Available online at www.sciencedirect.

com
Available online at www.sciencedirect.com
ScienceDirect
Available ScienceDirect
online
Availableatonline
www.sciencedirect.com
at www.sciencedirect.com
Procedia CIRP 00 (2018) 000–000

ScienceDirect
Procedia CIRP 00 (2018) 000–000 www.elsevier.com/locate/procedia
ScienceDirect www.elsevier.com/locate/procedia

Procedia CIRP 00 (2017)

Procedia CIRP 000–000
78 (2018) 61–66
www.elsevier.com/locate/procedia

6th CIRP Global Web Conference

“Envisaging the future manufacturing, design,Web
6th CIRP Global technologies and systems in innovation era”
Conference
“Envisaging the future manufacturing, design, technologies and systems in innovation era”
Novel method
28th CIRP to systematically
Design Conference, implement
May 2018, Nantes, France
Novel method to systematically
lean production in machining areas implement
A new methodologylean to analyze
production the functional
in machining andareas
physical architecture of
a, b a a
Stefan Seifermann
existing products*, Jörg
forBöllhoff
an , Carsten Schaede
assembly oriented , Michael
productKutzen
familyand Joachim Metternicha
identification
a, b a a a
Stefan Seifermann *, Jörg Böllhoff , Carsten Schaede , Michael Kutzen and Joachim Metternich
Institute of Production Management, Technology and Machine Tools (PTW), Techniche Universität Darmstadt,
a

InstitutePaul StiefManagement,
of Production
a
*, Jean-Yves Dantan,
Otto-Berndt-Str.2,
Technology 64287
and Alain
Darmstadt,
Machine Etienne,
Tools Germany
(PTW), AliUniversität
Techniche SiadatDarmstadt,
Herrmann Ultraschalltechnik GmbH & Co. KG, Descostraße 3-11, 76307 Karlsbad, Germany
b
Otto-Berndt-Str.2, 64287 Darmstadt, Germany
Écoleauthor.
* Corresponding Nationale
Tel.:Supérieure
b d’Arts
Herrmann et Métiers, Arts etGmbH
Ultraschalltechnik
+49-6151-1620081; Métiers
fax: +49-6151-1620087; ParisTech,
& Co. LCFC [email protected]
EA 3-11,
KG, Descostraße
E-mail address: 4495,76307
4 Rue Karlsbad,
Augustin Fresnel,
GermanyMetz 57078, France
* Corresponding author. Tel.: +49-6151-1620081; fax: +49-6151-1620087; E-mail address: [email protected]
* Corresponding author. Tel.: +33 3 87 37 54 30; E-mail address: [email protected]

Abstract
Abstract
The major goal of lean production is implementing continuous flow production value streams according to just-in-time principles, having one-
Abstract
piece-flow
The as anofideal.
major goal The implementation
lean production is implementingof suchcontinuous
systems isflow quite widespread
production in streams
value assemblyaccording
areas, whereas implementations
to just-in-time principles,for machine-
having one-
Inintensive
today’s manufacturing
piece-flow business
as an ideal. processes
environment,
The theare lacking
trend
implementation ofbehind
towards suchmoredue to a considerably
product
systems is variety andhigher
quite widespread technical
customization iscomplexity.
in assembly unbroken. A
Due
areas, whereas simple transfer of the methods
to implementations
this development, used
themachine-
for need of
for
agile creating
intensive flow in assembly
and reconfigurable
manufacturing areas are
production
processes issystems
not possible.
lacking emerged
behind In order
to totoaenhance
duecope with lean production
various
considerably products andimplementation
higher technicalproduct forAmachining
families.
complexity. To design
simple areas, ofa novel
and optimize
transfer method
production
the methods has
used
been
systems developed
as well
for creating tointo
flowas transform
choose batch
assembly the
areas size-oriented
optimal
is not product
possible.machining
In order into
matches, synchronous
product
to enhance analysis one-piece
methods are
lean production flow machining.
needed.
implementation It comprises
Indeed,
for most of the
machining a areas,
six-phase
known approach
methods
a novel based
aim
method to
has
on
beenexisting
analyze practices
a product
developed in product
lean batch
toortransform
one production.
family onThe
the approach
size-oriented physical
machining starts
level. with generating
Different
into product
synchronous process understanding
families,
one-piece however,
flow may
machining. and Itremoving
differ largelybarriers
comprises in termsforofflow
a six-phase the production,
number
approach and
based
and
nature then
on existing addresses
practices process
of components. inThis
leanfactstabilization
impedesThe
production. anbefore
efficient
approach establishing
comparison
starts with thegenerating
prerequisites
and choiceprocessfor understanding
a continuous
of appropriate product material
family
and flowbarriers
via group
combinations
removing technology
forflow
for and
the production,
production
manufacturing
system.
and thenA new cells. Afterwards,
methodology
addresses information
processisstabilization
proposed to analyze flow is facilitated
existing products
before establishing and man-machine
in view of their
the prerequisites interaction
forfunctional is
a continuous designed,
and physical before flexibility
material architecture.
flow via group towards
The aim changes
is to cluster
technology andin
customer
these demand
products in new are implemented.
assembly oriented The method
product has
families been
for validated
the both
optimization inofthe Process
existing Learning
assembly Factory
lines and
manufacturing cells. Afterwards, information flow is facilitated and man-machine interaction is designed, before flexibility towards changes in CiP
the at TU
creation Darmstadt
of future as well as
reconfigurable in
several
assembly
customer industrial
demandmanufacturing
systems. Based on Datum
are implemented. companies.
Flow
The Chain,
methodthe hasphysical structureboth
been validated of the
in products is analyzed.
the Process LearningFunctional
Factory CiP subassemblies
at TU Darmstadt are identified,
as well as andin
aseveral
functional analysis
industrial is performed.
manufacturing Moreover, a hybrid functional and physical architecture graph (HyFPAG) is the output which depicts the
companies.
© 2018 The
similarity Authors.
between Published
product by Elsevier
families B.V. This
by providing designis an open access
support to both,article under the
production CC BY-NC-ND
system planners and license
product designers. An illustrative
© 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(https://creativecommons.org/licenses/by-nc-nd/4.0/).
example
© 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND licensefamilies of steering columns of
of a nail-clipper is used to explain the proposed methodology. An industrial case study on two product
(https://creativecommons.org/licenses/by-nc-nd/4.0/)
Selection
thyssenkrupp andPresta
peer-review under
Franceunder responsibility
is thenresponsibility
carried out toof
(https://creativecommons.org/licenses/by-nc-nd/4.0/).
Selection and peer-review ofgive
the ascientific
the scientific committee
first industrial
committee of the
the 6th
evaluation
of 6ththe
of CIRP
CIRP Globalapproach.
proposed
Global WebConference
Web Conference“Envisaging
“Envisagingthe thefuture
future
manufacturing,
©manufacturing,
2017 Theand
Selection design,
Authors.
peer-review
design, technologies
Published and systems
systems
underbyresponsibility
technologies Elsevier
and B.V.of inin innovation
theinnovation era”.
scientific committee
era”. of the 6th CIRP Global Web Conference “Envisaging the future
Peer-review under
manufacturing, responsibility
design, technologiesof theandscientific
systemscommittee
in innovation of the 28th CIRP Design Conference 2018.
era”.
Keywords: lean production, machining, continuous flow production
Keywords:Assembly;
Keywords: Designmachining,
lean production, method; Family identification
continuous flow production

1. Introduction economic publications to date [3]. Implementing the principles

1.1.Introduction
Introduction of
of leanproduct
the
economic production
range enables
publications and continuous
characteristics
to date flow production
manufactured
[3]. Implementing and
and/or
the principles
The principles of lean production nowadays are widespread assembled helps
of lean to production
reduce
in thiswaste
system. in In
enablesproduction
this areas.
context,
continuous the
flowIn assembly
main challenge
production areas,
in
and
among
Due different
to the industries
The principles fast
of and all over
leandevelopment
production inthethe
nowadays world.
aredomain With the
widespread of lean
helpsproduction
modelling methods
and analysis
to reduce waste and
nowcontinuous
inisproduction
not only flow
areas.to production
Incope with single
assembly are
areas,
Toyota different
among Production
communication and System
an ongoing
industries (TPS)
and [1]
overbecoming
alltrend of world.famous
the digitization Withand by products,
the quite production
lean widespread.
a limited Particularly
product
methods range
and in the automotive
or existing
continuous flow product industry,
families,
production are
giving guidance
digitalization,
Toyota Production onSystem
manufacturinghow to (TPS) achieve
enterprises superior
are
[1] becomingfacing results
important
famous for
by but processes to are systematically
alsowidespread.
quite be able analyze improved
to Particularlyand to thebased
in compare on lean principles.
products
automotive to define
industry,
manufacturing
challenges
giving guidance quality
in today’s on andhow throughput-time,
market environments:
to achieve researchers
superior a continuing
results began
for new However,
product
processes arein machining
families. It can be
systematically areas,
observed
improved the implementation
that
based classical
on existing
lean principles. of
to scientifically
tendency towards
manufacturing analyze
reduction
quality the throughput-time,
and different
of product aspects of
development lean production.
researcherstimesbegan
and continuous
product
However, flow
families
in are production proves
regroupedareas,
machining to
thebeofimplementation
in function rather
clientsdifficult [4].
or features. of
“The
shortened
to Machine
product
scientifically That
analyze Changed
lifecycles. In the
the different World”
addition,
aspects [2]
thereof isbyanWOMACK,
lean increasing
production. Due to a higher
However,
continuous flowtechnical
assembly oriented
production complexity,
product tosignificant
proves families machine
are hardly
be rather set-
to find.
difficult [4].
ROSSMachine
demand
“The andcustomization,
of JONES brought
That Changed beingthe worldwide
at World”
the same attention
[2]time in atoglobal
by WOMACK, lean upOn
Due times
tothe and inevitable
product
a higher machine
family level,
technical downtime,
products
complexity, lean
differ mainly
significant methods
machinein two
set-
production
competition
ROSS andand
with iscompetitors
JONES considered
broughtto be onethe
allworldwide
over of the mostThis
world.
attention influential
totrend,
lean main cannot
up simply be transferred
characteristics:
times (i) the machine
and inevitable numberto this area. Furthermore,
ofdowntime,
components andmethods
lean the
(ii) the
which is inducing
production the development
and is considered to be onefrom of themacro to micro
most influential type of components
cannot (e.g. mechanical,
simply be transferred to thiselectrical, electronical).the
area. Furthermore,
markets,
2212-8271 results in diminished
© The Authors. Publishedlot sizes due
by Elsevier B.V.toThis
augmenting
is an open access Classical
article undermethodologies
the CC BY-NC-ND considering
license mainly single products
product varieties (high-volume to low-volume production) [1].
2212-8271 © The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license families analyze the
(https://creativecommons.org/licenses/by-nc-nd/4.0/). or solitary, already existing product
Selection
To cope and
withpeer-review under responsibility
this augmenting variety ofasthewell
(https://creativecommons.org/licenses/by-nc-nd/4.0/). scientific
as tocommittee
be able oftothe 6thproduct
CIRP Global Web Conference
structure “Envisaging
on a physical level the future manufacturing,
(components design,
level) which
technologies and systems in innovation era”.
identify possible
Selection and peer-reviewoptimization potentials
under responsibility
doi:10.1016/j.procir.2017.04.009
in the
of the scientific existing
committee of the 6thcauses difficulties
CIRP Global regarding
Web Conference an the
“Envisaging efficient definition design,
future manufacturing, and
technologies system,
production and systems in innovation
it is importantera”. to have a precise knowledge comparison of different product families. Addressing this
doi:10.1016/j.procir.2017.04.009
2212-8271 © 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(https://creativecommons.org/licenses/by-nc-nd/4.0/)
Selection and
2212-8271 peer-review
© 2017 underPublished
The Authors. responsibility of theB.V.
by Elsevier scientific committee of the 6th CIRP Global Web Conference “Envisaging the future manufacturing,
design, technologies
Peer-review and systems
under responsibility of in
theinnovation era”.
scientific committee of the 28th CIRP Design Conference 2018.
10.1016/j.procir.2018.09.059
62 Stefan Seifermann et al. / Procedia CIRP 78 (2018) 61–66
2 Stefan Seifermann, Jörg Böllhoff, Carsten Schaede, Michael Kutzen and Joachim Metternich / Procedia CIRP 00 (2018) 000–000

machinery in use requires considerably high investments, Based on the preferred layout of lean manufacturing cells, these
which leads to a focus on a maximized degree of capacity are often also called “U-cells”. [12]
utilization. As a consequence, batch production is commonly
used. [5] 3. Introduction and description of the novel method
One possible approach to integrate lean production
principles in machining areas is cellular manufacturing, which Based on the goals defined in section 2, the transformation
significantly increases flexibility [6]. Previous research on is distinguished in small, iterative steps grouped into six phases.
cellular manufacturing mainly focused on product clustering Each phase can be considered as a stand-alone phase and is
[7,8], automation [9] and economic aspects [10]. A holistic performed once at a time. During the transition, the lean
approach for the implementation of lean principles in machining cell serves as an overall target state.
machining areas via manufacturing cells has not been Approaching the target state step by step, it might be
developed yet. Hence, this paper provides a method in order to necessary to adapt the transformation process implicating
systematically implement continuous flow production in changes in requirements and in general conditions. Hence the
machining areas. method encourages to return to an already completed phase to
Starting with the identification of different goals of lean repeat it in accordance with the altered requirements. Dealing
machining, key elements for successful continuous flow with the arising challenges, skills are developed throughout the
production in machining areas are identified. Based on these workforce and a lean mindset can evolve.
elements, a six-phase approach addressing all goals is In addition to the six consecutive phases, the method is
developed and introduced. Following the systematic approach, constructed in accordance with basic requirements (phase 0)
every phase is defined and exemplified. In the final part of the and comprises an element for continuous set-up time reduction,
paper, the testing and validation in the learning factory at TU that accompanies the phases 4 and 5. Fig. 1 gives an overview
Darmstadt as well as in various industries is described. on the phases of the method. They are explained briefly in the
following sections.
2. Definition of goals for lean machining

Transforming a job shop environment with large batch sizes

into lean machining cells working in one-piece-flow is a time-
consuming and challenging task. Beside huge planning effort,
the transfer from a capacity-utilization driven job shop into a
takt-controlled, synchronous production requires a paradigm
shift throughout the entire organization. This complicates or
even hinders the successful transformation. In order to facilitate
the transition, the presented method aims at continuous
adaptions in different phases and – even allowing halts in
between – therefore has a higher prospect of success.
In order to define different phases in a sensible way, the
approach is taken to first identify independent, complementing
goals that can then be addressed by the phases one after the
other. The goals are:
1. Establish knowledge on the processes and equipment used
in the machining area that is considered for the lean
Fig. 1. Phases of the introduced method
implementation
2. Stabilize the processes with regard to product quality and
reliability
3. Establish the prerequisites for continuous flow production 3.1. Phase 0: Basic requirements for lean machining
with the vision of one-piece-flow (focus: material)
4. Link the processes to customer takt time and facilitate the For the implementation of lean production in machining
control (focus: information) areas, a specific level of organizational standards, deviation
5. Improve manual work for the operator (focus: personnel) management and basic stability is crucial. Basic requirements
6. Establish flexibility regarding capacity comprise both, worker involvement and an adequate level of
If all goals are addressed, a lean manufacturing cell as it is worker qualification [13]. Workers must be able to perform
typically described in literature will be the result. According to their tasks independently according to their job requirements.
HYER and WEMMERLÖV [11], a manufacturing cell is “a Based on the existing practice of shop floor management [14],
group of closely located workstations where multiple, routines for lean machining can be developed. For instance, it
sequential operations are performed (…)”. For lean is crucial to track and discuss key performance indicators (KPI)
manufacturing cells in particular, the material flow is directed on the shop floor in order to be aware of potential gaps between
and synchronized with customer demand by a given takt time. actual situation and target state. Afterwards, continuous
 Stefan Seifermann et al. / Procedia CIRP 78 (2018) 61–66 63
Stefan Seifermann, Jörg Böllhoff, Carsten Schaede, Michael Kutzen and Joachim Metternich / Procedia CIRP 00 (2018) 000–000 3

improvement activities need to be launched to ensure a stable processes that can perform without interruptions caused
sustainable implementation of lean production. Workers must by quality problems, unplanned downtime or insufficient
be involved in problem solving activities on the shop floor in material supply. The required process stability has to be
order to further develop expertise and competencies. checked before proceeding to the next phase.
Furthermore, orderliness and cleanliness need to be established
and sustained (e.g. by using 5S as a method to eliminate basic
waste at workplace level [1,13]). Applying the proposed 3.4. Phase 3: Prerequisites for continuous flow production
method, standards will be extended and refined continuously.
After basic stability regarding quality and reliability have
been achieved, first efforts towards a lean material flow can be
3.2. Phase 1: Generate knowledge of the process tackled. The driving vision of all these efforts is the continuous
one-piece-flow. However, due to technical or organizational
In order to optimize machining processes and to apply lean reasons, the vision might never be reached and a one-set flow
principles in machining areas, a sufficient knowledge of both, is rather preferred [17]. Creating the prerequisites for a
products and processes is crucial. To gain knowledge on a continuous flow production can be subdivided into the
manufacturing process, a detailed analysis of the process is following six successive steps:
necessary. This comprises an in-depth engagement in all (i) grouping products to families
technical and organizational processes used in the area (ii) dedicating product families to machines
considered for the transformation towards a lean machining (iii) setting up a clear and directed material flow
cell. (iv) dedicating equipment to one product at the same time
Beyond, relevant information like process capacity, overall (v) designing new equipment “right-sized”
equipment effectiveness (OEE) and customer takt time (CTT) (vi) locating the machining equipment in close proximity to
need to be determined. Whereas process capacity is a measure each other
used to express the output of a process, e.g. 25 units/shift, the Grouping products to families, also known as “group
OEE provides an indicator on the effectiveness of the process, technology”, is a long discussed topic in manufacturing theory
e.g. how much time it actually took to manufacture the 25 units and practice. Its basic idea is the classification and
compared to how much time it could have taken in an optimal identification of products that are similar regarding geometry
situation. Therefore, the OEE facilitates the identification of or manufacturing sequence [18, 19]. One early and popular
major losses caused by machine downtime, reduced speed, set- example for a geometrical classification is OPITZ’s geometry
up and defects. It is calculated as [15] index [20]. For identifying similarities in manufacturing
sequences, work processing documents need to be reviewed.
𝑂𝑂𝑂𝑂𝑂𝑂 =
𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔 𝑢𝑢𝑢𝑢𝑢𝑢𝑢𝑢𝑢𝑢 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 ∗ 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 𝑝𝑝𝑝𝑝𝑝𝑝 𝑢𝑢𝑢𝑢𝑢𝑢𝑢𝑢
(1) After this initial step, a number of designated product families
𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡
is defined, and each product to be manufactured is assigned to
one family.
The CTT is defined as [1]
Dedicating product families to machines now defines, which
𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 𝑛𝑛𝑛𝑛𝑛𝑛 𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑁𝑁𝑁𝑁𝑁𝑁 of the identified product families are allocated to which
𝐶𝐶𝐶𝐶𝐶𝐶 = = (2) machining equipment in an optimized way. In order to solve
𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 𝐶𝐶𝐶𝐶
this issue, a large number of methods have been discussed,
This parameter describes how much time is available for spanning from intuitive, experience-based methods, setting-up
producing one piece of finished product. Knowledge of the machine-component matrices, cluster-analyses or heuristics up
customer takt is crucial for the streamlining of processes. to computer-based mathematical optimization algorithms (for
Before carrying on to the next phase, it has to be checked an overview see [21] or [22]). Especially higher developed and
whether the process knowledge is sufficient and both, potentials complex methods seem to be rather driven by academia and are
and needs of action have been identified. rarely used in industrial practice [8].
Once the product-equipment-allocation is finalized
(dedication of product families), directed material flows from
3.3. Phase 2: Process stabilization one machine to the next need to be set up. This requires that
machine sequences have to be fixed. In order to create
Continuous flow production requires reliable and stable transparency and clarity in the process flow, it is encouraged to
processes. Hence, process quality, availability of equipment as avoid backtracking and to ensure a one-directional flow.
well as supplier quality and supplier reliability have to be The next goal is to ensure also a dedication of the equipment
stabilized. By refining standards, unexpected performance to single product families or to single products manufactured at
losses can be reduced as quality can be improved. Also the the same time. This means an abdication of machines being
implementation of total productive maintenance (TPM) [16] shared by various value streams simultaneously. Different,
enhances the equipment availability and has positive impacts independent value streams can never be aligned and controlled
on the OEE. It is essential to treat internal suppliers just as in such a way that a smooth utilization of shared resources can
external suppliers. The declared goal of this phase is to ensure be guaranteed. Sooner or later, conflicting demands from the
64 Stefan Seifermann et al. / Procedia CIRP 78 (2018) 61–66
4 Stefan Seifermann, Jörg Böllhoff, Carsten Schaede, Michael Kutzen and Joachim Metternich / Procedia CIRP 00 (2018) 000–000

different value streams will arise, leading to delays and thus, to 3.5. Phase 4: Cell control according to customer takt time
inventory. However, in some cases, the set-up of equipment
just for one value stream at a time might not be the most While phase 3 established the physical prerequisites of a
suitable approach, e.g. if utilization rates are very low (and thus continuous flow production in the lean machining cell, the goal
waiting times of products are minimal) or if the equipment is of phase 4 is to set up the actual control of the material flows.
costly and providing dedicated equipment would create The determining parameter is the CTT (see formula (2)). All
significant investments. In these cases, using shared equipment single processes in the machining cell (including the time for
will still be the cost-optimal solution. manual operation) necessarily need to be aligned to exactly this
If investments in new machinery are considered, “right- time. In case the operational time of any of the processes for
sized”-equipment approaches are mandatory. “Right-sized” machining one part exceeds the CTT, the planned output for a
comprises both, technological/dimensional adaptation as well time period is not realistically achievable. Then, and following
as adaptations regarding maximum required processing times. the determining parameters of formula (2), either customer
The guiding idea should be driven by the actual need and not orders can be shortened or the available capacity can be
by technological feasibility or desires. Therefore, a clear increased (e.g. via expanded operation times). Another option,
statement of technical requirements is necessary in order to especially valid when change-over times (COT) for product
execute a best-value-option-analysis. Potential future changes are excessively high, is adjusting lot sizes (LS) in a
developments of products (e.g. larger dimensions, higher way that the available capacity on each machine i is balanced
accuracy, longer processing times) should be considered for with the required capacity of the products being manufactured:
sustainability reasons but should not be overestimated. “Right-
𝑁𝑁𝑁𝑁𝑁𝑁 − 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝑚𝑚𝑎𝑎𝑐𝑐ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡
sized”-equipment in combination with dedication of the # 𝑜𝑜𝑜𝑜 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠 = (4)
𝐶𝐶𝐶𝐶𝐶𝐶
equipment to single product families often allows the purchase
of cost-efficient, basic equipment. The minimum number of 𝐶𝐶𝐶𝐶
𝐿𝐿𝐿𝐿 = (5)
machines required can be calculated as [23] # 𝑜𝑜𝑜𝑜 𝑐𝑐ℎ𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎−𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜

∑
𝑛𝑛
𝑡𝑡𝑜𝑜𝑜𝑜,𝑖𝑖 + ∑
𝑛𝑛
𝑡𝑡𝑡𝑡𝑡𝑡,𝑖𝑖 Fig. 3 gives an example of a feasible combination for a cell
# 𝑜𝑜𝑜𝑜 𝑚𝑚𝑚𝑚𝑚𝑚ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 𝑖𝑖=1 𝑖𝑖=1
(3) with three machines.
(𝐶𝐶𝐶𝐶𝐶𝐶∗𝑦𝑦)−𝑡𝑡𝑚𝑚𝑚𝑚𝑚𝑚

with 𝑡𝑡𝑜𝑜𝑜𝑜,𝑖𝑖 stating the time for machining operations at

machine i, 𝑡𝑡𝑡𝑡𝑡𝑡,𝑖𝑖 stating the time for automated tool changes, y
as buffer factor towards CTT and 𝑡𝑡𝑚𝑚𝑚𝑚𝑚𝑚 for the manual time to
load and unload a machine. The assignment of the machining
operations to the machines of the cell can be done using the
time diagram and a precedence graph indicating the sequence
of the machining operations. A method for process planning in
cellular manufacturing can be found in [23].
Finally, for eliminating waste for transport and for creating
maximum transparency, the machining equipment for each Fig. 3. Exemplified time diagram for a machining cell with three machine
product family is located in close proximity to each other. For tools
this purpose a material flow matrix can be applied. All
preceding steps would also allow a “virtual” continuous flow The actual material flow in the cell is organized from the last
production. Only after this step, real physical continuous flow machine in the cell backwards with ideally no work-in-progress
production cells are at hand. Fig. 2 shows a comparison of stocks (WIP) in between the single processes. Only if the
material flows before and after locating the machinery one- successive machine in the cell is free to receive a part from the
directed according to the material flow and in close proximity. preceding one, the machining operation will be started. If
holding zero WIP is not reasonable, e.g. due to long change-
over times or fluctuating operating times of different variants
balancing out within certain periods, pull-systems with both,
little and defined stocks should be installed. Typical examples
are first-in-first-out- or supermarket-racks [15]. The former are
(a) (b) preferred due to the shorter waiting time of idle products. This
Fig. 2. Material flows before (a) and after (b) locating the equipment one- procedure requires an informational link between the
directed and in close proximity – industrial example consecutive machines in the cell as well as between the cell and
its environment, e.g. via Kanban-cycles [24].
The six successive elements within phase 3 described above
might be undertaken iteratively for optimized results and
adjustment due to changed conditions. 3.6. Phase 5: Improve manual work for the operator

After the material and the information control flow have

been designed, the next phase is focusing on the manual work
 Stefan Seifermann et al. / Procedia CIRP 78 (2018) 61–66 65
Stefan Seifermann, Jörg Böllhoff, Carsten Schaede, Michael Kutzen and Joachim Metternich / Procedia CIRP 00 (2018) 000–000 5

in the cell. Until this point, manual operations in the lean For long term adaptions of cell output, the machining
machining cell are out of scope. As a result, one worker might capacity has to be altered. Therefore, a machine can be
be dedicated to each process or machine, representing the worst extracted from the cell with a decreased cycle time and be
case. In phase 5 now, the manual work is adapted. Therefore, added to the cell with increased cycle time (rebalancing). Fig.
two areas of design are important: 4 shows the adaptability of a lean machining cell in terms of
(i) training and qualification of operators for competencies changing customer demand. The distribution of machining
in more than one process, ideally in all processes in the operations has to be adapted to the new number of machines
cell (multiskilled operators) according to formula (3).
(ii) adapted (lean) automation of the cell for optimized
operation
With regard to (i), multiskilled operators are the prerequisite
for being able to align the capacity of personnel to the overall
amount of manual work required for operating the cell
(independent of the nature of the tasks to be performed) and
thus reducing waiting time of the operator to a minimum [25].
Training and qualification of operators in multiple skills are
established and fostered via a rigid job rotation system, ideally
including functional changes several times a day [24].
Part (ii) of phase 5 focusses on the content of the task and
deals with the question, which operational functions should be Fig. 4. Adaptation of a lean machining cell to changed demand
automated and how. In line with the overall philosophy of lean
production, operational tasks are initially analyzed for waste
before any automation efforts [24]. A detailed and 3.8. Parallel phase: Improve change-over times
recommended approach for an adapted lean automation with
optimum results has been developed by SEIFERMANN [26]. The reduction of change-over times is an iterative and
After these optimization efforts and assuming that manual ongoing task. Short change-over times are essential to enable
work and automatic machine work in the cell are decoupled to synchronous production through lot size reductions. The
a certain degree [27] (i.e. that the operator does not need to overall aim is to reduce change-over times down to zero in
supervise a machine every second – a state that is achieved after order to enable a one-piece-flow with a lot size of one.
phase 2 and is easy to realize given the high degree of Achieving this goal is a time consuming task and might be
automation of today’s machine tools), the optimum number of impossible for particular cases. Therefore, the phase of
multi-skilled workers in the cell can be calculated as reducing change-over times runs in parallel to the phases 4 to
6. For a systematic reduction of change-over time, four basic
𝑡𝑡𝑚𝑚𝑚𝑚𝑚𝑚,𝑖𝑖 +𝑡𝑡𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤,𝑖𝑖
# 𝑜𝑜𝑜𝑜 𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 = ∑𝑛𝑛𝑖𝑖=1 𝐶𝐶𝐶𝐶𝐶𝐶∗𝑦𝑦
(6) steps can be applied in a continuous process.
(i) Standardize change-over process
with 𝑡𝑡𝑚𝑚𝑚𝑚𝑚𝑚,𝑖𝑖 for the manual working time and 𝑡𝑡𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤,𝑖𝑖 for the (ii) Stabilize change-over time
waiting time at machine i, and y as buffer factor towards CTT. (iii) Reduce internal change-over time
(iv) Reduce external change-over time
Within the SMED-method (Single Minute Exchange of
3.7. Phase 6: Improve adaptability for changing demands Dies) [28], the transformation of internal into external steps has
a central function to reduce the standstill of equipment.
The final phase is to prepare the cell for shifts in demand. However, since the worker is considered as the bottleneck
Three measures can be taken to flexibly balance capacity: within the cell only the reduction of the total change-over time
(i) Adapt working hours (short term) implicates an overall change-over improvement.
(ii) Shift operator capacity (short term)
(iii) Shift machine capacity (long term) 4. Application and validation of the method
The easiest way to adapt to a changing demand is to adapt
the working hours of the cell. This is limited by labor Applying the sequential steps of the proposed method, some
regulations and can only be applied for short term adaptations. steps may require minor adaptation. However, the overall order
In case of a decreasing workload for one cell and an increasing of the sequential steps can remain unchanged. During the
workload for another cell, operators can be shifted between practical application, the method has been implemented in the
different cells. To provide the necessary qualification, constant Process Learning Factory CiP at TU Darmstadt. For a piston
job rotation between the cells should be practiced. If the total rod and a bottom plate of pneumatic cylinders, a cellular
working hours cannot be adapted in order to increase output, an manufacturing cell has been installed.
additional operator can reduce the cell’s cycle time by shifting Phase 1: The customer takt time of the cylinder is calculated
the bottleneck from the operator to the machine with the longest to 90 seconds in assembly, indicating that for synchronous
cycle time. production one piston rod and one bottom plate must be
produced every one and a half minute. In the initial situation,
66 Stefan Seifermann et al. / Procedia CIRP 78 (2018) 61–66
6 Stefan Seifermann, Jörg Böllhoff, Carsten Schaede, Michael Kutzen and Joachim Metternich / Procedia CIRP 00 (2018) 000–000

the complete machining has a takt time of 172 seconds Factory CiP and various industries, showing the actual practical
requiring two shifts in machining to cover the demand for one applicability of the method.
assembly shift. The processes were analyzed in-depth.
Phase 2: To stabilize the process and to reduce quality References
feedback time, checking gauges were designed that enable
testing directly at the machine (worker self-control). [1] Ohno T. Toyota Production System, Beyond Large-Scale Production.
Portland: Productivity Press; 1988.
Phase 3: To synchronize machining with assembly, six [2] Womack JP, Jones DT, Roos D. The machine that changed the world. New
right-sized-machines (four mills and two lathes) have been York: Rawson Associates; 1990.
installed as an alternative machining process. The machines [3] Willenbrock H. Womacks Weisheiten – Interview with J. Womack. McK
Wissen 2003; 5:72-7.
were supposed to operate in a directed one-piece-flow. [4] Glass R, Seifermann S, Metternich, J. The Spread of Lean Production in
Therefore, the machines form a cell. the Assembly, Process and Machining Industry. Procedia CIRP 2016;
Phase 4: For the production of the eight variants of the 55:278-83.
[5] Metternich J, Bechtloff S, Seifermann S. Efficiency and Economic
piston rod and the two variants of the bottom plate, a first-in- Evaluation of Cellular Manufacturing to enable Lean Machining. Procedia
first-out flow with zero WIP was chosen, setting the sequence CIRP 2013; 7:592-7.
at the first machine. [6] Irani S. Handbook of cellular manufacturing systems. New York: John
Phase 5: To facilitate the manual work for the operator, low Wiley & Sons; 1999.
[7] Hon K, Chi H. A new approach of Group Technology Part Families
cost automation solutions were implemented. In combination Optimization. Ann CIRP Vol. 1 1994; 43:425–8.
with zero change-over times, this enables the operator to [8] Hon K, Lopez-Jaquez F. Configuration of Manufacturing Cells for
unload, load and start a machine in less than 10 seconds. Dynamic Manufacturing. Ann CIRP Vol. 1 2002; 51:391–4.
[9] Seifermann S, Böllhoff J, Metternich M, Bellaghnach A. Evaluation of
Phase 6: Manual work and machine cycle times have been Work Measurement Concepts for a Cellular Manufacturing Reference Line
balanced to a cell cycle time of 90 seconds. In this scenario, one to enable Low Cost Automation for Lean Machining. Procedia CIRP 2014;
worker operates the line at a possible takt time of 84 seconds, 17:588-93.
representing the bottleneck within the cell. For short-term [10] Metternich J, Abele E, Bechtloff S, Seifermann S. Static total cost
comparison model to identify economic fields of applica-tion of Cellular
periods of an increased demand, a second operator can join the Manufacturing for milling and drilling processes versus done-in-one-
cell in order to operate it at a cell cycle time of 75 seconds. In concepts. Ann CIRP Vol. 1 2015:471-4.
this scenario, a milling machine is the bottleneck. In case of [11] Hyer N, Wemmerlöv U. Reorganizing the Factory. Competing through
decreasing demand, the machining processes can be allocated Cellular Manufacturing. New York: Productivity Press; 2002.
[12] Miltenburg J. U-shaped production lines: A review of theory and practice.
to three milling machines with a resulting cycle time of 115 International Journal of Production Economics 2001; 70:201-14.
seconds. [13] Liker JK. The Toyota Way. New York: McGraw-Hill; 2004.
In addition, the method has been applied successfully to [14] Suzaki K. The New Shop Floor Management: Empowering People for
different scenarios throughout the industry ranging from a line Continuous Improvement. New York: Free Press; 1993.
[15] Roth NG, zur Steege C. Excellent Lean Production. The way to business
dedicated for one single product for the aerospace industry stability. Hamburg: Deutsche MTM-Vereinigung e.V.; 2014.
(mass production) to customer individual products in lot size [16] Nakajima S. Introduction to TPM: total productive maintenance.
one for the machine and equipment industry. The cells installed Cambridge: Productivity press; 1988.
covered between three and nine machine tools. It was shown [17] Baudin M. Working with Machines. New York: Productivity Press; 2007.
[18] Mitrofanov SP. The Scientific Principles of Group Technology.
that applying the proposed method enhances process stability, Yorkshire: National Lending Library for Science and Technology; 1966.
enabled the implementation of one piece flow and thereby [19] Burbidge JL. The Introduction of Group Technology. New York: John
lowered the throughput time significantly. Consisting of Wiley & Sons; 1975.
approved practices, the method ensures high practicability. It [20] Opitz H.Werkstückbeschreibendes Klassifizierungssystem. Essen:
Girardet; 1966.
can be applied in different production scenarios and is [21] Singh N, Rajamani D. Cellular Manufacturing Systems. Design, planning
independent from the initial state of the machining area. Still, and control. London: Chapman & Hall; 1996.
some limitations apply for the proposed method. Sufficient [22] Narendran TT, Srinivasan G. Machine-Component Matrix Clustering
knowledge of the applied techniques and practice is necessary. Methods for Cell Formation. In: Irani SA. Handbook of Cellular
Manufacturing Systems. New York: John Wiley & Sons, 1999; 35-68.
Furthermore, the method is focusing only on machining and [23] Böllhoff J, Schaede C, Seifermann S, Krause F, Metternich J. Auslegung
does not replace a holistic value stream management. von Sequenzfertigungszellen zur schlanken Zerspanung.
PRODUCTIVITY Management 3 2015; 20:27-30.
5. Summary [24] Monden Y. Toyota Production System. An integrated approach to Just-in-
Time. London: Chapman & Hall; 1994.
[25] Shingo S. A Study of the TPS system. Cambridge: Productivity Press;
This article gave an overview on the sequential phases 1989.
proposed and validated for implementing continuous flow [26] Seifermann S. Methode zur angepassten Erhöhung des
Automatisierungsgrades hybrider, schlanker Fertigungszellen. Aachen:
production in a machining environment by using lean Shaker; 2018.
machining cells, an area that has been neglected so far. After [27] Takeda H. LCIA – Low Cost Intelligent Automation. mi-Verlag,
an initial section describing the motivational background and Landsberg; 2006.
defining a lean machining cell, the six-phase method was [28] Shingo S. A Revolution in Manufacturing. The SMED system. Stanford;
1985.
introduced and all phases were thoroughly described. The
paper closed with real-life examples from the Process Learning