clear how a safety gear can be adjusted continuously – or
how this can be accounted for in testing methods, so why the applicant needs to state this is unclear.
The test procedure for the safety gear certified for a single mass is outlined in section 5.3.3.2.2.1 as
follows:
“The laboratory shall carry out four tests with the mass( + ) . Between each test, the
friction parts shall be allowed to return to their normal temperature.
During the tests, several identical sets of friction parts may be used.
However, one set of parts shall be capable of:
a) three tests, if the rated speed does not exceed 4 m/s;
b) two tests, if the rated speed exceeds 4m/s.” [2]

Carrying out four tests if one set of friction parts is to be capable of either three or two tests,
depending on speed seems moot. Whilst it may be argued that conducting four tests for units
intended to be used two or three times provides an additional assurance about their longevity, it
would seem more sensible for the number of tests carried out to correspond with the number that
one set of friction parts should be capable of; then if deemed necessary, an additional run of the
same number of tests can be carried out using new braking parts.
It could be argued that the same set of friction parts should be capable of more than two or three
tests anyway. The problem with this arises when one considers that safety gears installed on site are
tested once at an acceptance test and, sometimes, another time at a witness test, which means that
after one more engagement, the braking parts may need to be changed. Some safety gears are
capable of upward of 100 tests without changing the friction parts, or any component for that
matter. Among these are the VG (variable geometry) range of safety gears supplied by Atwell
International, as demonstrated at the 2007 Interlift Trade Fair in Augsburg, when 121 tests were
carried out without loss of performance.
This also relates to a later section of the standard, contained within section 5.3.3.2.3.1. It says:
“NOTE Tests have shown that the coefficient of friction could be considerably reduced if
several successive tests were carried out on the same area of a machined guide rail. This is
attributed to a modification in the surface condition during successive safety gear

operations.” [2]
As already alluded to, the 2007 Interlift Trade Fair in Augsburg saw a VG4 safety gear supplied by
Atwell International work repeatedly without a reduction in braking performance over the course of
121 drop tests followed by similar performance over 30 more tests at Atwell International premises.
This phenomenon is most likely due to the type of friction that occurs between the braking parts and
the guide rail. When the safety gear first brakes on a new section of guide rail, the toothed carbide
inserts are “cutting through” the rail as they travel. Hence, there are two types of friction acting
between the inserts and the guides over these tests: ploughing friction and sliding friction. The teeth
of the inserts are both “ploughing” a groove into as well as “sliding over” the guide rails. After the
first few tests (how many depends on the guide rail characteristics and the load being put upon each
insert by the spring force), the teeth will have penetrated to their full capacity and the only type of
friction that will be taking place is sliding friction. It is, if anything, slightly more efficient to have
simply sliding friction taking place than a combination of that and ploughing friction. In my
opinion, the statement in section 5.3.3.2.3.1 may not be the case in all circumstances and this is
backed up by the test results gained in Augsburg in 2007.
The next section, 5.3.3.2.2.2, covers the testing procedure for safety gears certified for different
masses. It states that “two series of tests shall be carried out” for the maximum and minimum values
applied for. The terminology is quite vague and open to interpretation. It would be reasonable to
assume that, as in the previous section of the standard, a “series of tests” means 4 tests, with one set
of friction parts being capable of either 2 or 3 tests depending on speed.
Moving on to type testing of overspeed governors, and section 5.4.1 of the EN81-50 standard,
which covers the general provisions for testing. The standard calls for the applicant to indicate the
type of safety gear to be used with the governor. While this is understandable, as the co-ordination
between overspeed governor and safety gear is crucial, section 5.4.2.1, which covers the test
samples to be submitted, does not list a safety gear. If the standard considers the safety gear
sufficiently crucial to the operation of the overspeed governor to merit the applicant disclosing the
type of safety gear, it would seem logical that the test also demonstrate this co-ordination in
physical terms as well.
Section 5.4.2.2.2 relates to test procedure of the governor and states that:
“The acceleration to reach the tripping speed of the overspeed governor should be as low as
possible, in order to eliminate the effects of inertia.
In addition a minimum of two tests shall be made with an acceleration of between 0.9 and
1 in order to simulate a free fall situation and prove no further deterioration of the
governor has been caused.” [2]
It is not made fully clear whether the acceleration of the governor rope or the governor pulley shall
be measured. It would, perhaps, be a reasonable assumption that this arrangement is flexible
dependent on the individual set-up of any given test. Although, in any system involving friction,
true free fall is not possible due to what is known as ‘system losses’ (hence the standard stipulating
a minimum of 0.9 ), what is not clear is how the oft-uncontrollable, or at least difficult to control
and quantify, system losses can be limited with any reasonable certainty to no more than 10%.
Section 5.5.3.1.1 of the standard deals with the test procedure for energy dissipation buffers. In it, it
says:
“The acceleration and the retardation shall be determined as a function of time throughout
the movement of the weights.” [2]
This is all very well and good but section 5.5.3.1.2.4, dealing with the measurement of the
retardation of energy dissipation buffers, says:
“If there is a device for measuring retardation (see 5.5.3.1.1), it shall be placed as near as
possible to the axis of the buffer, and shall be capable of measurement with the tolerances of
5.1.2.6.” [2]
The fact that it says if there is a device for measuring retardation seems slightly contradictory when
put into the context of the earlier section dealing with the test procedure. The retardation clearly is
to be determined, but not necessarily by measuring it directly, it would seem. It would probably be
fair to say that one of the key objectives of the type examination is to measure the retardation
supplied by the buffer. Surely this cannot be done, at least with the required accuracy, without a
device for measuring retardation first-hand. Sensors, load cells or other instrumentation devices that
calculate deceleration from other measured parameters generally lose a certain degree of accuracy
as a result of these calculations in this author’s experience. Elevator buffers have to meet with a
variety of specifications but, surely, the most important of these is the manner in which the buffers
must bring an impacting elevator car to rest. Not measuring the retardation supplied by the buffer
directly during testing seems not to be conducive to finding out if the buffer meets with
specifications on bringing an impacting elevator car to rest.
I notice that in section 5.7.2 relating to the statement and test sample of the ascending car overspeed
protection means type examination, the standard says:
“As defined between the applicant and the laboratory:
- either a complete assembly consisting of both elements, braking device and

speed monitoring device; or
- only that device which was not subject to verifications according to 5.3, 5.4
and 5.6;
shall be provided by the applicant.” [2]
To put this into context, earlier in the standard, in section 5.7.1.2, it says:
“The applicant shall state the range of use provided:
a) minimum and maximum masses, or torque;
b) minimum (if applicable) and maximum rated speed;
c) use in installations with compensating ropes.” [2]

If the applicant has already had the braking device certified and, therefore, only the speed
monitoring element is the subject of the examination, surely it is no longer necessary for the
applicant to provide the maximum and minimum masses. Mass does not have any effect on the
operation of an overspeed protection means. The function of any excessive speed protection means,
taken in isolation, is to trigger a braking device upon detection of a certain speed, regardless of the
mass of the lift car. If the applicant was providing both elements, they would need to state the range
of masses because the mass does affect the performance of the braking device, but the same cannot
be said for the overspeed protection means alone.
The type examination of unintended car movement protection means is covered in section 5.8, and
section 5.8.1 concerns the method of the test:
“The unintended car movement protection means shall be type tested as a complete system
or the subsystems for detection, activation and stopping may be submitted to an individual
type examination.” [2]
The standard guidelines for this is phrased very differently compared to the equivalent standard for
overspeed protection means; the standard refers to three subsystems, for detection, activation and
stopping, which is understandable, because protection means against unintended movement needs
to incorporate a control circuit to detect movement with the car doors open, whereas the means for
protection against ascending car overspeed does not. However, what is not so clear is what the
standard for unintended movement means by “interface conditions between the subsystems if
integrated into a complete system”. [2]
Despite acknowledged critical differences between the two, protection means against unintended
movement and ascending car overspeed have, in essence, to perform fairly similar functions.
Protection means against unintended movement has to detect 150-200 mm of car movement with
doors open and stop it according to the guidelines laid out in the standard. Meanwhile, protection
means against ascending car overspeed has to detect excessive speed of an ascending car and stop
or slow the lift to such a speed for which counterweight buffers are certified. The relative similarity
of these two means is demonstrated by the earlier discussion in this report as to the difference
between excessive speed in the up direction and unintended car movement with doors open in terms
of the speed monitoring, speed reducing and prevention or stop devices allowable by the standard.
The conclusion was drawn that identical devices, namely an overspeed governor in tandem with a
rope brake or safety gear, can be used for both scenarios. The only differences are that for
unintended movement the means has to incorporate a control circuit to detect movement with the
car doors open before actuating of the stopping means when the doors are open. The fact that such
similar devices are interchangeable for dealing with the two situations speaks for itself.
The unintended movement standard also states that, among others, the minimum and maximum
fluid pressure, if applicable, and limits of temperature and humidity of the design and any other
relevant information agreed between the applicant and test laboratory shall be stated by the
applicant. None of this is mentioned in the section of the standard covering protection means
against ascending car excessive speed. Firstly, it is unclear to what fluid pressure the standard is
referring here and why it does not apply to the standard for ascending car overspeed. In addition,
what is meant exactly by the temperature and humidity of the design is not clear: does it refer to one
or more of the components, the lift car itself, the lift machine, or an amalgamation of these? The
fundamental differences between the two scenarios do not seem to merit such discrepancies in
standard guidelines between them.
Another difference between the two standards concerning unintended movement and ascending car
overspeed comes in the method of test section. The unintended movement section calls for
measurements to be made of the stopping distance, response time of the detection, actuation,
stopping element and control circuits, whereas the ascending car overspeed section does not. In
addition, a figure is provided showing the acceleration and deceleration of the lift car with response
times labelled, but no equivalent figure is given for the section of the standard covering ascending
car overspeed. Control circuits apart, it appears not to make sense that ascending car overspeed
protection tests do not require these measurements and figures. Whilst the two situations are
different with fundamentally different requirements, and unintended movement is a potentially
more dangerous and serious situation than ascending car overspeed, the extent of the inconsistency
between the two standards it is somewhat surprising. Surely, the detection device is a fundamental
aspect of the unintended car movement protection means and it would go without saying that, as it
does in the standard for the ascending car overspeed protection means, its operation is to be tested.
Another inconsistency crops up in the next two sections, dealing with devices certified for a single
mass, torque or fluid pressure (5.8.3.2.2 and 5.8.3.2.3). [2] Once again there is a mention of fluid
pressure, to which the same query raised earlier applies. Additionally, this title makes reference to
torque. Again, the reasons for including torque in this title are not clear. At what stage in the
detection, actuation and braking process there is torque involved for unintended car movement that
there is not for ascending car overspeed protection is a mystery. If this title is as it is, why is the
corresponding title in the ascending car overspeed section not “Device certified for a single mass or
torque” instead of “Device certified for a single mass”. [2]
Staying with this particular section of the standard (5.8.3.2.2), the standard defines that:
“The laboratory shall carry out 10 tests with the system mass or torque or fluid pressure
representing an empty car in up direction and 10 tests with the system mass or torque or
fluid pressure representing an empty car carrying the rated load in down direction.” [2]
The unintended car movement protection means is to act in both directions, whereas the ascending
car overspeed protection means only acts in the up direction, so the fact that this section of the
standard dictates tests are to be carried out both with the equivalent of an empty car in the up
direction and the rated load in the down direction is plausible. However, one disparity between the
two standards that is not is the fact that 10 tests are to be carried out for each, when only 4 are
carried out for the ascending car overspeed testing. That unintended car movement testing is
deemed to require six more tests being carried out in each direction than ascending car overspeed
testing is mysterious. What is also strange is that one set of friction parts must be capable of 5 tests
minimum here, in comparison to 2 or 3 (depending on speed) for ascending car overspeed.
Comparison of the two sections dealing with checking after the tests in the two standards yields
more apparently unnecessarily pronounced differences. Although points a), b) and c) appear to
roughly correspond to each other, the unintended movement standard has an extra aspect to it:
“d) it shall be checked that the retardation with the minimum mass has not exceeded 1
.” [2]
Again, the fact that is stated as a constraint for this standard but not for the device for protection
against ascending car overspeed appears illogical; surely this is a criteria that either applies to all
speed monitoring, speed reducing and prevention or stop devices, or none of them. It is clear that
unintended car movement and ascending car overspeed protection are fundamentally different
situations, and unintended car movement is a potentially more serious and dangerous situation than
ascending car overspeed, for one, because people may have been getting in and out of the lift when
it started to move. However, despite this, the nature and extent of many of the inconsistencies in the
standards is very strange and surprising.
4 CONCLUSION
Bearing in mind the key differences between unintended car movement and ascending car
overspeed, the devices that are allowable for each are slightly different accordingly, although they
need, in essence to perform similar functions of detecting the situation and stopping or slowing the
lift. To conclude, an overspeed governor that can detect 150 mm of movement with the car doors
open in tandem with a rope brake or safety gear is one example of a compliant protection means
against unintended car movement with car doors open, along with electronic control circuits, shave
brakes and, potentially, solenoids. Again, an overspeed governor in tandem with a rope brake or
safety gear, in addition to counterweight buffers is a compliant means for protection against
ascending car overspeed.
A mechanical linkage to the car to assist the means in its performance is permitted for ascending car
overspeed protection, and it would seem also for unintended car movement, as long as it meets
requirements called for elsewhere in the standard.
As well as discussing the implications brought about by the wording of the standard in certain
sections, as throughout the paper, section 3.2 mused what this means as far as testing and
certification of the compliant devices is concerned, which is covered in EN81-50. In addition to
this, details of test procedures were confronted and evaluated, such as factor of safety on
calculations, and first-hand measurement of deceleration. This analysis incorporated energy
dissipation and accumulation buffers as well as detection and stop devices.
The number of tests to be carried out on both ascending car overspeed protection (ACOP) and
unintended car movement protection (UCMP) means was also appraised. Arguments were put
forward regarding the number of tests carried out in relation to the number of tests one set of
braking parts should be capable of. This, in turn, brought up issues of friction between braking parts
and the braking surface; including how some safety gears are capable of many more tests than
called for by the standard, even over used sections of surface. Finally, the paper examined and
compared the testing procedures laid out in the standard for both ACOP and UCMP means, with
respect to the devices allowable by the standard for each, and questioned the, what it deemed,
relatively large disparities between them.
GLOSSARY
MRL: machine room-less lift
ACO: ascending car overspeed
UCM: unintended car movement
REFERENCES
[1] BSI Standards Ltd., BS EN81-20:2014 - Safety rules for the construction and installation of
lifts. Lifts for the transport of persons and goods. Passenger and goods passenger lifts. BSI
Standards Ltd., 2014
[2] BSI Standards Ltd., BS EN81-50:2014 - Safety rules for the construction and installation of
lifts. Lifts for the transport of persons and goods. Passenger and goods passenger lifts. BSI
Standards Ltd., 2014
[3] David G. Ullman. The Mechanical Design Process Second Edition. Available online:
http://www.maelabs.ucsd.edu/mae3/Assignments/Energy_Analysis/factor_of_safety/FactorOfSafet
yGuidelines-Ullman.pdf
Installing and Calibrating Loop Amplifiers to EN81-70 so that Test
Certificates can be Produced for Audio Frequency Induction Loop
Systems (AFILS) in Accordance with BS EN 60118-4
John Trett
C.E. Electronics Ltd., P.O. Box 1679, Marlow, Bucks, SL7 3ZG, UK
[email protected]
Keywords: Loop, Amplifier, Installation, Calibration, AFILS, EN81-70, BS EN 60118-4,

IEC6010118-4
Abstract. Service providers have to make "reasonable adjustments" to the physical barriers to gain
access in all buildings. The summary of main provisions for disabled access includes: f) Emergency
telephone and inductive coupler for hearing aid users. Inductive loop amplifiers need to be installed
and calibrated correctly so that test certificates can be produced for Audio Frequency Induction
Loop Systems (AFILS) in accordance with BS EN 60118-4. If they are not powerful enough or
incorrectly set up they produce distorted sounds. Often, installed systems are simply a loop behind
the Car Operating Panel (COP) with limited range, so they cannot be heard by a deaf person at the
other side of the car or collapsed on the floor. This leaves users with impaired hearing at a
dangerous disadvantage as they cannot hear normal and telephone lift messages. This paper
provides guidance on how existing loop amplifier specifications and installations can be improved.
1 INTRODUCTION
Passengers with impaired hearing are often not able to properly ‘hear’ both the emergency
telephone messages and the lift speech messages including emergency messages. Approximately 1
in 6 of the UK population (that is 10 million people!) have hearing loss and would benefit from
additional assistive devices to recognize spoken messages.
Those passengers wearing hearing aids will have a ‘Telecoil’ (T-coil) fitted in the aid (Figure 1),
which, if switched on, can pick up audio inductive signals and amplify them in the aid to the
passengers hearing requirements. Hearing aids amplify the volume and also compensate for the
loss of specific frequencies. The passenger knows to switch their ‘Telecoil’ ON when they see the
hearing loop fitted sign. (Figure 2)
Figure 1 ‘Telecoil’ in hearing aid Figure 2 Signage
The signal from the loop, if set up correctly, should improve on the signal picked up by the hearing
aid itself. It should not distort or clip the signal or it will not improve the signal and will often make
it worse.
An AFILS driver or induction loop system has a linear current amplifier of at least 2.2 amps rms
with a 1kHz signal and a voltage output of at least 4.5V peak. The input is connected to the speech
sources (telephone and lift messages). A low impedance loop coil is connected to the output, and
mounted on the perimeter of the area where the field is required, to generate an audio field in the lift
car. A passenger standing in this field picks up the audio signals. The equipment and design should
be sufficient so there is no clipping or distortion of the signal and have metal compensation.
A ‘Telecoil’ in the hearing aid is a small magnet and coil which vibrates and picks up these audio
signals. The signals are amplified by the hearing aid to the needs of the wearer, to compensate for
both volume and frequency loss.
The orientation of the loop field is important and should be at 90 degrees to the plane of the
‘Telecoil’ for the best results. Ideally the loop should be mounted horizontally, above or below the
passenger, or the ‘Telecoil’ will not be in a good field and the audio signal will be very weak.
(Figure 3)
Figure 3 Loop Magnetic field Hearing Aid ‘Telecoil’ in the magnetic field
2 INSTALLING LOOP AMPLIFIER CONSIDERATIONS

There are a number of ways loops have been installed in lift cars. The best way is to have a loop in
the ceiling or on top of the car as it gives a field in the whole car. Some manufacturers put a small
coil behind the COP. This will have limited range and power and will only be picked up by the
passenger who is near the COP and not collapsed on the floor or at the back of the car.
The installer should ask if the induction loop amplifier is to have the field coil on top of the car to
fill the whole car, or if the coil is to be mounted behind the COP to have a small field just in-front
of the COP only. The loop system should ideally have suitable input facilities to accept signals from
both the telephone system and the car speech system to give the passenger with impaired hearing all
the audio messages. This includes emergency speech messages such as fire recall and door closing
etc. Otherwise the passenger will be at a disadvantage in these situations. (Figure 4) (Figure 5)
The lift car is a metal box which will absorb the signal if the loop is mounted near the walls. This
“Faraday cage” effect hinders loop installations, so we require the loop amplifier to be located and
to have sufficient power and metal compensation for the loop to be mounted outside the lift car
metal box yet to allow the audio field signals to pass through. For best effect the loop is mounted on
the roof, about 150mm in from the car top edge. It should be above the passenger and no higher
than 1.8 metres from their ears. The loop wire containment conduit on top of the car should be of
non-metallic construction to avoid short circuit earth paths.
The loop electronics should be in an earthed metal box and mounted where it is not vulnerable to
mechanical damage and the controls can be accessed through the cover. It must be close to the
loop.
The connections should be of a type suitable for the application. For example, if the input is a
telephone system it should have the correct isolation and impedance match. If taken from the
speaker feed, it should be a twisted pair.
Installing and Calibrating Loop Amplifiers to EN81-70 so that Test Certificates can be Produced for Audio 23-3
Frequency Induction Loop Systems (AFILS) in Accordance with BS EN 60118-4
Figure 4 Basic Wiring
Figure 5 Car top loop fitting
3 CALIBRATING LOOP AMPLIFIERS

Manufacturer’s loop amplifier instructions include setting up and adjusting the input and output
signals and strengths of the loop electronics.
The signals from the audio sources should be set to the input threshold requirements of the loop
amplifier. If too weak the amplified background noise will come through. If too strong the signal
will be clipped and distorted. This should be resolved at source of installation.
The current through the loop should be set to give a field strength of 400mA/m sine wave at 1kHz.
If it is too high the hearing aid will be overloaded. If it is too low the signal to noise ratio is
reduced.
Standard IEC6010118-4 (BS EN 60118-4) is prepared by the International Electrotechnical
Commission (IEC). It provides a standard for system performance, and specifies the use of the T-
sign logo. It provides an expectation of quality.
Induction Loop Testers (Figure 6) are sophisticated field strength meters with digital displays and
menu selection and good quality headphones. They are designed to simplify the setup of an
induction loop system to the latest version of BS EN 60118-4 and to check the performance. The
tests are very comprehensive so that test certificates (Figure 8) can be produced for AFILS in
accordance with BS EN 60118.
Designed to test Magnetic field strength in audio-frequency induction loops for hearing aid
purposes, the kit includes an induction loop tester with intuitive display and simple to follow test
menus, a calibrated signal generator with pre-loaded test tones and a set of headphones.
As well as checking the magnetic field strength of an induction loop system, it also measures
amplified background noise, frequency response, metal compensation loss and also allows you to
listen to the loop signal.
Figure 6 Field strength meters
Acceptable coverage should be the whole area where passengers of different heights could be
standing or sitting in a wheelchair (or even collapsed on the floor). If total coverage is not possible
(e.g. with loops in the COP) then the hearing aid user needs to know where the loop is. Hearing
loss is a hidden disability and audio loop fields are not visible.
Amplified background noise should be -32dB or lower (A-weighted). Noise should not affect
intelligibility.
Metal degrades magnetic fields. A 2.5-amp current test in the centre of the loop should give 0db
loss with respect to a signal with no metal loss. E.g. if the signal is -6dB in the centre of the loop,
the metal loss equals 6dB.
The field strength should be tested at 400mA/m sine wave at 1kHz over the whole required volume.
If the signal is too high the hearing aid will be overloaded. If it is too low the signal to noise ratio is
reduced. The signal strength should not deviate more than ±3 dB over the listening area, i.e. the
signal should be consistent from floor to head height and over the whole floor area.
The frequency response should be within ±3dB from 100Hz to 5kHz with reference to the signal at
1kHz. Factors such as effective drive current and metal structures will affect this.
Listener headphone receivers will not do these tests and cannot be relied on for signal quality.
Signage is essential so the user does not have to ask. Signs must be clearly visible. (Figure 7)
Figure 7 Signage
Figure 8 Typical test certificate produced on installation

4 STANDARDS SUMMARY
As part of the performance standard the system must meet standards for:
 Low amplified background noise

 The correct field strength
 Even field strength
 Flat frequency response
Just as critical are:
 Input audio quality to separate signal from noise

 Acceptable coverage
 Clear signage, no user request necessary
 Training, monitoring and maintenance.
Site testing should be an integral part of achieving standard compliant loop systems. Contractors
need to plan and manage and train their employees and provide the information necessary to
comply.
5 MAINTANANCE OF LOOP AMPLIFIERS

Ignorance is the most cause of loop failure and incorrect installation. Staff must be trained to test
the system and help customers. Loop systems must be regularly monitored. This includes staff
access to the monitoring equipment and regular maintenance and testing by trained staff.
Site testing should be an integral part of routine maintenance procedures.
1) Regularly (monthly) check the signal. This can be done using an audio listener headphone
receiver. This should be held VERTICALLY to be the same orientation as the “Telecoil” in
the hearing aid. (Figure 9)
2) Annually the system should be checked for quality of sound using the full field strength
meter calibration again. (Figure 6)
Figure 9 Audio headphone induction loop receiver

6 CONCLUSION AND ALTERNATIVE TECHNOLOGIES

This paper has looked at current technology and given guidance at how to apply it well. When
installing loop amplifiers, it is really important that they are set up and calibrated so the passenger
can understand the audio signal clearly anywhere in the lift car.
The draft revision of EN 81-70 recently out for public comment is based on induction loops in all
lift cars (the current standard has this subject to negotiation – i.e. provided when agreed / specified);
however, this has been very heavily commented.
There are other solutions which have not been reviewed in detail as the technology is more recent
and not yet so widely available. However, consideration has been given to other technologies
including speech to text recognition on mobile devices using Bluetooth or similar. This would be
restrictive for passengers with hearing loss as this technology is not as universally available as is a
“Telecoil” in hearing aids. Another solution could be to have a speech to text screen in the car
taking its inputs directly from the telephone and lift speech systems.
The author is grateful for the input and literature from the following loop system manufacturers
Ampetronic
C.E. Electronics
Contacta
Deaf Alerter
SigNET AC
REFERENCES
Standards IEC6010118-4; BS EN 60118-4; EN81-70
John Trett is the Managing Director of C.E. Electronics Ltd in the UK. They make and supply
many electronic ‘signage’ devices for lifts to give passengers information, including indicators, TFT
screens, speech units and induction loops. John has an electrical engineering degree BSc (Eng)
from Nottingham University and other electronic and lift qualifications. He was trained by Otis.
John became severely deaf from a virus 3 years ago so needs induction loops to be set up correctly
to understand what is going on. He has a personal interest in promoting good installations.

Liftand Escalator Symposium Proceedings 2016

Uploaded by

Copyright:

Available Formats

Liftand Escalator Symposium Proceedings 2016

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Liftand Escalator Symposium Proceedings 2016

Uploaded by

Copyright:

Available Formats

6TH SYMPOSIUM ON

LIFT & ESCALATOR

Dr Jonathan Adams, The University of Northampton

Professor Lutfi Al-Sharif, The University of Jordan, Jordan

Professor Stefan Kaczmarczyk, and Dr Richard Peters

Towards a Systematic Methodology for the Design of Elevator Traffic Systems in 2

Lifts in Health: Health Technical Memorandum 08-02 Revisited 3

Your Lift Journey – How Long Will You Wait? 5

Evaluating a Holistic Energy Benchmarking Parameter of Lift Systems by using 6

The Report of Thorough Examination as a Management Tool for Maintenance 7

London Underground Escalator Passenger Safety Strategy Improving Passenger 10

Pilot for Standing on Both Sides of Escalators 11

An Overview of India, Travelling Tall 12

How Current Technology Trends are Empowering Us All to Drive Innovation 13

Modelling of a Rope-Free Passenger Transportation System for Active Cabin 15

Dynamic Lift Control for Improvements in Energy Efficiency 16

Fire Lifts, Escalators & Moving Walks Management System (FEMS) in an 17

Map-Based Active Compensation of Lateral Vibrations in Elevators 19

The Role of Economic Factors in Traffic Planning and Selection of Lift 20

Understanding the Requirements of the New EN-81 Standards with Respect to 22

Installing and Calibrating Loop Amplifiers to EN81-70 so that Test Certificates 23

𝜕(𝜌𝑢𝑖 ) 𝜕(𝑢𝑖 𝜌𝑢𝑗 ) 𝜕𝑝 𝜕𝜏𝑖𝑗

3 THE MODEL DESIGN AND COMPUTER SIMULATION

4 RESULTS AND DISCUSSION

Figure 3 The time sequence of the relative position of each lift

𝜌: the air density (1.2041 𝑘𝑔𝑚−3) (at 20 °𝐶)

𝑉: the lift car velocity (20 𝑚𝑠 −1 )

𝐿: the lift car width (2 𝑚)

𝜇: the dynamic viscosity of air (1.81 × 10−5 𝑘𝑔𝑚−1 𝑠 −1)

(2) Professor Stefan Kaczmarczyk

Lutfi Al-Sharif1, Ghazi Al Sukkar2, Asma' Hakouz2, Nama' A. Al-Shamayleh1

• The number of floors in a single zone should not exceed 18 to 20 floors.

2 SECTORS, ZONES AND STACKS

3 OVERVIEW OF EXISTING WORK ON HIGH RISE BUILDING

Another rule uses natural language descriptions:

Howkins [16] then:

4 GUIDANCE TO THE DESIGNER RATHER THAN AUTOMATED DESIGN

5 EFFECT OF DESTINATION GROUP CONTROL

6 PARAMETERS AND NORMALISATION

• Equal elevators in each zone (for symmetry).

Number of floors above Net area: Area taken up by elevators

Key words: lifts, healthcare, hospitals

1 HOW THE DEPARTMENT OF HEALTH MANAGES ITS ESTATE

Figure 1: Whole building lifecycle

Figure 2: Health specific documents

Table 1: Health Technical Memoranda

HTM00 Policies and principles

2 THE HEALTH TECHNICAL MEMORANDA

HTM 08-01 [Specialist services] Acoustics

3 THE HISTORY OF HTM 08-02

4 THE PROBLEM OF REVISING HTM 08-02 IN EARLY 2016

Important note to users of this edition of HTM 08-02

5 THE CONTENTS OF HTM 08-02: 2016

6 PARTICULAR CONSIDERATIONS IN THE HEALTHCARE ENVIRONMENT

Example 2: Medical assistance

Example 3: Special access controls

Example 4 Emergency bed service (code blue control)

7 PROFESSIONAL ROLES AND RESPONSIBILITIES

Designated Persons (Lifts)