International Journal of Project Management 20 (2002) 271–277

www.elsevier.com/locate/ijproman

Effects of resource management regimes on project schedule

Saad Al-jibouri*
University of Twente, Construction Process Management Department, PO Box 217, 7500 AE, Enschede, The Netherlands

Received 4 October 2000; received in revised form 7 February 2001; accepted 21 February 2001

Abstract
This paper deals with the problem of resource scheduling within given resource constraints on a sectionalised construction pro-
ject. A computer model has been developed which realistically simulates the progress of projects. A project has been selected and
divided into sections whereby section managers are allocated part of the site resources to carry out their work and are made
responsible for the performance of the work on their sections. The major part of the model is a heuristic scheduler that has been run
under three types of project organisations, these are; central management, section management with co-operation between the
sections and independent section management. The resultant project duration under fixed resource level has been used in each of
these project management regimes as a measure of the efficiency of the resource scheduling process. It has been found that the
scheduling is more effective in the case of section management with co-operation between the sections than the other two cases.
# 2002 Elsevier Science Ltd and IPMA. All rights reserved.
Keywords: Managing programmes; Time; Management structures; Resource scheduling

1. Introduction developed to solve the problem of resource-constrained

project scheduling including theoretical optimisation
Construction projects are often large and complex; and heuristic rules, see for examples Kelly [1], Pascoe [2]
they consist of many different types of work; change and Mawdesley [3]. Heuristic scheduling approach is
rapidly; and are carried out in a changing and fre- one of the simplest and effective methods, which has
quently harsh environment. Despite this, clients want been proposed for the allocation of limited resource to
them completed in the minimum time possible and to the activities of a project. Researchers have used heur-
budget. In an attempt to achieve this, project managers istic approaches to get an acceptable solution that may
often rely on the use of various common planning and not necessarily be optimal [4,5]. In general these
control techniques available. approaches employ decision rules to decide the order
The common techniques of network planning such as for the activities to proceed at any stage of the project.
the Critical Path Method (CPM) and PERT are con- This implies that, depending on the decision rules to be
cerned with minimising project duration. They assume used, a priority number will be associated with each
no limit on the availability of the resources to be activity at any time during the project.
employed to complete all project activities on schedule. Heuristic methods of allocating limited resources for
This unrealistic assumption can lead to ineffective project scheduling also allow construction firms or
resource usage and project delays. In practice, most managers to apply these techniques for example, when
projects are limited to the number of resources they can they wish to carry out several projects with the same
use and in order to schedule activities with resource resources. In most cases, the total availability of these
constraints, other techniques have to be employed. resources become constraints and must therefore be
Resource allocation has been a major concern of recognised in the planning of the individual project.
research since the advent of the CPM/PERT network Usually when the projects are large, the resource allo-
techniques. Since late 1950, many algorithms have been cation is carried out in two ways: the allocation of
resources to all the projects is integrated into one multi-
* Corresponding author. Tel.: +31-53-489-4887; fax: +31-53-489-
project plan, or the resource allocation for each project
2511. for only the resources being shared is integrated in one
E-mail address: [email protected] multi-project functional plan.
0263-7863/02/$22.00 # 2002 Elsevier Science Ltd and IPMA. All rights reserved.
PII: S0263-7863(01)00020-5
272 S. Al-jibouri / International Journal of Project Management 20 (2002) 271–277

In carrying out multi-project scheduling, construction number of each type of resource required by the activ-
firms or managers strive to provide more efficient bases ities are available for the use of the project. The sche-
for project resource planning and scheduling because, duler uses the latest start time as a major sort. This
quite often, special skills and resources are required to implies that, at any one time, if two or more activities
be utilised across many networks simultaneously. Multi- compete for the same resources when there is insufficient
project scheduling techniques permit the movement of number of these resources to operate all of them con-
these skills and resources from one project to another, currently, priority will be given to the activity with the
as if the networks of the different projects are linked earliest latest start time. If required, this decision rule
together, even when they appear to be independent of can be changed without any alteration to the structure
the each other. The movement of resources from one of the model.
site to another may be considered as a job and time for The scheduling of the activities in the model is con-
transferring of resources between projects should be sidered daily. If an activity needs either resources or
taken into consideration [6,7]. materials that cannot be satisfied, then that activity has
Despite the fact that abundant literature exists on to be abandoned for that day. As a natural consequence
both the relationship and importance between project
organisation structures on one hand and planning and
scheduling on the other hand, very little attempts have
been made in the past to measure and quantify the
effects of the these two aspects on each other [8,9].
This paper describes the development of a compu-
terised simulation model that considers the problem of
multi-section scheduling which is based on the same
principle as multi-project scheduling. In this paper, the
whole project is represented by a single precedence net-
work. Sections are assumed to be located within the
same site, and therefore, the time needed by resources to
move between sections is regarded to be too small to be
taken into consideration.
This paper also describes the use of the model to
investigate the effects of three types of management
regimes in a sectionalised project on the project sche-
dule. The three resource management regimes used for
this purpose are:
1. Central Management;
2. Independent Section Management; and
3. Section Management with Co-operation.
The resultant project duration, under limited resource
levels, has been used in each of the above cases as a
measure of the efficiency of the scheduling process.

2. Model structure

The model described here is part of a more compre-

hensive computer simulation program, which has ori-
ginally been designed for the purpose of providing
information about the progress of the project to be used
by different project control systems. An outline flow
chart of the model is shown in Fig. 1.
The model is designed to be a day-by-day scheduler
that has variable resource levels and which gives vari-
able duration for the activities. It is a serial sort sche-
duler in which the network calculations are performed
only once at the beginning of scheduling [1,10,11]. It is
based on the assumption that a certain predetermined Fig. 1. The model flowchart.
 S. Al-jibouri / International Journal of Project Management 20 (2002) 271–277 273

of this, an activity can be stopped due to the shortage of The activity duration is taken to be the largest of the
either resources or materials and then restarted when resulting resource durations.
there are enough of both of them to schedule the activ- The minimum duration (M) for an activity can be
ity. This means that an activity can be split over two or determined for example on the basis of assumed physi-
more periods of time and then proceed at different rates cal constraints such as space. An upper limit of each
of progress depending on the level of resources available. resource is assumed above which an increase of that
The types of events that can take place at any time in resource will not affect the speed of the activity. This
the scheduler are the start or finish of an activity, mate- limit differs from one resource to another. This
rial delivery and stock out, resource hiring and firing. assumption is made because, in real life, the number of
The activity model used here allows changing of the resources working on an activity is usually affected by
duration of an activity by altering the amount of the type of work involved and the space available for
resources between specified limits. Several authors in the these resources to work on that activity. There is a
past have suggested this type of model, since variable physical limit, for example, to the number of excavators
duration activity, in general, is closer to real life than and labourers that can work in or near a hole to be
fixed duration model [1,7]. To facilitate the estimation excavated on a construction site. If the number of
of activity duration on the basis of such a model, a resources on the activity exceeds a certain limit, it will
mathematical relationship between the number of result in extremely inefficient working conditions and
resources working on the activity and the duration is might result in no work being carried out at all. Also,
assumed. This relation is represented by a hyperbola as too many resources in a limited space will undermine
shown in Fig. 2. The relationship is approximated as safety on site.
shown by the dotted line on the same figure in order to As well as the upper limit of resources that can be
simplify the calculations. The resource requirement of used by any activity, a minimum limit of resources
each activity is specified in terms of the quantity of work below which work will not proceed is assumed, such as
required by each resource to finish the activity. The one labourer, 1/10 of a machine. The allocation of part
duration of an activity is calculated as follows: of a machine to an activity means either a full machine
is working in a part of the duration or a full machine is
wi
Di ¼ þM ð1Þ shared with other activities. For convenience of simula-
ni tion, the division of machines has been made into 10
parts.
Where Di=the duration required by resource i to com- In addition to the limits of resources on activities, the
plete the activity wi=work quantity for resource i model also simulates the effect of supervision on the
ni=number of resource type i used M=minimum progress of the work and the productivity of labour. It
duration of the activity uses exponential mathematical relationship between the
The above mathematical relation is applied in the productivity of labour and the ratio of the number of
scheduler for every type of resource used by the activity. foremen to the number of operatives. It assumes labour
productivity to be 50% of the optimum if no super-
vision is involved in the work. Using such assumption, a
‘productivity factor’ is calculated using the following
formula:

0:5
PF ¼ 1  ð2Þ
eR

Number of foremen
where PF=Productivity factor and R ¼ Number of operatives
The final duration of an activity can then calculated
as being the maximum duration based on resources
divided by the productivity factor as indicated in Eq. (3)
below.

Di ðmaxÞ
D¼ ð3Þ
PF

It may be worth mentioning here that the productivity

factor may differ from one section of the work to
another as a result of the degree of supervision available
Fig. 2. Activity’s resource–duration relationship. in each section.
274 S. Al-jibouri / International Journal of Project Management 20 (2002) 271–277

2.1. Model representation of resources The problem of orders and deliveries of materials in
real life has been simulated in the model by keeping a
Physical resources in the model have been divided predetermined list of order and delivery times for all the
into three categories: materials on site. The delivery of each kind of material
then takes place automatically at the specified time.
1. operatives, skilled and unskilled;
Control of orders and delivery of materials is a very
2. machines; and
important factor on any construction site, in order to
3. subcontractors;
ensure the arrival of materials in time and at the same
Each of the above categories was given a different time to avoid excessive storage requirements. The deliv-
code number in order to be able to distinguish between eries of consumable and re-usable materials in the
them during scheduling. model take place regularly throughout the project. On
The utilisation factor of the resources in the model is the other hand, just like in real projects, special materi-
calculated as being the ratio between the total resources als have to be ordered sometime before they are needed.
employed and the total resources provided. In order to simulate this, the model assumes that the
To be realistic, the model also addresses the problem start of any activity, that requires a special material, to
of non-attendance of resources by using a uniformly be dependent on the time taken to deliver that material.
distributed random number between 0 and 0.2 for the This time represents the duration of the order activity of
allowance of absenteeism within labour. This means the special material.
that there is a possibility of non-attendance of 20% of The fact that the model has been originally designed
the total number of labour every day as a maximum. to provide progress information to be used by various
The hiring and firing of resources in the model take control systems means that the model contains facilities
place at the end of every week if required. The decision for categorisation and representation of finance. It also
of hiring and firing is made in accordance with the includes various tools for retrieving and reporting of
planned schedule times of these resources. The proce- information to the project management. These topics,
dure is performed in such a way that the type and however, will not be discussed here because they are not
number of resources are associated with the section relevant to the subject of this paper.
number for or from which these resources are hired or
fired. 2.3. Project management regimes

2.2. Model representation of materials A project should be organised to facilitate produc-

tion. The project manager is responsible to his superiors
In order for scheduling to take place in the model, for the site in its totality. On small sites, he will carry
requirements of activities from materials should be also out most of the management work himself and part of
gratified. Materials in the model have been classified as this work may involve planning and controlling the
follows: work of the plant and labour employed on the various
activities. The manager will agree a long-term plan to
1. Consumable materials: these are the kind of mate-
fulfil the major objectives of the project and also pre-
rials that are usually used throughout the project
pare the short-term plans to ensure efficient use of pro-
by all the activities, such as sand, cement, etc.
ject resources.
Materials of this kind usually become part of the
On large sites, the project manager usually divides the
work they have been used to construct, and there-
work into sections and sections managers are made
fore, can be used only once.
responsible for totality of the work on their sections.
2. Special materials: special materials are those that
The project manager’s personal work, in this case, has
are not in common use among many activities.
commonly less to do with the detailed short term plan-
They are used for special kinds of activities and
ning of the work, except to ensure its conformity with
not for the others. An example of such a material
his overall plan. The section managers are allocated a
would be bridge bearings.
proportion of the site resources to carry out their work
3. Re-usable materials: this type of material differs
and account for the performance of their sections to the
from the others in that materials of this kind can
project manager.
be used more than once (e.g. shuttering and scaf-
There are many ways in which a project can be orga-
folding).
nised, in particular the balance between section auton-
A different wastage factor for each material is omy and central direction can vary widely. In this model
assumed from experience and wastage is considered to however, three types of project organisation have been
occur on delivery, although in reality it is generally a used for the purpose of investigating the effect of dif-
function of duration and condition of storage as well as ferent resource management regimes on the project
usage. This has been done to simplify the model. schedule, these are:
 S. Al-jibouri / International Journal of Project Management 20 (2002) 271–277 275

— Sub-contractor
1. Central Management: in this kind of project orga-
nisation, the project is not split into sections and

Resource 11
the project manager is assumed to be the person
who is responsible for allocating the resources to
all the activities of the project. The model also

10
1
2
4
6
8
assumes that there is a single resources pool avail-
able for the whole site.

Resource 10
— Crane
2. Independent Section Management: in this case, the
model simulates the project as being divided into
sections. Each of the sections has its own resource

12
18
24
3
3
6
pool that is independent of the others. For the

— Dumper
purpose of the experiments, it has been considered

Resource 9
in this case that there is no kind of co-operation
between sections of the project concerning the

10
16
22
28
3
5
sharing of resources between them.
3. Sections Management with Co-operation: in this

— Labourer
case, co-operation between various sections with

Resource 8
respect to the movement of resources between
them is considered. Co-operation between sections

13
25
50
58
66
72
is based on the idea of having a common or resi-
dual pool into which all the surplus resources that

— Carpenter
are not used by any section are kept to be available

Resource 7
to other sections.

18
36
42
48
54
9
To perform the simulation, a single project has been

— Excavator
chosen and resource scheduling experiments were then
carried out on it to examine the effects of dividing it into Resource 6
sections. The project is treated under the above three

12
16
20
types of organisation. As a result of each experiment, 2
4
8
project duration is produced and used as an indicator of
— Steel fixer

how efficient the resource scheduling process is. The

Resource 5

resultant project duration from the various experiments

are presented and discussed in a later section of the paper.
17
34
42
50
58
9

2.4. Sample project

Resource 4
— Pile rig

The model is designed to simulate any project but the

project chosen for the experiments consists of 40 activ-
16
24
32
4
4
8

ities, which represent the construction of a footbridge.

Resource 3

A precedence network has been used to represent the

— Lorry

relations between the activities.

Resource levels used in the scheduling experiments

In practice, sections of the project are usually chosen

12
20
28
36
4
6

according to their areas on site or to the types of work

involved. In the model, the project was divided into five
Resource 2
— Shovel

main sections.
The activities of the project have been divided between
10
12

the different sections of the project in such a way that there

2
3
6
8

is no sharing of any activity between two or more sections.

Resource 1
— Grader

Each section with a different number of activities has a

resource pool, which is independent of the others.
10
14
18
2
3
6

2.5. Experiments and findings

Resource
Table 1

For each of the management regimes discussed ear-

level

lier, six experiments have been carried out varying the

1
2
3
4
5
6
276 S. Al-jibouri / International Journal of Project Management 20 (2002) 271–277

resource level in each. In the ‘section’ case, each of the for all the six resource levels used, section manage-
resource levels shown in Table 1 represents the sum of ment with co-operation always gave shorter project
the number of each kind of resource in the section pools. duration, and hence, a better schedule than in the
In all the experiments, priority was given to the cases of the other two types of project management
activities with the earliest–latest start time independent regimes.
of their section. Priority based on latest start time was The reported results of the above experiments indicate
chosen because it seems to produce a better schedule [2]. that for the same resource level, the scheduler gave dif-
The results are shown in Table 2. This table shows that ferent project durations under the different types of
organisation. They also indicate that project duration
decreases as the resources level increases but there is no
Table 2 linear relationship between the two.
The results of project duration (in days) under different types of In the case of independent section management where
organisationa
there was no co-operation between sections, the project
Resource Central Independent Sections with duration was always the longest. The indications pro-
level management section management Co-operation vided by these results seem to be realistic, since the
1 820 1705 736 activities in this case can only use resources from their
2 597 854 477 own section pool, which represent only a portion of the
3 379 518 369 site total resources.
4 358 468 349 On the other hand, the results have indicated that the
5 341 439 339
scheduler gave shorter project durations in the case of
6 332 419 326
section management with co-operation than in the case
a
Priority is given to the activities with the earliest–latest start time. of central management. This, however, may appear to

Fig. 3. Part of a bar chart of the project under different types of organisation.
 S. Al-jibouri / International Journal of Project Management 20 (2002) 271–277 277

be unrealistic since the resources available to schedule However, as stated earlier, this is not expected to always
the activities in the latter case represent the sum of the be the case. The efficiency of resource utilisation is gen-
number of resources in all the section pools. Never- erally dependent on the methods of selecting priorities
theless, the analysis of the results has shown that in the and sections as well as the methods of distributing the
case of central management, resources were made site resources among these sections. Nevertheless, the
available to all activities in the project and that sche- reduction in project duration by as much as 20% in the
duling was carried out according to the priorities set for case of section management with co-operation as com-
these activities. This resulted in a very narrow selection pared to central management is significant enough to be
of activities to be scheduled first which prevented the worthy of more extensive research into this problem.
start of some remaining activities. On the other hand, in
the case of sections with co-operation in which resour-
ces were allocated to sections, there was an opportunity References
for greater number of activities to start at the same time.
An example of such effect is represented by the bar [1] Kelley J, Walker M. Critical path planning and scheduling. Pro-
ceedings of the Eastern Joint Computer Conference, December
chart shown in Fig. 3. This figure shows that although 1959.
the durations of individual activities are longer in the [2] Pascoe TL. An Experimental Comparison of Heuristic Methods
case of sections with co-operation, at the period for Allocating Resources, PhD thesis, Cambridge: University
between day 91 and day 135, more activities are engaged Engineering Dept., 1965.
[3] Mawdesley M. Resource Scheduling, PhD thesis, University of
than in the case of central management. This has resul-
Nottingham: Department of Civil Engineering, 1973.
ted in shorter project duration. [4] Tam PWM, Palaneeswaran E. Use of enhanced positional weight
method for constrained resource projects scheduling. Canadian
Journal of Civil Engineering 1999;26(2):242–7.
3. Conclusions [5] Khattab MM, Soyland K. Limited-resource allocation in con-
struction projects, Proceedings of the 19th International Con-
ference on Computers and Industrial Engineering, Miami, FL,
This paper has shown that the efficiency of the 1996.
method of allocating resources to the activities is influ- [6] Abdul Ahfur N. Multi-Project Scheduling, MPhil thesis, Uni-
enced by the type of project organisation employed. versity of Nottingham: Department of Civil Engineering, 1984.
In this simulation model, three types of project orga- [7] Battersby A. Network analysis for planning and scheduling. 3rd
nisation for managing resources have been used and the ed. Basingstoke: Macmillan, 1970.
[8] Harrison F. Advanced project management: a structured
results of the experiments have shown that project approach. 3rd ed. Gower, Publishing, 1992.
duration is always longer in the case of independent [9] Brod E, Pohani, B. Integrated project management organisation
section management than in the other two types. The for multiple projects: management by projects, 10th Internet
scheduler has also given shorter project duration in the World Congress, 1990.
case of sectionalised management with co-operation [10] Lockyer K, Gordon J. Project management and project network
techniques. 6th ed. London: Pitman Publishing, 1996.
than in the centralised type of management. This is [11] Kolisch R. Serial and parallel resource-constrained project sche-
because the first case provides an opportunity to sche- duling methods revisited: theory and computation. European
dule a greater number of activities at the same time. Journal of Operation Research 1996;90:320–33.