CONSTRUCTION CREW DESIGN: A LEAN APPROACH
By
Naveen Nerwal

THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the Degree of
MASTER OF SCIENCE
Construction Management
2012

ABSTRACT
CONTRUCTION CREW DESIN: A LEAN APPROACH
By
Naveen Nerwal

Construction work is executed by a variety of speciality trades, using group of workers
commonly known as crews. Performance of these crews is very crucial to the work being
executed and consequently for the overall project performance. Yet decisions regarding how to
design a crew (makeup and how work itself will flow) and improve their performance are
addressed in an ad hoc fashion, if at all. Also, despite the clear significance of crew design, only
a modest scholarly interest in this area has been seen in the last two decades. Developing leanbased guidelines for crew design will allow us to answer questions such as: Can we
choreographically design construction work crews, and how? The aim of this research is to
address this question by developing lean-based crew design guidelines for construction
operations by conducting an extensive literature review focused on “work design/team design”
that have well developed theoretical and empirical foundations in Lean Production, Lean
Construction, Socio-Technical System theory, and Organizational Psychology. Synthesizing
these existing production theories into a new unified model to develop a set of lean-based crew
design guidelines that allows experimenting with different crew interactions and interrelations,
and not only sizes with the help of using simulation models between discrete event simulation
and system dynamics modeling is expected to lead to better overall performance of crews
achieving lean construction ideals of minimizing waste and maximizing value for construction
operations and projects at large.

Dedicated to
My Loving Parents
&
My Mentor Dr. Tariq Sami Abdelhamid

iii

TABLE OF CONTENTS
List of Tables ................................................................................................................................ vii
List of Figures .............................................................................................................................. viii
1 Introduction ................................................................................................................................ 1
1.1
Background ...................................................................................................................... 1
1.2
Need Statement ................................................................................................................ 5
1.3
Research Questions .......................................................................................................... 6
1.4
Research Goals and Objectives ........................................................................................ 6
1.5
Research Hypothesis ........................................................................................................ 7
1.6
Research Scope ................................................................................................................ 7
1.7
Research Method ............................................................................................................. 8
1.8
Research Benefits and Contribution .............................................................................. 12
1.9
Chapter Summary .......................................................................................................... 13
2 Literature Review ..................................................................................................................... 14
2.1
Introduction .................................................................................................................... 14
2.2
Construction as a Complex System ............................................................................... 14
2.3
Existing Production Planning Models ........................................................................... 19
2.3.1 Entity Interdependence Model .................................................................................. 20
2.3.2 Conversion Model ..................................................................................................... 21
2.3.3 Method Productivity Delay Model (MPDM) ........................................................... 23
2.3.4 Control Model ........................................................................................................... 24
2.3.5 Factor Model ............................................................................................................. 25
2.3.6 Conceptual Construction Process Model .................................................................. 28
2.3.7 Integrated Building Process Model ........................................................................... 30
2.3.8 Flow Process Model .................................................................................................. 31
2.3.9 The Last Planner® System ....................................................................................... 32
2.3.10
Construction Process Analysis model (CPA) ....................................................... 34
2.4
Existing Crew and Work Analysis ................................................................................. 35
2.5
Existing Production System Theories ............................................................................ 44
2.5.1 Theory of constraints ................................................................................................ 44
2.5.2 Lean Production Theory ........................................................................................... 45
2.5.3 Sociotechnical System Theory.................................................................................. 62
2.6
Crew and Work Design .................................................................................................. 68
2.7
Chapter Summary .......................................................................................................... 80
3 Modeling and Simulation in Construction ............................................................................... 81
3.1
Introduction to Modeling and Simulation ...................................................................... 81
3.2
Classification of Simulation Models .............................................................................. 83
3.2.1 Deterministic and Stochastic Simulation Models ..................................................... 84
3.2.2 Static and Dynamic Simulation Models ................................................................... 84
3.2.3 Discrete and Continuous Simulation Models ........................................................... 84
3.3
Discrete Event Simulation (DES) .................................................................................. 87
3.3.1 DES Introduction ...................................................................................................... 87
iv

3.3.2 DES Modeling Concepts........................................................................................... 87
3.3.3 DES World Views..................................................................................................... 89
3.3.4 DES Software Tools ................................................................................................. 92
3.3.5 DES Application ....................................................................................................... 95
3.3.6 DES Advantages and Limitations ............................................................................. 95
3.4
Continuous Simulation - System Dynamics (SD) ......................................................... 96
3.4.1 System Dynamics (SD) Introduction ........................................................................ 96
3.4.2 SD Modeling Concepts ............................................................................................. 98
3.4.3 SD Modeling Tools ................................................................................................. 102
3.4.4 SD Application........................................................................................................ 104
3.4.5 SD Advantages and Limitations ............................................................................. 105
3.5
Chapter Summary ........................................................................................................ 106
4 Lean Based Crew Design ....................................................................................................... 107
4.1
Introduction .................................................................................................................. 107
4.2
Crew and Work Design Concept ................................................................................. 108
4.3
Lean Crew Design Characteristics ............................................................................... 109
4.3.1 Flow Characteristics................................................................................................ 112
4.3.2 Work Characteristics ............................................................................................... 116
4.4
Lean Crew Design Guidelines ..................................................................................... 124
4.5
LBCD Model Development ......................................................................................... 140
4.6
Lean Crew Design Process .......................................................................................... 141
4.7
Chapter Summary ........................................................................................................ 142
5 Case Study - Model Building, Simulation & Experimentation .............................................. 143
5.1
Introduction .................................................................................................................. 143
5.2
Case Study Background ............................................................................................... 143
5.3
Discrete Event Simulation Model ................................................................................ 146
5.3.1 DES Base Crew Model ........................................................................................... 146
5.3.2 DES Improved Crew Model ................................................................................... 156
5.3.3 DES Lean Crew Model ........................................................................................... 170
5.4
System Dynamics Model ............................................................................................. 181
5.4.1 SD Qualitative Model ............................................................................................. 182
5.4.2 SD Quantitative Models .......................................................................................... 183
5.5
Crew Design Analysis and Comparison ...................................................................... 192
5.6
Chapter Summary ........................................................................................................ 200
6 Conclusion .............................................................................................................................. 202
6.1
Research Overview ...................................................................................................... 202
6.2
Research Contributions ................................................................................................ 203
6.3
Conclusion ................................................................................................................... 205
6.3.1 Quantitative Findings from Case Study .................................................................. 205
6.3.2 Qualitative Findings from Literature Review ......................................................... 207
APPENDICES ............................................................................................................................ 209
Appendix A: DES “Base” Crew Model Formulation and Simulation Results. ...................... 210
Appendix B: DES “Improved” Crew Model Formulation and Simulation Results. .............. 211
v

Appendix C: DES “Lean” Crew Model Formulation and Simulation Results ....................... 216
Appendix D: SD “Base” Crew Model Formulation ............................................................... 219
Appendix E: SD “Improved” Crew Model Formulation ........................................................ 228
Appendix F: SD “Lean” Crew Model Formulation ................................................................ 242
BIBLIOGRAPHY ....................................................................................................................... 252

vi

LIST OF TABLES
Table 5.1: DES "Base" crew model data …………………………………………………......148
Table 5.2: Simulation results of the “Base” crew of installation of light fixtures …………....151
Table 5.2: DES "Improved" crew model data ……………………………………………......158
Table 5.3: Simulation results for the “Improved” crew of installation of light fixtures ……...163
Table 5.4: DES "Lean" crew model data …………………………………………………......172
Table 5.5: Simulation results of the “Lean” crew of installation of light fixtures ……..……..176
Table 5.6: Comparison of DES and SD simulation results ……………………………….…...191
Table 5.7: Motion in 3 crew designs …………………………………………………………..192
Table 5.8: Non-processing time in 3 crew designs ……………………………………….……193
Table 5.9: Cycle times of assembling and installation activities in three crew designs...196

vii

LIST OF FIGURES
Figure 1.1: Percent changes in real value added by construction (For interpretation of the
references to color in this and all other figures, the reader is referred to the electronic version of
this thesis.) ...................................................................................................................................... 1
Figure 1.2: Representation of research method phases................................................................... 8
Figure 1.3: Synthesis of existing crew and work design approaches to develop lean crew design
characteristics. ............................................................................................................................... 10
Figure 1.4: Lean-based Crew Design (LbCD) Model .................................................................. 11
Figure 2.1: Entity Interdependence Model, Thompson (1967). ................................................... 21
Figure 2.2: Conversion Model adapted from Alexander (1974), Koskela (1992). ....................... 22
Figure 2.3: Conversion model, Walker (1985) ............................................................................. 23
Figure 2.4: Control Model Sanvido (1984)................................................................................... 25
Figure 2.5: Actual productivity vs. ideal productivity curve, Thomas et al (1987). ..................... 26
Figure 2.6: Factor Model, Thomas et al (1990). ........................................................................... 27
Figure 2.7: Conceptual Construction Process Model, Sanvido (1988) ......................................... 29
Figure 2.8: Schematic representation of the Function Box, Chung (1989) .................................. 30
Figure 2.9: Flow process model, Koskela (1992) ......................................................................... 31
Figure 2.10: Traditional PM Vs. Last Planner system, Ballard and Howell (1997) ..................... 32
Figure 2.11: Last Planner® System, Ballard and Howell (1997) ................................................. 33
Figure 2.12: Work Study Process, Drewin (1985) ........................................................................ 38
Figure 2.13: Graphical representation of Delay Model (Bernard et al. 1973) .............................. 39
Figure 2.14: Graphical representation of Activity Model (Thomas et al. 1983) .......................... 40
Figure 2.15: Graphical representation of Work Team Autonomy (Banker et al 1996) ................ 72
Figure 3.1: Graphical representation of discrete and continuous models (Jain 1991) .................. 85
Figure 3.2: Graphical representation of event, activity and process in DES (Page 1994) ............ 88
Figure 3.3: Graphical representation of event scheduling paradigm/view Pidd (1998) ............... 89
viii

Figure 3.4: Graphical representation of activity scanning paradigm/view Pidd (1998) ............... 90
Figure 3.5: Graphical representation of Three-Phase (AS) paradigm/view Pidd (1998) ............. 91
Figure 3.6: Graphical representation of the modeling elements in EZStrobe (Martinez 2001) ... 94
Figure 3.7: Graphical representation of patterns of behavior of systems (Kirkwood 1998) ........ 99
Figure 3.8: Graphical notation of positive and negative feedback or causal loops diagram. ..... 100
Figure 3.9: Sales Growth example for combination of positive and negative causal loops
(Kirkwood 1998). ........................................................................................................................ 101
Figure 4.1: Synthesis of existing crew and work design approaches. ........................................ 109
Figure 4.2: Lean Crew Design Characteristics ........................................................................... 111
Figure 4.3: Causal relationships between primary & intermediate crew design (flow)
characteristics and the final crew performance outcomes (Part A). ........................................... 113
Figure 4.4: Causal relationships between primary & intermediate crew design (flow)
characteristics and the final crew performance outcomes (Part B)............................................. 114
Figure 4.5: Causal relationships between primary & intermediate crew design (work)
characteristics and the final crew performance outcomes (Part A). ........................................... 121
Figure 4.6: Causal relationships between primary & intermediate crew design (work)
characteristics and the final crew performance outcomes (Part B)............................................. 122
Figure 4.7: Causal relationships between primary & intermediate crew design (work)
characteristics and the final crew performance outcomes (Part C)............................................. 123
Figure 4.8: Hard factors affecting the crew performance. .......................................................... 124
Figure 4.9: Soft factors affecting the crew performance. ........................................................... 124
Figure 4.10: Lean-based Crew Design (LbCD) Model .............................................................. 140
Figure 4.11: Lean-based crew design (LbCD) process ............................................................... 141
Figure 5.1: Type of light fixture installed (Figure 9 in Howell et al. 1993) ............................... 144
Figure 5.2: Original vs. modified process of installation of light fixtures .................................. 145
Figure 5.3: Sequence of activities in the “Base” crew case of installation of light fixtures ....... 147
Figure 5.4: Simulation model (Assembling Cycle) for the “Base” crew case of installation of
light fixtures ................................................................................................................................ 149

ix

Figure 5.5: Simulation model (Restocking Cycle) for the “Base” crew case of installation of light
fixture .......................................................................................................................................... 149
Figure 5.6: Simulation model (Installation Cycle) for the “Base” crew case of installation of light
fixtures ........................................................................................................................................ 150
Figure 5.7: Production rates for the "Base" crew case................................................................ 152
Figure 5.8: Sequence of activities in the “Improved” crew case of installation of light fixtures 157
Figure 5.9: Simulation model for the installation activity cycle in the “Improved” crew case of
installation of light fixtures ......................................................................................................... 159
Figure 5.10: Simulation model for the wire and clip activity cycle in the “Improved” crew case
of installation of light fixtures..................................................................................................... 160
Figure 5.11: Simulation model for the lamp & finish activity cycle in the “Improved” crew of
installation of light fixtures ......................................................................................................... 161
Figure 5.12: Production rates for the "Improved" crew case ...................................................... 165
Figure 5.13: Sequence of activities in the “Lean” crew case of installation of light fixtures ..... 171
Figure 5.14: Simulation model for the apprentice, assembling & restocking activity cycle in the
“Lean” crew of installation of light fixtures ............................................................................... 173
Figure 5.15: Simulation model for the journeyman, installation cycle in the “Lean” crew of
installation of light fixtures ......................................................................................................... 174
Figure 5.16: Production rates for the "Lean" crew case ............................................................. 177
Figure 5.17: Representation of assembling cycle for DES VS. SD “Base” case........................ 184
Figure 5.18: Pattern of behavior for the apprentice in assembling cycle for “base” crew case in
SD (Part A). ................................................................................................................................ 185
Figure 5.19: Pattern of behavior for the apprentice in assembling cycle for “base” crew case in
SD (Part B) .................................................................................................................................. 186
Figure 5.20: Pattern of behavior for the light fixture in assembling cycle for “base” crew case in
SD (Part A) ................................................................................................................................. 186
Figure 5.21: Pattern of behavior for the light fixture in assembling cycle for “base” crew case in
SD (Part B) .................................................................................................................................. 187
Figure 5.22: Representation of the quantitative SD model for “Base” crew case (Part A) ........ 188
Figure 5.23: Representation of the quantitative SD model for “Base” crew case (Part B) ........ 189
x

Figure 5.24: Representation of the quantitative SD model for “Base” crew case Part (C) ........ 190
Figure 5.25: Resource Idle/Wait times in 3 crew designs .......................................................... 194
Figure 5.26: Resource Utilizations in 3 crew designs................................................................. 195
Figure 5.27: Work-in-progress (WIP) in the three crew designs ................................................ 196
Figure 5.28: Operation production rates between Base and Lean crew design .......................... 197
Figure 5.29: Impact of crew size on the production rates between the Base and Lean crew design
..................................................................................................................................................... 198
Figure 6.1: Unit Cost and Time of Operation Savings in Lean crew ......................................... 206
Figure 6.2: Overall Crew productivity (production rate) improvement in Lean crew................ 206

xi

1 INTRODUCTION

1.1

Background

Construction industry plays an important role in contributing to the economic health of the
United States. According to U.S. Bureau of Economic Analysis, “Construction fell for the sixth
consecutive year and the 3.2 percent drop was the smallest since the 3.0 percent drop in 2006”
(Teresa L. Gilmore et al 2011) (Figure 1.1).

Percent Changes in Real Value Added by
Construction

8

3

-2

2007

2008

2009

2010
-3.2

-7

-5.5

-6.6

-12
-12.9
-17

Figure 1.1: Percent changes in real value added by construction (For interpretation of the
references to color in this and all other figures, the reader is referred to the electronic version of
this thesis.)
Unlike other industries such as manufacturing which have seen a steady increase in the
value added to the U.S. Gross Domestic Product (GDP), construction industry has gone

1

under a substantial decline over the past decade (U.S. Bureau of Economic Analysis). This
decline in the value added by construction industry could be attributed to a lot of reasons and it
would not be an exaggeration to say that the ever increasing size, scope and complex nature of
the construction projects have played a major role towards this decline.

However, the

construction industry has displayed great difficulty in coping with the increasing complexity of
major construction projects (Stocks and Male, 1984). Also, as the construction projects are
becoming increasingly complex; that traditional project management methods are proving
inadequate (Williams, 1999). Further, Dalchar, 1993 claims that ``Contemporary project
management practice is characterized by: late delivery, exceeded budgets, reduced functionality
and questioned quality. As the complexity and scale of attempted projects increases, the ability to
bring these projects to a successful completion dramatically decreases''.
Conventional construction management has focused for many years on overcoming the
complexity and improving project performance by primarily focusing on project management
techniques that rely on an outmoded time-cost-quality tradeoff paradigm. The failure and
inability of the conceptual models of construction management to deliver on the mantra of ‘ontime, at budget, and at desired quality’ is discussed at length in Koskela’s seminal 1992 report
and in Koskela (1999, 2000). Another paradigm-breaking anomaly was that observed by Ballard
(1994), Ballard and Howell (1994a and 1994b), Howell and Ballard (1994a and 1994b) and
Howell (1998). Analysis of project plan failures indicated that about 50% of weekly tasks were
not completed (Ballard and Howell 2003). The preceding observations have led to the birth of
Lean Construction as a discipline that subsumes the transformation-dominated contemporary
construction management (Abdelhamid 2004).

2

According to Howell and Ballard (1999), the design of work methods under a Lean
Construction paradigm happens in two phases; the first involves structuring of work during
the product design stage before the start of construction operations; and the second phase
carries forward from master scheduling to weekly work.
Examples of techniques used for production system design during the design stage
include first run studies, computer simulation, and recently BIM. These techniques are used
to minimize waste and improve system throughput. However, the design of crews has
received less attention, and is typically considered adequate if the available resources are
provisioned (Nerwal and Abdelhamid 2010).
The production process on construction jobsites may be considered the most complex
undertaking in any industry (Bennett, 1991, El-Gafy and Abdelhamid 2008). These processes
consist of multitude of interdependent, uncertain and variable flow of resources. These
resource flows may produce varied outcome of operations in terms of cost, time and other
performance indicators, with no clear understanding yet of underlying casual relations.
According to Halpin, (1973) the flow of resources are present in the operations at the
micro and macro level of construction and may be classified depending on the resource type
being considered. The flows at the macro level are of more strategic importance than the
micro level flows which are present at the operational level.
Due to the complex interaction among units on the construction job site and in the
construction environment, mathematical models (e.g. queuing systems) can be applied to
only a limited number of special cases (Zayed and Halpin, 2000). But with advances in
computer technologies, simulation models have been used and developed for decades to help
analyze complex operations. Simulation is one of the most widely used operations research
3

and management-science techniques (Law and Kelton 2000). The use of simulation for studying
construction operations was first proposed and developed by Halpin in the 1970s (Halpin 1973).
Simulation modeling has traditionally been carried out using discrete event simulation
(DES). And it has proved successful in providing detailed analysis of systems involving linear
processes and modeling discrete changes in the system behavior over the last three decades. But
DES has its limitation of modeling only at the operation level where the entities are passive
objects present at a low level of abstraction. When the systems get more complex and integrated,
discrete event simulation faces serious challenges to keep up with the high level of abstraction
present at the macro level and at times modeling a system at the strategic level keeping the
operational level details and analysis seems impractical (Nerwal and Abdelhamid 2010).
In discrete event simulation, it is also difficult to model the "softer" aspects of projects such
as learning, pride or morale, and incorporate managerial actions which are important in
determining the project behavior (Williams 2002, El-Gafy and Abdelhamid 2009). Moreover,
along with modeling the softer aspects of projects, the various nonlinear cause-and-effect
relationships, feedback loops, and time delays existing in the complex construction processes
could be analyzed more effectively using the methods of systems thinking.
Systems thinking methods have been used for over thirty years (Forrester 1961) and are now
well established. However, these approaches require that we move away from looking at isolated
events and their causes and start to look at the organization as a system made up of interacting
parts (Kirkwood, 1998).
System Dynamics (SD) is a system thinking approach that follows an integrative perspective
in modeling systems while recognizing the information feedback characteristics so as to show
how organizational structure (in policies), and time delays (in decisions and actions) interact to

4

influence the behavior of the system (Forrester, 1965). It is an appropriate modeling
approach to simulate complex construction systems by capturing the cause-and-effect
relationships of the various factors affecting the behavior of the system in a casual loop
diagram.
1.2

Need Statement

A problem facing estimating and scheduling teams working on early stages of a construction
project is the need to determine a reasonable number and composition of crews that can be
effectively used in the construction process (Hassanein 1997).
Can we choreographically design construction work crews, and how? This is the question
that this research takes on. This will be referred to as Lean-based Crew Design (LbCD).
In an effort to have a better crew design as an integral part of a production system, two
main systems-based theories are used in this research, namely, Lean Construction and SocioTechnical systems theory.

These two theories along with other existing production

management theories are used to develop guidelines for Lean-based Crew Design.
There are numerous factors which affect the crew executing the various interconnected tasks
involved in any operation. How to form or design a crew to carry out an operation in a
dynamically changing construction environment is a complex task and should be approached
with formal methods rather than as assumptions that rarely materialize. There is a need to have a
unified model based on the existing production theories that allows experimenting with different
crew interactions and interrelations, and not only sizes with the help of using simulation models
between discrete event simulation and system dynamics modeling.

5

In this research, discrete event simulation and system dynamics models are used to
demonstrate an example of the use of proposed lean crew design guidelines for construction
processes.
1.3

Research Questions

Traditionally, production systems have been designed as transformation of inputs into outputs
neglecting the flow and value generation in the construction. There is modest number of research
efforts in which production system are designed based on the transformation-flow-value theory
leading to maximum value and minimum waste in construction. While crew design has a major
impact on any production system model, there is no existing production system model which
supports Lean Crew Design.
This research will primarily address the following central questions:
1. What are the important soft and hard factors affecting production system, crew and work
design?
2. What improvements could be made to existing crew and work design models to develop a
lean crew design?
3. What are the primary Lean-based Crew Design guidelines for construction operations?
4. How can different dynamic simulation methods be integrated to develop a have a better
lean crew design model?
1.4

Research Goals and Objectives

The overall goal of this research is to develop crew design guidelines for construction operations.
The goal will be achieved by fulfilling the following objectives:

6

1. Synthesizing Lean Construction and Socio-Technical production system theories,
supplemented by other existing production system theories, into crew design guidelines.
Objective 1 will be achieved by performing the following tasks:
a.

Conduct literature review of existing construction models, production system
theories, and crew & work design models.

b.

Identifying the limitations of existing production system, crew and work design
models in the construction industry.

c.

Identifying the important soft and hard factors affecting the production, crew and
work design models in the construction industry.

2. Develop crew design guidelines based on literature review and hypothesized inferences.
3. Demonstrating an example of the use of proposed guidelines using simulation models
between discrete event simulation and system dynamics.
1.5

Research Hypothesis

In this research, it is hypothesized that lean-based crew design guidelines will give rise to better
overall crew performance. Hence, using the developed guidelines for crew design, the following
will be achieved, or at least approached:
1. Executing the construction operations in the most efficient way.
2. Minimizing waste in the construction operations.
3. Maximizing value generation in the construction operations.
1.6

Research Scope

The scope of this research would mainly be limited to the design of crews executing operations
on site. The study will contrast operations performance using a Lean-based Crew design versus a
conventional (non-Lean) Crew design. Lean-based crew design guidelines will be developed
7

based on analysis of performance obtained using the simulation models between discrete event
simulation and system dynamics modeling. For the scope of this research, sensitivity analysis is
not conducted. However, it would be next step for future research to conduct sensitivity analysis
and is acknowledged as one of the limitation of this research.
1.7

Research Method

To complete this research, five phases were followed (Figure 1.2). In the first phase, literature
review on the existing construction models, production system theories, and crew & work design
models was examined (Chapter 2 and 3). The second phase involved synthesizing the existing
production system theories, crew and work design models into crew design characteristics and
development of crew design guidelines (Chapter 4).

• LITERATURE REVIEW
PHASE
1
• CREW DESIGN CHARACTERISCTICS & GUIDELINES
PHASE
2
• LEAN BASED CREW DESIGN MODEL
PHASE
3
• DES & SD MODEL BUILDING & EXPERIMENTAION
PHASE
4
• SIMULATION ANALYSIS
PHASE
5

Figure 1.2: Representation of research method phases.
8

The third phase involved creating a Lean-based Crew Design (LbCD) Model (Chapter 4). The
fourth phase involved development of DES and SD computer simulation models with
experimentation about crew interaction/design attributes (Chapter 5). In the fifth phase, analysis
of the simulation results took place to finalize the impact of the different crew design guidelines
(Chapter 5).
st

In the 1 phase, an extensive review of the concepts related to this research was conducted.
Existing construction models present in the construction industry were reviewed. Next, existing
crew and work analysis methods followed by production theories such as Lean Production, Lean
Construction and Socio-Technical System theories were reviewed to identify the important soft
and hard factors or characteristics affecting the construction crew design.
Literature on crew design or crew formation mainly revolved around how to increase the
productivity of workers of the crew involved in the construction operations. Literature reviewed
on crew design or crew processes defined input-output or control methods for improving crew
performance and optimizing crew on the site. Further, simulation approaches such as Discrete
Event Simulation and System Dynamics modeling were reviewed in detail along with discussing
the applications, advantages and limitations of each of the modeling techniques in order to gain
better understanding of the process of building simulation models.
This research aims to develop a Lean-based Crew Design model that integrates both the
existing crew design and work design research; hence, in the 2

nd

phase research studies on

“work design/team design” that have well developed theoretical and empirical foundations in

9

lean production, lean construction, socio-technical system theory, social and organizational
psychology were synthesized into lean crew design characteristics (Figure 1.3).
The lean crew design characteristics are comprised of “primary crew design characteristics”,
“intermediate crew design characteristics and outcomes” and “final crew performance
outcomes”. The primary crew design characteristics are divided into two groups: (1) work
characteristics and (2) flow characteristics that are again divided into waste, task, social and crew
characteristics. The crew design characteristics are based on the TFV (transformation-flowvalue) concept outlining important design characteristics and performance outcomes that should
be considered while designing a crew.

Socio-Tech
Social &
Organization

Lean
Construction

Lean
Production

Lean Crew Design Characteristics
Figure 1.3: Synthesis of existing crew and work design approaches to develop lean crew design
characteristics.
10

After, synthesizing the existing crew design and work design research into lean crew design
characteristics, lean-based crew design guidelines are proposed to guide the crew design. The
proposed guidelines lead to the joint optimization of both the crew and the work performed,
resulting in better overall crew performance by optimizing the following crew performance
outcomes: (1) duration of work performed; (2) cost of work performed; (3) productivity; (4)
quality of work performed; (5) worker satisfaction; (6) worker learning; (7) worker motivation;
(8) pride in work performed.
rd

The 3

phase involved creating a Lean-based Crew Design (LbCD) Model. The structure of

the Lean-based Crew Design (LBCD) model (Figure 1.4) is patterned after a transformation
process structure where inputs are converted to outputs by means of a conversion process. The
model input consists of the lean crew design characteristics, developed based on existing crew
and work design approaches. The transformation process is comprised of three main phases: (1)
formation, (2) execution, and (3) refining.

INPUT

PROCESS

OUTPUT

Lean Crew
Design

Crew Design
Process

Lean-based
Crew Design

FORMATION

EXECUTION

REFINING

Figure 1.4: Lean-based Crew Design (LbCD) Model
The three phases are proposed to guide the crew design process in a production setting
involving

initial

formation

of

the

crew,
11

structuring

of

the

work,

execution

of

work/implementation (actual work/simulation model execution) where data collection of primary
and

intermediate

crew

design

indicators

is

conducted.

Finally,

the

design

indicators/characteristics are analyzed in the third phase to refine the initial crew design and
maximize the crew performance based on the lean crew design guidelines to produce output of
the process as the lean crew design.
th

The 4

phase involved development of DES and SD computer simulation models with

experimentation about crew interaction/design attributes in order to demonstrate how the crew
design characteristics affect the performance of the crew. For the purpose of modeling and
simulation, a case study involving a well-documented construction operation from a prior
research work conducted by Howell et al (1993) was selected. The case study consists of a real
time construction operation focusing on the installation of light fixtures by a construction crew.
To develop a simulation model that can effectively identify and demonstrate both the hard
and soft crew design characteristics, a combination of Discrete Event Simulation (DES) and
System Dynamics (SD) is used. The simulation models are developed first using Discrete Event
Simulation (DES) and then moving to System Dynamics (SD) providing the qualitative and
quantitative crew design models.
th

In the 5 phase, analysis of the simulation results took place to finalize the impact of the
different crew design guidelines.
1.8

Research Benefits and Contribution

This research is expected to contribute in developing a set of guidelines to design construction
crews to achieve Lean Construction ideals. The introduction of the use of simulation models

12

between discrete event simulation and system dynamics in this study will also be an addition to
existing literature on simulation, which typically uses discrete event simulation only.
It is expected that a crew designed based on the developed guidelines would achieve higher
performance levels on the short and long term. This research would benefit the construction
industry by providing basis for designing crew interaction and roles during performing a
construction operation.
1.9

Chapter Summary

This chapter provided an overview and scope of this research. It outlined and explained the need
to develop Lean-based Crew Design guidelines for construction operations. Further, it briefly
proposed a five phased research method to conduct this research.

13

2 LITERATURE REVIEW

2.1

Introduction

This chapter presents and discusses the concepts related to this research. The review starts with
discussing the dynamic and complex construction environment. This is followed by a brief
review of the existing crew design and construction models present in the construction industry.
Next, existing crew and work analysis methods are reviewed. Finally, production theories such
as Lean Production, Lean Construction and Socio-technical theories are reviewed.
Literature on crew design or crew formation mainly revolves around how to increase the
productivity of workers of the crew involved in the construction operations. Literature reviewed
on crew design or crew processes defined input-output or control methods for improving crew
performance and optimizing crews on the site. Production theories reviewed such as Sociotechnical, Lean Production, and Lean Construction theories formed the basis for developing the
guidelines for lean crew design. Also, simulation approaches such as discrete event simulation
and system dynamics modeling are presented in more detail together with discussing the
applications, advantages and limitations of each of the modeling techniques.
2.2

Construction as a Complex System

The real world is a world full of complexity, chaos, uncertainty, change, flow and processes and
cannot be pinned down to the simple, solid, certain, uniform or unchanging objects (Merry,
1995). And construction industry is considered to be one of the most complex systems existing in
the real world environment. It has been shown that construction system and the processes
involved in construction projects are highly complex and dynamic in nature (Wood and Gidado
14

2008; Bertelsen 2003a, 2003b; Antoniadis 2008; Baccarini 1996). In fact, the construction
process may be considered the most complex undertaking in any industry (Bennett, 1991) which
is also supplemented by Mills (2001) who describes the construction industry as one of the most
dynamic, risky and challenging businesses.
In order to deal with complex construction systems, it is first important to comprehend what
complexity actually means. At times complexity in general usage is confused for the term
complicated, for example, The Collins English Dictionary (2006) defines complexity as “the
state or quality of being intricate or complex”, or “something complicated” and the term complex
here is defined as “intricate or complicated”. But it is to be noted that complexity and
complicated do not mean the same, as complicated means something physical or social which is
difficult to understand, analyze or manage for one person which may not be true for another
person.
The study of complexity and complex systems holds much scope for modern problems where
traditional practices have struggled. However, the study of complexity is still relatively recent
and it is difficult to reach a standard definition of complexity (Wood and Gidado, 2008).
Kauffman (1995) has stated: The efforts are still so new that there is not yet even a generally
accepted, comprehensive definition of complexity. Williams (1999) states: While many project
managers use the term complex project, there is no clear definition about what is meant – beyond
the general acceptance that it is something more than simply a ‘big’ project.
However, a generally accepted definition of complex systems can be presented as systems which
are composed of interconnected parts that exhibit one or more properties or behaviors may not be
obvious from the individual parts.

15

Williams (1999) proposed that the overall project complexity can be characterized by two
types of uncertainty such as “structural uncertainty” in terms of the number of elements involved
and interdependence of these elements and “general uncertainty” which could exist in the form
of uncertainty in goals and uncertainty in methods adopted. Williams (1999) classification was
very similar to the one given by Jones and Deckro (1993) who defined technical complexity in
projects in terms of variety of tasks, the degree of interdependencies within these tasks, and “the
instability of the assumptions upon which the tasks are based”.
Baccarini (1996) proposes a definition of project complexity as “consisting of many varied
interrelated parts and can be operationalized in terms of differentiation and interdependency.”
The terms differentiation and interdependency are related to the number of varied elements
present and the degree of interrelatedness between these elements. He discusses these concepts in
terms of two types of project complexity such as organizational complexity and technical
complexity. Differentiation in an organizational complexity would mean the number of levels in
the organizational hierarchal structure, number of formal organizational units, division of tasks,
number of specializations, etc, whereas in technical complexity it would mean variety or
diversity of some aspect of a task, such as number and diversity of inputs and/or outputs, number
of separate and different actions or tasks to produce the end product of a project, and number of
specialties (subs or trades) involved on a project.
Gidado (1996) conducted a structured interview with selected experts in building industry to
determine what they consider as project complexity and their views and opinions have been
presented below:
1. That having a large number of different systems that need to be put together and/or that
with a large number of interfaces between elements;
16

2. When a project involves construction work on a confined site with access difficulty and
requiring many trades to work in close proximity and at the same time;
3. That with a great deal of intricacy which is difficult to specify clearly how to achieve a
desired goal or how long it would take;
4. That which requires a lot of details about how it should be executed;
5. That which requires efficient coordinating, control and monitoring from start to finish;
6. That which requires a logical link because a complex project usually encounters a series
of revisions during construction and without interrelationships between activities it
becomes very difficult to successfully update the program in the most efficient manner.
Form the above definitions, complexity as defined by Gidado (1996) was divided into two
perspectives of project complexity: (1) The managerial perspective, which involves the planning
of bringing together numerous parts of work to form work flow and; (2) The operative and
technological perspective, which involves the technical intricacies or difficulties of executing
individual pieces of work.
Also, from the structured interviews conducted, Gidado (1996) determined that complexity
factors that affect managerial actions in construction originate from a number of sources which
can be categorized into two types: (1) Category A, which deals with the components that are
inherent in the operation of individual tasks and originate from the resources employed or the
environment and; (2) Category B, which deals with those that originate from bringing different
parts together to form a work flow. Grouping the sources of complexity factors affecting
managerial actions forms a clear distinction between the various originating sources such as
employed resources, the environment, number and interaction of different parts in the work flow

17

etc, which makes it much easier to manage, control or reduce complexity, once we know from
where it is originating.
Gidado (2004) also deduced that the project complexity can be divided into six main
components, each of which is made up of various intersecting factors. The six main components
to project complexity are: (1) Inherent complexity; (2) Uncertainty; (3) Number of technologies;
(4) Rigidity of sequence; (5) Overlap of phases or concurrency; (6) Organizational complexity
Also, Sven Bertelsen (2003) discussed how the three group of characteristics of complexity such
as autonomous agents, undefined values and non-linearity may be directly related to the three
different perspective of construction process such as transformation, value generation and flow
provided by Koskela (2000).
According to conventional production theory, production is simply considered as the
transformation of inputs to outputs, which ignores the flow and value generation concept.
However, according to Koskela (2000), three different conceptualizations should be
simultaneously used: production as transformation, flow and value generation. By using this
conceptualization, the transformation model is in an auxiliary position, and the other two flow
and value generation models are successfully able to address both the time-dependent and the
time-independent complexity (Pennanen and Koskela 2005).
The proceeding literature seems to suggest how the construction industry and the processes
involved are considered to be quite complex. The construction environment is affected by many
dynamic and uncertain variables, such as weather, space congestion, crew absenteeism,
regulatory requirements, and design changes and reworks (Ahuja and Nandakumar, 1985). These
factors along with others such as project size, technical tasks involved, interactions among the
resources (equipment, labor and material), team structure, work structure etc lead to even more
18

complexity in construction. This high complexity may lead to difficulty in comprehending the
interaction between the various processes involved in construction.
2.3

Existing Production Planning Models

The successful planning of the production process is critical to the success of any construction
project; it has to address strategies for site and crew set-up, and a whole host of policies,
methods, and procedures (Gidado, 1996). While project management focuses on macro level,
strategic issues (Lee et. al. 2006; Pena-Mora et al. 2008), production (or operation) management
focuses on the micro level.
Project management can simply be defined as the “process that takes into account a
company’s way of doing business, allowing for the possibility of a significant payoff with fewer
risks” (Callahan & Brooks, 2004) whereas, production management can be defined as the
management actions incorporated to meet a project’s target by adjusting time, cost, and resources
(Lee et. al. 2006). The basic difference between the two is that project management is the use of
skills, tools, and management techniques incorporated together to meet company’s goals and
objectives while production management deals with the skills, tools and actions employed to
deliver the project objectives.
Since the early 1970’s, researchers and practitioners have developed various construction
models to deal with both project and production issues in construction. Kartam et al. (1997)
suggested that the existing models could be divided into types of construction models such as
system models and process models depending on whether the models emphasize whole system
or the on-site construction process. Kartam et al. (1997) differentiated between process models
and system models as:

19

"A process is defined as a set of consecutive steps or activities with an end product or service
being delivered … On the other hand, a system model neglects the steps that constitute the
process; instead it focuses on the process environment. Portraying the process with its inputs,
outputs, directives, feedback loops, and interactions with other processes is a major objective of
system modeling. Thus, a process model gives a different perspective and accordingly serves a
different purpose than does a system model".
In the following section, the existing process and system models present in the construction
literature are discussed in more detail. We discuss how these models correlate with the crew
design process and operations design.
2.3.1

Entity Interdependence Model

Entity Interdependence Model was proposed by Thompson (1967), which is part of one of the
best known work on organization theory containing the most widely accepted typologies of
technology in management. The term technology as used by Thompson refers to the work flows
or the transformation processes. In this model, the interdependencies existing between the
various entities were explicitly explained. His classifications are based on flow of work or how
the work is to be divided between individual members of the work group. Thompson identified
three types of interdependencies such as: 1) pooled (no coordination), 2) sequential (simple
coordination), and 3) reciprocal (complex coordination).
These three types of interdependence existing between the entities are shown in Figure 2.1.
Pooled interdependence requires little to no coordination between the entities. Pooled
interdependence means that an entity A or B or Z would make a discrete contribution to the
whole; each entity would be involved in accomplishing individual tasks, and the collective
output of these entities is what constitutes production for the unit/project. Sequential
20

interdependence is a one way directional interdependence with simple coordination, which
means each entity has to accomplish their individual tasks before the next or any other entity in
the work unit could start their task/work. This means that output of an entity C becomes the input
for the next entity D and the first entity C must finish before the next entity D could start.
Reciprocal interdependence is a two way directional interdependency requiring complex
coordination, which means two entities such as E & F could be dependent on each other; the
output of each of the entity could be the starting point for the other, meaning that both the
entities are dependent on each other to finish. The knowledge of these interdependencies is very
important for the purpose of coordinating the work units/crews working on the construction site.
A

C

D

2) Sequential Interdependence

B
1) Pooled Interdependence

3) Reciprocal Interdependence
E

F

Figure 2.1: Entity Interdependence Model, Thompson (1967).
2.3.2

Conversion Model

One of the focuses of the past research in traditional project management has been to follow the
transformation theory converting the inputs into outputs. The dynamic system model (Figure 2.2)
defined by Alexander (1974) best describes this conversion process where he showed that the
inputs are converted into output by going through a process and this process is regulated by a
controller, which establishes the operating characteristics of the process.

21

Regulate Inputs

Feedback
CONTROLL
Operating

INPUT

PROCESS

OUTPUT

(Material/Labor)

(Products)

Sub-Process

Sub-Process

Figure 2.2: Conversion Model adapted from Alexander (1974), Koskela (1992).
According to Kartam (1995), Alexander’s conversion model is more appropriate to describe
the construction process than other static models such as bar charts, PERT diagrams or decision
trees. This model was mostly dominant in the manufacturing setting; and was extended by
Koskela (1992), where he stated that production process was aimed at maximizing the
conversion of inputs such as material and labor into outputs such as completed units of work;
where the work was further divided into sub tasks and activities; again seen as a conversion
process. In the conversion model, it was assumed that the cost of any process could be
minimized by minimizing the cost associated with respective sub-processes.
Walker (1985) extended the conversion model provided by Alexander (1974) and applied it
to the construction process. He divided the transformation process in two groups, namely the
construction process and client process; where each of them, while converting inputs to outputs,
is affected by various factors from the environment. The model was sufficient and seemed

22

acceptable to show the construction process but only in terms of transformation of processes. It
neglected the flow of material and information and value generation components associated in a
construction process, which Koskela advocated in his work (Koskela 1992). Further, it did not
explain how the construction processes are interdependent with considerable performance
variation. The conversion model developed by Walker (1985) is shown below in Figure 2.3:

Environmental Forces

To Environment

TRANSFORMATION

INPUT

Finish Point

Start Point

Environmental
Forces

OUTPUT

To Environment
INPUT

TRANSFORMATION

OUTPUT

Figure 2.3: Conversion model, Walker (1985)
2.3.3

Method Productivity Delay Model (MPDM)

Adrian and Boyer (1976) introduced the method productivity delay model (MPDM). This model
was developed for application by a small to mid-size construction firm that cannot afford
professional services. Simple statistical methods were used in this model to make it accessible to
construction personnel. It was developed to measure, predict and improve the productivity of the
construction operations. For this purpose, this model incorporated elements of other techniques
such as work sampling, production function analysis, statistical analysis, time study, and
balancing models (Halpin and Riggs 1992).

23

The MPDM measures, predicts and improve the productivity of the construction operations
in four stages, namely, data collection, data processing, model structuring, and finally model
implementation. The construction delays were classified into five major types: environmental,
equipment, labor, material, and management. By collecting the data on these delay types
occurring in the construction operation, efficiency of construction operations is measured by
identifying the problematic areas lowering the productivity. MPDM does not consider the other
interactions and interdependencies occurring between crews on the site, which may lead to
optimizing processes but not the whole system.
2.3.4

Control Model

To address the limitations in the conversion model given by Alexander (1974), Sanvido (1984)
suggested a control model that addresses the dynamic construction environment and identifies
eight major activities that constitute on-site construction. This model described the various
important functions of the construction process and how to regulate or control them (Figure 2.4).
The main focus of this model was to take corrective actions once mistakes or inefficiencies
are encountered rather than taking preventive actions before their occurrences (Kartam 1995).
This model was significant as it established a hierarchical way to address and control the
construction process by highlighting the interdependence between the important on-site
construction activities and allowing the management to address the dynamic construction process
in a much better way.

24

Figure 2.4: Control Model Sanvido (1984).
2.3.5

Factor Model

Crew productivity has been an important factor in determining crew effectiveness and
construction operations performance, and in turn the overall project performance in achieving
intended aims (typically time, cost, and quality). A lot of factors act together which may affect
the productivity and hence the performance of construction operations. Thomas and Yiakounis
(1987) developed a factor model to capture this multi-factor influence. This model was

25

developed for labor-intensive construction activities based on the fact that work executed by a
crew is affected by a number of factors which may lead to incidental disturbances, leading to low
crew performance.

Figure 2.5: Actual productivity vs. ideal productivity curve, Thomas et al (1987).
According to Thomas et al (1987) these disturbances could be due to the following factors:
environmental factors, site factors, management factors, and design factors which may add up to
give an actual productivity curve that may be irregularly shaped and difficult to interpret. They
argued that if these disturbances can be mathematically discounted from the actual productivity
curve, one is left with an ideal productivity curve (Figure 2.5). Based upon design requirements
and construction practices, it is theorized that this ideal productivity curve can be predicted
before the start of work.

26

Further, based on U.N. Report ("Effect" 1965), Thomas et al (1990) extended the factor mode to
divide the factors affecting the labor productivity into two groups such as 1) Work environment
and 2) Work to be done. The work to be done and work-environment categories are analogous to
catalysts in the form of resources and conditions needed to efficiently convert inputs (workhours) to outputs (quantities) (Thomas et al 1990). The detailed representation of the factor
model is shown in Figure 2.6.

Figure 2.6: Factor Model, Thomas et al (1990).
This model is different from previous models as it focuses on productivity at the crew level
and mathematically calculates the actual productivity considering the most common factors
affecting the crew productivity. It is similar to the method productivity delay model where both
attempt to account for the factors which affect the productivity and then predict the actual
productivity. Both models attempt to improve the performance of crews by focusing on the crew
27

productivity but lacked the consideration of interdependencies existing among crews or the
variation in the flow of work within the crew.
2.3.6

Conceptual Construction Process Model

To overcome the limitations of the previous models, Sanvido (1988) extended the control
model (Sanvido, 1984) focusing on the functions that a crew must perform and the management
support needed to perform that particular work. The CCPM (Sanvido, 1988), shown in Figure
2.7, introduced the following features to improve the productivity and performance of site
construction operations: (1) definition of the basic tasks of the crew and the input resources
required by them; (2) identification of interrelationships between different functions involved in
supporting the field construction process and specification of rules to govern their performance;
(3) definition of the scope and boundaries of the on-site construction process; and (4)
categorization of external influences on the construction process that are beyond the control of
the site personnel. One of the other features where this model differs from the Alexander's
(1974) model is that here the control function is performed by the crew and supervisors and not
by some automatic controller. In the CCPM, the craftsman/crew worker has control over the
process of taking the inputs and converting them to outputs. The transformation of inputs into
outputs by the craftsman/crew worker is restricted by three main functions (Figure 2.7) such as
(1) Influences on the supply of resources, (2) Influences on planning, (3) Influences of project on
participants and environment.

28

Figure 2.7: Conceptual Construction Process Model, Sanvido (1988)
The CCPM presented the structure of functions that should be performed on a project. The
research showed that projects that better matched to the ideal case specified by the CCPM
perform better in terms of schedule, cost, and quality than those who did not (Sanvido 1988).
However, the research did not define the flow of resources for the operations, how these
resources would be shared among the crew.

29

2.3.7

Integrated Building Process Model

Sanvido et al (1990) developed a hierarchical model known as Integrated Building Process
Model (IBPM) to provide a generic representation of the building construction process from
initial conception stage to the operations stage. This model is based on the Structure Analysis
and Design Technique (SADT) which is a graphical language consisting of a set of boxes, each
of which represents a conversion process (Kartam et al. 1997). This model provides the primary
function “Provide Facility” which is divided into five main sub-processes such as Plan Facility,
Design Facility, Construct Facility, Operate Facility and Manage Facility defining the initial
hierarchical structure of this model. Each of the process consists of four main elements such as
input, control, mechanism and output representing the conversion process associated with that
process (Figure 2.8).

Figure 2.8: Schematic representation of the Function Box, Chung (1989)
The IBPM has its advantages over the other models such as conversion model (Alexander
1974, Walker 1985) and the control model (Sanvido 1984) in terms that it covers the building
construction process from initial conception stage to the operations stage in detail and also by
including control and mechanism elements apart from just the input/output elements. However it

30

provides a broad overview of the construction process and lacks the consideration for the subprocesses at the operational, activity or task level. Also, consideration for formation and
operational processes of the crew is not defined.
2.3.8

Flow Process Model

Koskela (1992) developed the Flow Process Model (Figure 2.9) where construction operations
were considered not just as a conversion of inputs into outputs but as flow of resources such as
material and/or information from the starting to the end product. The model was proposed to
cover all the important features of the production process and overcome the limitations of the
conversion model.

Figure 2.9: Flow process model, Koskela (1992)
According to Koskela (1992), in the production process, material and/or information goes
through a number of other functions/activities such as waiting, inspection, or movement apart
from the traditional view of being just processed (i.e. conversion or transformation); however
these functions/activities are inherently different and have not been addressed in the earlier
construction models. The processing activities fall under the conversion aspect of production
whereas inspecting, moving and waiting fall under the flow aspect of production. In the flow
process model, any activities other than processing, such as moving, waiting, inspection etc. are
31

considered as waste not adding any value to the project. The flow process model is a good model
for the representing the production process, however, it does not take into account the system
concept and lacks the feedback and interdependencies existing in the production process. Also, it
focuses more on the flow of resources such as material and/or information being processed but
the reasons for the variation in the performance of the crews and the reasons for the inherent
waste in the crews executing these activities hasn’t been discussed.
2.3.9

The Last Planner® System

The Last Planner® System was introduced by Ballard and Howell (1997) to overcome the
limitations of the conversion process models and traditional project management practices to
coordinate and control the field production process (Figure 2.10).

Figure 2.10: Traditional PM Vs. Last Planner system, Ballard and Howell (1997)
This system was implemented to stabilize the production work environment by reducing the
upstream variation and management uncertainty. Due to the huge gap, almost 50%, between the
32

number of planned activities that SHOULD be done to those which are actually DONE, Ballard
and Howell (1997) suggested that the traditional project management practice of planning &
controlling the work should be modified from “Should-Did” to “Should-Can-Will”. They argued
that differentiating between what SHOULD be done and what CAN be done is of importance in
order to improve the reliability of plans.

Figure 2.11: Last Planner® System, Ballard and Howell (1997)
Moreover, allowing the task performers, last planners, to commit to work that can be done, i.e.,
going from CAN to WILL, the opportunity to reflect the most up-to-date information about the
project at the production level instead of over committing to work that will not be completed yet
expected by downstream trades. In the Last Planner® System , Make-Ready (look ahead)
schedules are developed for the upcoming 5-6 weeks by selecting assignments/work packages
from the master/phase schedule that was prepared jointly by the construction manager and the
33

involved trade partners. The assignments from the make-ready schedule which are cleared with
respect to resources, directives and pre-requisite work constraints are then selected to be included
in the weekly work plans.

The rest of the assignments, enter a workable backlog to be

considered for next week work plans. At the end of each week once the work is executed, the
weekly work plan is then compared with the actual work completed to determine any reasons for
a low percent plan complete (PPC). PPC is defined as the ratio of the number of activities
completed to the total number of activities on the weekly work plan. This model has gained a lot
of acceptance due to its simple conception and successful implementations. It is a preferred
system to coordinate and control the field production process by stabilizing the work flow.
However, it still lacks the consideration for how the crews should be designed and the work flow
variation which could exist due to interactions existing within the crews.
2.3.10 Construction Process Analysis model (CPA)
This CPA model was proposed by Lee et al (1999) to identify and quantify waste in the
construction operations by analyzing the work processes. It implements process charts and topview flow diagrams common among process analysis techniques (Parker and Oglesby 1972).
The simple charts and diagrams used in this model are based on the standardized symbols
under Japanese Industrial Standards JIS Z 8206, effectively describing the flow of processes in
order to quickly determine the existing problem areas in the process Lee et al (1999). The basic
symbols used in the CPA model include Operation, Transportation, Storage, Delay, Volume
Inspection, and Quality. The process charts are used to record the steps of the construction
operation and the flow occurring within each unit, section etc. And flow diagrams are used to
supplement the process flow charts by observing the layout diagrams of the area in which that
process flows.
34

2.4

Existing Crew and Work Analysis

A significant amount of the research on crew and work analysis has been focused on productivity
studies and time and motion studies (work-study), work observation and activity sampling. The
focus of these studies has been to improve the performance of crews and optimize construction
operations. These studies determine the factors affecting the crew performance and then suggest
implementing corrective measures.
Productivity has been widely perceived to play a major role in affecting the performance of
crews. Higher productivity levels are considered to increase the performance of the crews and
hence productivity studies have been used to improve the low productivity within a crew.
Extensive amount of research (Adrian and Boyer 1976; Maloney 1983; Oglesby et al. 1989;
Thomas and Yiakoumis 1987; Diekmann and Heinz 2001; Dai et al 2009) has been conducted to
determine the factors affecting labor productivity. Listing these factors would easily run into
100 plus factors.
Adrian and Boyer (1976) classified construction productivity factors into five major types:
environmental, equipment, labor, material, and management. According to Maloney (1983),
labor motivation and performance factors have a significant influence on construction
productivity. He discussed how driving, induced, and restraining forces act on the construction
workers to affect their level of productivity. Maloney (1983) presented four major labor factors
influencing the productivity: (1) The intensity of the individual worker's effort; (2) the duration
of that effort; (3) the effectiveness with which the worker's effort is combined with technology
and other resources; and (4) the overall efficiency with which these inputs are translated into
usable outputs of acceptable quality.

35

Thomas and Yiakoumis (1987) developed a factor model to forecast the crew performance
by determining factors disturbing the crew performance. The factors affecting productivity were
divided into four groups: (1) Environmental factors; (2) Site factors; (3) Management factors; (4)
Design factors. The model was developed to improve productivity for repetitive construction
operations. The factor model recognizes that the labor productivity is not constant and varies
over time in a unique way but assumes it can be predicted because it takes different amount of
resources to complete different activities. Thomas and Sakarcan (1994) classified the factors
affecting construction productivity into two main groups: (1) Work to be done factors such as
work scope, work content, design features etc. and (2) Environmental factors such as congestion,
sequencing, equipment, material etc.
To estimate productivity rates with a higher degree of accuracy, Herbsman and Ellis (1990)
developed a statistical model to better illustrate the quantitative relationships existing between
the influencing factors and productivity rates. Productivity factors were divided into two main
groups such as technological factors and administrative factors based on the interviews with the
industry.
Liberda et al (2003) made a broad classification of factors in terms of human, external and
management issues that affect construction productivity and presented 51 major productivity
influencing factors. The factors presented were prioritized along with identifying measurement
criteria to improve future construction projects. Based on industry interviews, Liberda et al
(2003) identified management as the most influential factor affecting labor productivity. Dai et al
(2009) also developed factors affecting productivity from the construction craft workers
perspective by conducting a nationwide survey. The survey quantified the relative impact of 83
productivity factors showing that the factors involving tools and consumables, materials,
36

engineering drawing management and construction equipment have the greatest impact on
productivity. Latent factors representing the underlying structure of these 83 productivity factors
were identified in this study. The results were further compared to similar previous efforts
identifying significant differences that may impact future productivity improvement strategies.
Among other productivity studies, the effects of scheduled overtime on labor productivity
have been examined in the past (Oglesby et al. 1989; Thomas 1992; Thomas and Raynar 1997).
These studies described how the productivity data were collected, processed, and analyzed. It
was shown that as the number of hours of overtime increased to 50-60 hour weeks, there was a
10-15% loss in the efficiency which was comparable to other published data including Business
Roundtable (BRT) curves. Also, it was concluded that the losses of efficiency are caused by the
inability to provide materials, tools, equipment, and information at an accelerated rate.
Thomas and Napolitan (1995) and Hanna et al. (1999) documented the effect of changes in
construction on labor efficiency. It was found from these studies that lower labor efficiency is
strongly related to the presence of change work, disruptions and rework and on an average 30%
loss of efficiency occurs when change orders happen in the construction. Among the few
disruptions quantified such as sequencing, information availability, material availability,
equipment availability, and tool availability, lack of material availability was found to be the
most serious disruptions of all.
Thomas et al (1990) defined work study as the systematic study of work systems for the
purposes of finding and standardizing the least-cost method, determining standard times, and
assisting in training in the preferred method. Work study could also be called time-and-motion
study (Barnes 1980). A work study is composed of 2 main interrelated methods such as (1) a
work-methods study (also known as motion study) and (2) a work-measurement study (also
37

known as time study). The relationship between the two methods has been shown in the Figure
2.12.

Figure 2.12: Work Study Process, Drewin (1985)
The work-methods study or motion study is used to determine the best method of doing a
work/task whereas work-measurement study or time study involves determining the standard
38

time to perform a certain work/task. Both methods originate from industrial engineering and
involve in-depth examination of the production process (Thomas et al 1990). Thomas et al.
(1990) proposed three labor productivity models based on the work-study concepts to be
implemented at crew-level operations. These models were classified as delay model, activity
model, and task model. And to implement these models as labor productivity models, three
fundamental assumptions must be met (Thomas et al 1990): (1) productive time and delay times
are related and hence reducing delays will increase productive time; (2) productive time is
related to output and productivity and hence knowing productive time, production output can be
calculated; (3) waiting time is related to productivity and hence productivity is improved as
waiting time is reduced. The three labor productivity models proposed by Thomas et al (1990)
are discussed below:
1. Delay Model
The delay model is a very simple model. To quantify productive time and delays happening on
the site, a work day in this model is divided into three parts to record instances of delays.
Graphical representation of the delay model is shown in Figure 2.13.

Figure 2.13: Graphical representation of Delay Model (Bernard et al. 1973)
Thomas et al (1990) suggested that application of this model to most labor-intensive
processes is not feasible or difficult as the output-productive time relationship, even for simple
operations, is very complex.
39

2. Activity Model
To measure the time spent on various activities, work measurement technique such as work
observation, activity sampling or work sampling are employed. The activity model (Figure 2.14)
is a simple model based on these work measurement techniques (Thomas and Daily 1983;
Thomas et al. 1984; Liou and Borcherding 1986). This method can be applied to labor-intensive
activities to measure the time utilizations of the labor but it is only an indirect measure of the
actual productivity, since there is no measure of the output (Thomas 1981).

Figure 2.14: Graphical representation of Activity Model (Thomas et al. 1983)
Also, various studies have shown that activity model cannot be considered as a valid
productivity model, as the direct work time and output are unrelated to each other (Thomas 1981;
Thomas et al 1983). This was also supplemented by Rogge et al (1982) who presented data
showing that productivity and waiting time are only somewhat related. As work sampling
40

focuses more on the measurement of time utilizations for individual worker rather than a crew;
direct work can be considered as a measure of individual worker activity, and not the
productivity itself (Thomas et al 1983).
3. Task Model
The Task Model extends from the previous delay and activity models and was developed by
researchers at Loughborough University, England. This model introduced the concept of
dividing the work/tasks according to their importance or necessity; certain activities are
considered as basic or necessary, others are additional but necessary, and a third group are
unnecessary (Thomas et al. 1983).
A number of analysis methods have been implemented in construction for crew and work
analysis. Job analysis is an important aspect of any design or redesign process. Morgeson and
Campion (1997) defined job analysis as the number of systematic techniques for collecting and
making judgments about the job information. Information derived from job analysis can be used
to make informed decisions about the design process; can be used to make selection decisions,
determine training and development needs develop performance appraisal systems, as well as to
analyze various tasks and jobs (Morgeson et al 2006). According to Morgeson et al (2006), job
analysis may also focus on tasks, worker characteristics, worker functions, work fields, working
conditions, tools and methods, products and services etc. Some of the common methods for job
analysis presented by Morgeson (2006) are briefly discussed below:
1. Interviews
2. Questionnaire
3. Checklists
4. Critical Incidents Reporting
41

5. Work Observation & Activity Sampling
6. Variance Analysis
7.

Time & Motion Studies
•

Process Charts

•

Flow Diagrams

•

Possibility Guides

•

Network Diagrams

8. Linkage Analysis
Interviews with job experts such as the field personnel involving the supervisors and workers
are often the first step towards a better crew design. During such meetings, important
information such as duties, tasks, skills, knowledge is collected and used for optimizing the field
operations. The questionnaire format is adopted where information needs to be collected from a
large number of people. Questionnaire are constructed and used to collect judgments about the
way the work is being done and how could it be improved. Checklists are a quicker method of
getting job information from the field personnel.
Critical incidents method focuses on incidents which happen or might happen on the job site
affecting the construction operation performance. They emphasize on the worker behavior that
are ineffective or effective in the outcome of the operations. Work Observation & Activity
Sampling are important form of observation methods which at times needs to be implemented
where job analysis demands actual observation of the work performed. More sophisticated
methods involve statistical sampling of work activities. Variance analysis is a tool of sociotechnical design used to identify areas of uncertainty in a production process. The variance
concept is applied to technical systems and involves five steps (Davis and Wacker, 1982):

42

1. List variances that could impede the production or service process.
2. Identify causals relationship among variables.
3. Identify and focus on key variances whose control is most critical to successful outcomes.
4. Construct a table of key variance control that contains a brief description of variances.
5. Construct a table of skills, knowledge, information and authority needed so that workers
can control key variances.
Time and motion studies include tools such as Process Charts, Flow Diagrams, Possibility
Guides, and Network Diagrams help the design engineers visualize operations in the study or
design of job in order to improve efficiency. Process charts graphically represent separate steps
or event that occurs during performance of a task. Flow diagrams differ from process charts
because they utilize drawings in which an activity takes place. Lines between the symbols help to
show the path of travel. Possibility guides help in examining consequences of suggestions to aid
in selecting the most feasible changes. Network Diagrams are used where there are a lot of
complexities such as (a) dependencies are tangled, (b) output has many components, (c) process
is too complex. Finally, linkage analysis is used to determine the various linkages present among
the components in an operation/process. Links can be classified into three types such (a)
communication links, (b) control links, (c) movement links.
The brief review of the previous research on productivity studies is not meant to be
exhaustive here; rather the review aim is to demonstrate the state of productivity research
conducted to optimize crew performance. Productivity studies are appropriate to improve low
productivity issues in construction and they also contribute in the knowledge of the various
factors affecting the labor productivity or low labor efficiency.

However, the current

productivity models are implemented in isolation of each other, and at times only considering a
43

few or single productivity factors. In addition, these studies do not model the crew or provide
knowledge about how a crew should be choreographed in terms of crew structure, crew size,
crew multi-skilling, how the work-flow is to be designed within a crew.
Formation of a crew is a consequential process, and it should not be designed based on adhoc basis. A crew design model or guidelines based on lean principles such as pride in final
product, waste reduction, learning, labor utilizations, and pull systems should be explored.
2.5

Existing Production System Theories

2.5.1

Theory of constraints

Theory of constraints is an approach developed for the manufacturing industry for improving
manufacturing operations. It focuses on managing a process or a series of manufacturing or
production steps required to create a product. The Theory of Constraints (or TOC for short) also
evaluates how projects are managed as well as provide estimates of completion, amount of safety
stock included and where buffers should be located. The five basic TOC are (Goldratt 1997):
1. IDENTIFY the system’s constraints.
2. EXPLOIT the system’s constraints.
3. SUBORDINATE everything else to the above decision.
4. ELEVATE the system’s constraints.
5. GOTO step one again.
By following these steps, bottlenecks could be identified in a process. This bottleneck could
either be an activity or a resource (person or equipment). By identifying the bottleneck, the rest
of the production process is managed to the capacity of the bottleneck or the bottleneck’s

44

production capacity is increased thereby removing that bottleneck. The bottleneck activity’s
schedule is detailed to minimize any disruptions or impacts on downstream activities.
Using TOC results in less amounts of inventory pileup, and identification of where to focus
efforts to increase production capacity, remove bottlenecks and thereby maximizing the
performance operations. In the context of construction operations and the crews executing them,
operations would not be optimized unless performance of all the workers in the crew is increased
or at least limited to only a very small part of the crew. It is worth noting also that the workers
which are working or executing sequential operations with interdependencies cannot finish the
whole operation or work faster than the slowest worker in that crew. And if capital is spent on
improving activities that are not bottlenecks, it will not optimize the system but only lead to
waste.
2.5.2

Lean Production Theory

Lean Production is a production practice developed for manufacturing, primarily aimed at
shortening the time from order to delivery by eliminating waste in the entire production system.
Any expenditure of resources which does not add value for the customer is considered waste.
Lean Production as presented by Womack et al. (1990) states that “Lean Production is ‘lean’
because it uses less of everything compared with mass production: half the human effort in the
factory, half the manufacturing space, half the investments in tools, half the engineering hours to
develop a new product in half the time”. The principles of Lean thinking as presented by
Womack and Jones (1996) are:
1. Specify value for the customer
2. Identify the value stream, and remove non-value added steps
3. Make the remaining value adding steps flow
45

4. Let the customer pull the product
5. Pursue perfection through standardization & continuous improvement
Lean production method has revolutionized operations not only for the manufacturing
industry but also for other production and service industries such as construction, banking,
insurance, and medical.
2.5.2.1 Toyota Production System
Lean production philosophy was derived mostly from the Toyota Production System (TPS)
(Shingo 1989; Ohno 1988). The simple concept of elimination of unnecessary inventory and
waste to simultaneously achieve high quality, low cost and deliver maximum value to the end
customer is at the heart of TPS (Monden 1983; Shingo 1989; Ohno 1988).
According to Taiichi Ohno, the main focus of the TPS was “the absolute elimination of
waste,” where waste is anything that prevents the value-added flow of material from raw
material to finished goods. Also, according to sensei Shigeo Shingo (1989) the significant
principle and unique feature of the Toyota Production System lies in the following:
1. Eliminating unnecessary inventory.
2. Total elimination of waste
3. Relentless efforts made to cut man-power costs.
The TPS aims to reduce overburden (muri) and inconsistency (mura), and eliminate waste
(muda) in the production system. The seven commonly defined wastes in the Toyota Production
System (TPS) are:
1. Over-production;
2. Waiting;
46

3. Transport;
4. Over-processing;
5. Unnecessary inventory;
6. Unnecessary motion;
7. Defects.
Overproduction is considered to be the most serious and worst waste of all, as it leads to the
disruption of continuous flow of products or material. Over-production encourages the push of
products down the line, which results in extra in-process inventory and excessive work-inprogress. As a result, more pressure on the work rate generates leading to overburden (muri) and
inconsistency (mura). In addition, overproduction may also lead to more number of defective
products down the line.
Waiting takes place whenever products or material are not being processed or worked on.
Also, the idle time spent by the workers waiting for the machine to complete an operation or
waiting for the next batch of material to process is considered as a waste.
Transport, involves the movement of materials, labor or equipment over distances in a
production process. Transportation does not add any actual value to the final product, and hence
any unnecessary or excessive transport should be considered as a waste. Unnecessary transport
adds extra time to the operations and therefore should be avoided.
Over-processing occurs when extra steps or processes are employed to carry out an operation
than required to deliver value. In addition, it involves any complex solutions used where simple
solutions would work. Any non-value added steps in an operation are considered as a waste.

47

Unnecessary Inventory involves any raw material or work-in-progress that is not being
processed or has not been converted to a finished product providing value to the end customer.
Excessive in-process inventory may lead to overburden among the workers and low quality. In
addition, excess inventory requires more transportation and more storage space and may get
damaged before being processed.
Unnecessary Motion involves any movement of labor, equipment or material which does not
add value to the product directly. The movements of the workers to process a material or while
using equipment should be simple and any complex movements which may lead to poor
productivity or low quality of the product should be avoided.
Defects are the direct source of investing extra capital to re-do the work. They could occur
due to various problems or due to the presence of other types of wastes in the system.
Further, apart from these seven common types of waste (non-value added waste) known as
“muda”, TPS also identified the negative impacts of “mura” and “muri”. Mura results from
inconsistency or unevenness in the system and “muri” occurs when there is overburden on the
labor or equipment. Presence of inconsistency or unevenness in the system may also lead to the
need of keeping extra buffers in the system to encourage continuous flow and avoiding any
disruptions to the work. Also, as compared to the “muda” of idleness or waiting, “muri” creates
the potential for safety or quality failures as people or machines are pushed beyond their limits
(Lichtig et al 2005).
In TPS, “process” and “operation” have very specific meanings. Knowledge of how to
improve a process or operation is of value in order to design a crew executing a job. In TPS, as
defined by Shingo (1989):

48

“a process is a flow of material in time and space; a transformation from raw material
to semi-processed components to finished product whereas, operations are the work or
actions performed to accomplish this transformation; the interaction and flow of
equipment and operators in time and space”.
Process Improvement
To maximize production efficiency, TPS emphasizes analyzing and improving the “process”
before improving the “operations”. Shingo (1989) defines four distinct types of “process”
elements: (1) processing, (2) inspection, (3) transport, and (4) delay.
“Processing” improvement is achieved by first redesigning the product to reduce
manufacturing costs while maintaining the required quality and then improving the
manufacturing process of the product itself.
Three types of “Inspections” are defined in TPS in order to maintain quality, reduce
defective products and save time and capital on rework. These are successive inspection, source
inspection and enhanced inspection. Successive inspection involves workers inspecting the
products passed to them from previous operation before processing them themselves;
implementing successive inspection on average could reduce the number of defects by 80-90%
Shingo (1989). Source inspections are also of value as they help in preventing the defects by
controlling the conditions at their source. Finally, enhanced inspections are best at providing the
fastest feedback; they involve the use of devices for automatically detecting the defects. These
devices are also known as “poka-yoke” or ''mistake-proofing'' devices. The three types of “pokayoke” control methods identified in the TPS (Shingo 1989) include:

49

1. Contact method - identify defects by whether or not contact is established between the
device and some feature of the product's shape or dimension
2. Fixed value method - determines whether a given number of movements have been made
3. Motion step method - determines whether the established steps or motions of a procedure
are followed
The above inspection methods identified in the TPS leads to up to 100% reductions in the
number of defects occurring in an operation, faster feedback for the mistakes, reduction in the
amount of rework required and hence saves both time and capital. Therefore, while designing a
crew, consideration for incorporating these inspection methods should be made in order to
improve a “process” and hence improving crew performance.
TPS identifies the need to reduce process element such as “Transport” as much as possible,
as it does not add any real value to the finished product. Shingo (1989) discusses how improving
the layout of the process accounts for real elimination of the transport function. He explains that
“transport improvement” and “transport operation improvement” are two distinctly different
problems and that unavoidable transport work should be improved through mechanization only
after opportunities for layout improvement have been exhausted.
Finally, Shingo (1989) presents three storage types that lead to unnecessary “Delays” in the
process. Process delays occur due to unprocessed items waiting to be processed and also due to
the accumulated extra inventory that is waiting in line to be processed or delivered. To improve a
“process” following accumulations between the processes should be reduced or eliminated:
1. E (engineering) storage – this results from the unbalanced flow between processes and
can be eliminated by leveling the flow of items to match the output of all the processes.

50

Elimination of this storage would lead to continuous flow and reduction of any inventory
piling up.
2. C (control) storage – this results from the need of buffer or cushion stocks to avoid any
delay in the subsequent processes which may happen due to failure of machine or
equipment and any rejects or major defects that may happen down the line. This storage
can be reduced by employing better inspection processes to avoid any rejects or defective
work.
3. S (safety) storage – this results from the presence of over-producing beyond what is
needed for the control storage. Safety stock is useful to guard against any scheduling
errors, delay in delivery, damage to raw materials etc.
Operations Improvement
In TPS, “operations” are defined as the work or actions performed to accomplish the
transformation of raw material to semi-processed components to finished product. It involves the
interaction and flow of equipment and operators in time and space (Shingo 1989). In TPS, the
operations are improved by reducing or eliminating the waste present among the workers and the
equipment.
For example, waste of overproduction could be due to over staffing of workers and
producing more than required; this waste can be reduced by better managing the workers and
designing a crew in a way not leading to this waste. Also, waste of over-processing among the
workers and equipment can be reduced by eliminating any extra or complex steps that workers or
the equipment might be doing while executing the operations. Further, reduction of unnecessary
motion or transport would also lead to improving the operations. As the cost of workers waiting
is generally more than the idle machine or equipment cost, TPS aims to reduce labor cost by
51

improving worker motions, improving machine or equipment motions, and mechanizing worker
motions and thereby improving the overall “operations”.
In addition to removing overburden (muri) and inconsistency (mura) and waste (muda) to
improve the “operations”, Shingo (1989) mentions that the easiest way to improve the operations
is to separate the workers from the machine; following the "one worker, many process" theory.
TPS also emphasizes the concept of “multi-skilling” where one worker is not fixed to a single
machine or process, but is attending or operating other machines also. This is also supported by
the use of the “Nagara System” in TPS, where a worker performs more than one process at a
time by automating a number of processes. This results in reducing the waiting time, increasing
the productivity and reductions in labor cost. Hence “multi-skilling” is an important aspect for
improving operations and should be considered while designing crews for operations.
Further, various research efforts (e.g. Ohno 1988, Shingo 1989; Koskela 1992; Howell 1999;
Liker 2004) have defined and listed the lean production principles in different formats and level
of detail. For the purpose of this research, the most recent and simple grouping of the lean
production principles were considered and combined together to form crew design guidelines.
2.5.2.2 Application to Construction
2.5.2.2.1 Lean Construction
The application of above discussed production theories such as Theory of Constraints (Goldratt
1997) and Lean Production (Ohno 1988; Shingo 1989; Womack et al. 1990; Womack and Jones
1996) has demonstrated significant improvements in the performance and quality for the
manufacturing industry. Lean production has caused a revolution in the manufacturing and when

52

this production theory is applied to construction, it changes the way work is done throughout the
delivery process (LCI 2007).
The term "Lean Construction" was coined in 1993 by the International Group for Lean
Construction in its first meeting (Gleeson et al. 2007). According to Lean Construction Institute,
“Lean Construction is a production management-based approach to project delivery -- a new way
to design and build capital facilities”. Lean construction is a “way to design production systems
to minimize waste of materials, time, and effort in order to generate the maximum possible
amount of value” (Koskela et al. 2002).
Another definition by Abdelhamid (2004) states that “Lean Construction is holistic facility
design and delivery philosophy with an overarching aim of maximizing value to all stakeholders
through systematic, synergistic, and continuous improvements in the contractual arrangements,
the product design, the construction process design and methods selection, the supply chain, and
the workflow reliability of site operations”.
As noted by Abdelhamid (2008), “while Lean Construction is identical to Lean Production in
spirit; it is different in how it was conceived as well how it is practiced”. Also, the inherent
characteristics of the construction industry have been identified as a challenge to transfer all the
concepts of lean production to construction (Tommelein 1998).

However, both the lean

production and lean construction conform to the Socio-Technical construct (Niepce and
Molleman 1998) where both the human and technical elements work together to ensure higher
performance outcomes (Moore 2002). Lean Construction is the result of combining the existing
research in construction with an adaption of Lean Production principles and practices.
Despite the challenges to implement Lean Production to construction as a whole, certain
parallels can be made between the two industries. Similarities can be found in the production
53

process of manufacturing and construction. The conceptual basis for construction management is
the same as that of manufacturing, i.e. it is conversion oriented (Koskela 1992). For example, in
both settings, conversion of raw materials into finished products by expenditure of resources
over certain period of time takes place.
As in manufacturing, the conversion process in construction also generates a lot of waste,
making possible opportunities to apply the lean production principles to generate value by
eliminating the inherent waste in the process. This is supplemented by two similar characteristics
identified by Sanvido and Medeiros (1990): (1) they both produce engineered products that
provide a service to the user; (2) they both may include processing of raw materials and
assembly of many diverse pre-manufactured components in the final product.
The conventional theory of production and operations management both in manufacturing
and construction has been based on conversion of inputs into outputs or the transformation view
(Koskela 1992). However, in manufacturing since the adoption of the new production philosophy
based on the transformation and flow perspective embodied in Lean Production (TPS),
significant improvements in performance have been realized.
On the other hand, the construction industry is still facing an ongoing challenge of improving
the performance of projects at the production level. Koskela’s seminal 1992 report argued that
conventional Construction Management has focused for many years on improving project
performance by primarily focusing on project management techniques that rely on
transformation-based philosophy or the conversion model where the main focus is on
transforming inputs to outputs with typically no management of the transformation process itself
(Abdelhamid 2004).

54

Production based on the transformation-based philosophy or the conversion model can be
defined as follows (Koskela 1992):
1. A production process is a conversion of an input to an output.
2. The conversion process can be divided into sub-processes, which are also conversion
processes.
3. The cost of the total process can be minimized by minimizing the cost of each subprocess.
4. The value of the output of a process is associated with costs (or value) of inputs to that

process.
However, Koskela (1992) states that “the conversion process model, as applied to analyze
and manage productive operations, is misleading or even false.” This is supported by the
arguments provided by Ohno (1988) and Shingo (1989) that the conversion model neglects the
overall flow and value generation in a process. In the conversion model, physical flow between
the conversions is not considered; even the non-value-adding activities such as waiting, moving
and inspections, etc., are considered inseparable from the conversion process.
To overcome the limitations of the conversion model, Koskela (2000) presented the ‘TFV’
theory of production (Table 2.1) where the production was conceptualized in three
complementary ways: (1) Transformation (T) of inputs into outputs; (2) Flow (F) of materials
and information; (3) Value (V) generation for the customers. According to Koskela (2000), to
improve productivity and optimize production, it is important to consider all aspects of
production (transformation, flow and value).

55

Table 2.1: Integrated TFV view of production Koskela (2000)
Transformation view

Flow view

Value generation
view

Conceptualization of
production

As a transformation of
inputs into outputs

As a flow of material
composed of
transformation,
inspection moving and
waiting

As a process where
value for the
customer is created
through fulfillment of
his requirements

Main principles

Getting production
realized efficiently

Elimination of waste
(non value-adding
activities)

Elimination of value
loss (achieved value
in relation to the best
possible value)

Methods and practices Work breakdown
structure, MRP,
(examples)
Organizational
Responsibility Chart

Continuous flow, pull
production control,
continuous
improvement

Methods for
requirements capture,
Quality Function
Deployment

Practical contribution

Taking care of what
has to be done

Taking care that what
is unnecessary is done
as little as possible

Taking care that
customer
requirements are met
in the best possible
manner

Suggested name for
the practical
application of the
view

Task management

Flow management

Value management

Further, conventional project management in construction is inadequate because it relies on
the transformation theory or the conversion model rather than the TFV theoretical framework
(Howell and Koskela 2000; Koskela and Howell 2001).
In the conventional scenario, a project is broken down into activities by using the work
breakdown structure. Then, time and cost of each of these activities is calculated. Using criticalpath network methods, the logical order of performing these activities is determined. In this
scenario, all the activities are considered as value-adding activities which may not be true in
practice. Activities are typically scheduled to start on the earliest possible start date and then
project controls is applied to keep the schedule on track.
56

The practice as described completely ignores the flow of work within and between different
activities and that controlling by tracking activity completion and accrued costs fails to assure
reliable work flow (Howell and Ballard 1996; Koskela et al 2002). In addition, value is assumed
to be delivered to the customer by delivering the project on time and within initial budget and
providing the customer with the specified quality level. A notion of value delivery based on
extending and expanding the construction client’s needs is advocated in Lean Construction.
According to Koskela et al (2002):
“from the Lean Construction perspective, current practice rests on a defective
model of the project, the work involved, and its control … it fails even in the attempt to
manage activities and misses entirely the management of work flow, and creation and
delivery of value”.
The preceding observations have led to the birth of Lean Construction as a discipline that
subsumes

the

transformation-dominated

contemporary

construction

management

and

acknowledges the flow and value generation character of production (Abdelhamid 2004).
In Lean Construction, production system design serves the three main goals:
1. Produce a product in the most efficient way (Task management).
2. Eliminating waste (Flow management).
3. Maximizing value (Value management).
A main premise in achieving all three goals is the necessity of having reliable workflow.
Arriving at high work flow reliability, a conceptual framework titled Lean Project Delivery
System (LPDS™) was developed by the Lean Construction Institute (Ballard2000) to guide the
implementation of Lean Construction on project-based production systems. Although all the

57

parts of LPDS™ are crucial for success of a lean project, the heart of LPDS™ lies in Work
Structuring and Production Control (Abdelhamid 2008).
Workflow is the movement of information and materials through the networks of production
units (Ballard 1999; Ballard and Howell 1994). Koskela (1992 and 2000) provided some general
principles for how a ‘flow’ process should be designed, controlled and improved in practice.
Certain principles relevant to the crew design process are:
1

Reduce the share of non value-adding activities.

2

Reduce variability.

3

Reduce the cycle time.

4

Simplify by minimizing the number of steps, parts and linkages.

5

Increase output flexibility.

6

Increase process transparency.

7

Balance flow improvement with conversion improvement.

In Lean Construction, the concept of work structuring has been used to refer to the design of
the production system; which is designed to “achieve the purposes of both their customers and
those who ‘deliver’ the system” (Ballard, et al., 2001). Work structuring is the most fundamental
level of process design, answering the questions (Ballard 1999):
1. In what chunks will work be assigned to specialist production units (PUs)?
2. How will work chunks be sequenced through various PUs?
3. In what chunks will work be released from one PU to the next?
4. Where will decoupling buffers be needed and how should they be sized?
5. When will the different chunks of work be done?

58

Work structuring encourages the design of the production process during the product design
stage. Lean work structuring can be “accomplished using a myriad of tools such as BIM, Target
Costing, set-based design, CPM, simulation, etc.” (Abdelhamid 2008). Further, work structuring
is a complex process and includes the following sub-processes (Ballard 1999) such as: (1)
chunking; (2) sequencing; (3) releasing; (4) decoupling; (5) scheduling.
In the following section, lean principles and concepts aimed at the elimination of waste
(muda), variability (mura), overburden (muri) and maximization of value in construction are
reviewed.
2.5.2.2.2 Waste (Muda)
In a production process such as in construction, non-value-adding activities (waiting, moving,
inspection etc.) have a considerable share and reducing them to improve the production process
has been discussed at length in various studies (Koskela 1992; Koskela 2000; Ohno 1988;
Shingo 1989; Womack and Jones 1996; Liker 2004). Further, Ballard et al (2001) extended the
categorization of waste into defective products, lack of flow, lost capacity, and avoidable cost
apart from the waste generated by non-value adding activities and the activities other than
processing.
2.5.2.2.3 Variation (Mura)
Variation can be defined as the occurrence of output differences due to inconsistencies in the
work methods. It typically exists when workflow is out of balance and workload is inconsistent.
In simple terms, having zero variation would mean that the production process is completely
reliable with a continuous flow of work whereas, high variation would lead to disruptions in the

59

production process, creating unnecessary waste (Tommelein et al 1999, Abdelhamid and Everett
2002).
In the construction industry, lack of materials, late delivery, equipment breakdowns, tools
malfunctions, defects, improper sizing & composition of the crew, improper crew utilization,
schedule compression, etc., lead to variation and hence hampers the reliable work flow.
For example, the impact of work flow variation on the performance of construction trades
has been successfully illustrated by “The Parade Game” (Tommelein and Howell 1999). It is a
simulation of a construction process involving single line of processing stations where products
output by one are input required by the next one.
The game shows that reduction of waste and shortening of project duration could be achieved
by reducing the variation in work flow between trades. It was shown that variations leading to
unreliable work flow results in two kinds of waste: (1) Production stations cannot realize their
full production capacity due to starvation of resources; and (2) Intermediate buffers are larger
when high variation prevails. Finally, it was also deduced that high variations in flow would
result in reduction of throughput, delayed project completion, and increased waste.
It is generally accepted that variation in construction activity durations or production capacity
may lead to increases in both the volume of non-value-adding activities and the cycle time
(Koskela 1992). In addition, “variation within a production process affects the performance and
work flow reliability between the different processes and in turn affecting the overall project
performance,” (Abdelhamid and Everett 2002). It is also worth noting that variation affects
reliable workflow in any system that has processes and operations – whether it by a production
system or a product design system.

60

A significant amount of research has been conducted on how to stabilize and improve work
flow and the effects of flow variation in the production process (Howell and Ballard 1994;
Ballard and Howell 1994; Ballard 1999; Ballard and Tommelein 1999; Tommelein and Howell
1999; Horman et al 2002; Liu and Ballard 2009). One of the essential features of these studies
was how to ensure a continuous and reliable work flow. Improving work flow reliability is
important for the productivity of linked production units, and consequently for project cost and
duration (Liu and Ballard 2009). In a continuous flow process (CPF), work moves from one
work station/production unit to another in a linear sequence. The main idea was to balance the
processing rates of different work stations/production units to ensure a continuous and reliable
work flow while only keeping a modest amount of work-in-process (WIP) in between production
units. Continuous flow was achieved by reducing the resource idle time and WIP resulting in
maximum overall throughput.
Ballard and Tommelein (1999) presented the following essential elements in balancing a
production system to achieve continuous and reliable flow:
1. Adjust the crew sizes along with equipment and methods;
2. Redesign or restructure the operation;
3. Locate and size the in-line inventory buffers (output/in-process inventory)
4. Locate and size inventory buffers of delivered materials;
5. Locate and size capacity buffers.
Another study conducted by Zabelle and Ballard (1999) presented four simple tools to
improve performance of fabricators serving the AEC industry through a flow driver process: (1)
work mapping; (2) analysis; (3) compression and (4) prototyping. Steps to implement the
compression tool are listed below:
61

1. Expand bottleneck capacities to increase system capacity.
2. Establish decoupling buffers before and after bottleneck processes to take best advantage
of available system capacity
3. Implement pull techniques such as kanban or conwip to reduce work-in-process (WIP).
4. Reduce rework by implementing poka-yoke (mistake proofing) techniques.
5. Reduce transfer batch sizes.
6. Reduce process batch sizes where appropriate.
7. Organize production in cells using group technology and multi-skilled workers.
2.5.3

Sociotechnical System Theory

‘Sociotechnical’ (one word only) or ‘Socio-Technical’ (two words separated) is composed of two
main words: (1) Socio (of people and society) and (2) technical (of machines and technology).
Both variations mean the same; so should it be Sociotechnical System or Socio-Technical System
theory? Walker et al (2008). Both terms either Sociotechnical (Cherns 1976; Emery 1978;
Pasmore 1988; Niepcel and Molleman 1998; Clegg 2000) or Socio-Technical (Herbst 1974;
Bostrom and Heinen 19777; Cummings 1979; Trist 1981) have been continuously used in
organizational development literature and in practice can be been inferred as the same. But the
precise meaning of each one of them and if there is a difference among the two remains unclear.
The term “Socio-Technical System” was created by researchers Eric Trist, Ken Bamforth and
Fred Emery (Trist & K. Bamforth 1951; Trist 1960 and Emery) with a background in the
behavioral sciences at the Tavistock Institute in London around the end of the 1950s. During the
first decades of our century, a large amount of labor studies were conducted largely concerned
with the adaptation of humans to the organizational and technical framework of production;
however, there was a growing resistance of the workforce to the new technological systems.
62

The above researchers suggested that an organization was composed of two sub-systems
namely social sub-system and technical sub-system; and that a proper fit of both the sub-systems
was needed for achieving economic performance of the organization. When Socio-Technical
system theory was first developed “it was seen by its creators as a means for optimizing the
intelligence and skills of human beings and associating these with new technologies that would
revolutionize the way we live and work” (Mumford 2006).
Sociotechnical refers to the interrelatedness of social sub-systems and technical sub-systems
of an organization. The main goal of the socio-technical system theory is the “joint optimization”
(Emery 1959), which states that for an organization to function optimally, both the social and the
technical sub-systems of the organization should be designed to meet the demands of each other
and the environment. The social sub-system of an organization is comprised of people and the
relationships among them whereas the technical sub-system of an organization consists of tools,
techniques, procedure, skills, knowledge and other devices used by the people of the social subsystem to achieve their goals (Trist & K. Bamforth 1951; Emery 1959; Pasmore 1978; Pasmore
et al 1982; Cherns 1976).
One of the main concepts of Socio-Technical system is the notion of an “open system”
perspective. This concept recognizes that for an organization to run and survive effectively over
time it must thrive to interact with its surrounding environments and adapt to the changing
environmental conditions. “The term system means that all parts of the organization are
interrelated, so that the design of one necessarily effects the operation of another” (Pasmore et al
1982).
Sociotechnical theory is founded on two main principles: (1) the interaction of social and
technical factors creates the conditions for successful (or unsuccessful) organizational
63

performance; (2) optimization of each aspect alone, socio or technical tends to increase not only
the quantity of unpredictable, ‘un-designed’ relationships, but those relationships that are
injurious to the system’s performance.
In a Socio-Technical system, two types of interactions (linear and complex) between the two
sub-systems is eminent, whenever these two sub-systems are put together to work in a system.
The interactions could be of linear ‘cause and effect’ relationships that are normally ‘designed’
and partly from ‘non-linear’, complex, even unpredictable relationships that are often
unexpected. The inevitable consequence of putting these two together results in growing
complexity and interdependence; as the social subsystem (humans) do not behave in the same
manner as the technical sub-systems (machines).
The features that scholars hypothesized for the two sub-systems are (Watson, 2004):
“… the technical sub-system is much more than the sum of the equipment in the
organization; it can be identified with the process responsible for the conversion of
system inputs into system outputs” …. the social sub-system, on another hand, is
much more than the set of technical control tasks to be performed by people.
Technical tasks are combined with individual jobs and with responsibilities assigned
to groups. Any analysis and redesign of the social sub-system implies a revision of
the jobs and of the corresponding social roles, for the implications they have on the
technical sub-system …”
Overall performance for an organization can only be achieved if the interaction and
the interdependency among the two sub-systems are explicitly recognized. Any design or
redesign must be done keeping in mind the critical impact each subsystem might have on
the other and planning must ensure that both the sub-systems are in proper fit.
64

Crew design for construction operations has aspects of both the social subsystem (humans or
workers) and the technical sub-systems (machines or equipment). In construction operations,
there are number of both linear and complex interactions & interdependencies between the
workers and the machines/equipment. And optimization of only one sub-system cannot lead to
overall system performance for construction operations. Hence, a crew design derived from the
principles of Socio-Technical system theory ensuring the joint optimization of both the social
subsystem (humans or workers) and the technical sub-systems (machines or equipment) is of
value.
Socio-Technical principles have been developed to enhance the levels of organizational
performance by including operational measures such as effectiveness and productivity, along
with psychological indicators of human well-being and attitudes (Clegg 2000). Albert Cherns
(1976), an associate of Tavistock Institute presented nine principles of Scio-Technical system
design. The following principles represent a distillation of real-world experience and a synthesis
of writings by socio-technical icons such as Emery and Trist (1972) and Herbst (1974):
Principle 1: Compatibility
The process of design must be compatible with its objectives. “This means that if the aim is to
create democratic work structures then democratic processes must be used to create these”
(Mumford 2006). Cherns (1976) explained that if the design objective is to create an adaptive
system which is capable of self-modification then, a constructively participating organization is
required. For this to happen, people must be given an opportunity to participate in the design
process of the work they are required to perform.

65

Principle 2: Minimal Critical Specification
This principle has two parts (minimum & critical) with one being a negative aspect and the other
a positive one. The negative aspect states that no more should be specified than is absolutely
essential. And the positive aspect requires identifying what is essential. Cherns (1976) discusses
that specifying a lot arises out of the desire to reduce uncertainty in the job but it may results in
closing the options which might arise in the future due to the changing environment. This often
involves giving workers or group clear objectives but letting them decide how to achieve them.
Principle 3: The Socio-technical Criterion
This principle states that any variances or deviations should be eliminated. And if they cannot be
eliminated, they must be controlled as close to their point of origin (source) as possible. Cherns
(1976) defined variance as “any un-programmed event which critically affects the outcome.”
Variance could be present in the form of defects or deviations in the quality of raw material,
machine breakdowns, failure to take action at a critical time, etc. Also any variances occurring
should be eliminated or controlled by the group experiencing them and not by another supporting
group.
Principle 4: The Multi-functionality Principle
Work in organizations needs to constantly adapt to their changing environment. The work groups
have to be flexible in order to adapt to the changing environment. This can be achieved by
adding new roles, modifying old ones or adding a variety of skills.
Principle 5: Boundary Location
Boundaries should facilitate the sharing of knowledge, experience, information and learning.
They should occur where there is a need to group people and activities on the basis of – time,
technology change, etc. – in the work process. Need for Boundaries occurs when work from one
66

worker/group of workers passes on to another one and new set of skills or activities are required
to perform that work. For a favorable socio-technical system design, working groups should
learn to manage their own boundaries without hampering the coordination or communication
needed among the groups.
Principle 6: Information Flow
This principle states that information must be provided to all those who require it and whenever
they require it. Cherns (1987) suggested that information in the organization has three uses: for
control, for record keeping and for action. Information provided at the right moment may also
help in controlling the variances and improve the performance of the work groups.
Principle 7: Support Congruence
This principle states that systems of social support must be designed to reinforce the desired
social behavior and that management’s actions should be consistent with the expressed
philosophy (Cherns 1976). He suggested that systems such as conflict resolution, work
measurement, performance assessment, training, promotion etc. can all contradict or reinforce
the behaviors which are desired.
Principle 8: Design and Human Values
This principle states that an objective of design should be to provide high quality work. Cherns
(1976) suggested that quality is a subjective phenomenon involving responsibility, variety,
growth etc. and the broad objective should be to provide these to only those who want them
without subjecting others to peer control. Thorsrud (1972) identified six characteristics of a
good job or high quality of work: (1) jobs to be reasonably demanding; (2) opportunity to learn;
(3) an area of decision-making; (4) social support; (5) the opportunity to relate work to social
life; and (6) a job that leads to a desirable future.
67

Principle 9: Incompletion.
This principle identifies that the design is a re-iterative process and that it never stops. Every
design once implemented, because of the changing and uncertain environment will have
consequences indicating the need for redesign. Multifunctional, multi-level, multidisciplinary
team required for design is needed for its evaluation and review” (Cherns 1976).
The proceeding review of the existing Socio-Technical literature provides conceptual
foundations for developing construction –specific crew design (formation and composition)
criteria
2.6

Crew and Work Design

While designing “a team for work” or “work for a team” appear to be two different problems,
both involve the integrated design of “group of workers/people” and the “work/job” that needs to
be executed by them. The “group of workers/people” can be defined as a “team”, “group” or
“crew” and a work/job can be defined as a set of “activities” or “tasks” that are performed by the
workers/people.
Also, a job can be defined as “an aggregation of tasks assigned to a worker” (U.S.
Department of Labor 1972) whereas a task is a “set of actions performed by a worker who
transforms inputs into outputs through the use of tools, equipment, or work aids,” Medsker and
Campion (2007). And Job Design is the “content and structure of jobs that employees perform”
(Oldham, 1996, p. 33). In the literature, Job Design has been known by different names such as
“Work Design” or “Team Design”, but all are mean the same when generalized.

68

When work is executed by a “group of workers/people” known as a “team”, “group” or “crew”,
then the performance of the team/crew affects the performance of the overall work or the
operation. In addition there are numerous decisions to be made regarding:
•

How the work is to be executed;

•

How many workers need to be in a crew/team;

•

What should be the composition of workers in a crew, sequencing of tasks; and

•

How the work should flow among the workers in a crew.

For example in construction, one of the main problems facing estimating and scheduling
teams working on early stages of a construction project is the need to determine a reasonable
production rate for a crew based on their composition, experience, time working together, etc
(Hassanein 1997). In the present scenario, this complex job/work/team design (involving both
crew and work performed) is typically performed in an ad-hoc, random manner by the managers
or supervisors, which, at best, is equally likely to produce low or high performance outcomes.
Crew design is not a luxury but a necessity and an integral part involved in the process of
executing any job/work/operation. However, the design of either work/job/teams/crews hasn’t
received much attention in construction literature. As stated by Medsker and Campion (2007),
“Job design is one of those aspects of managing organizations that is so commonplace that it
often goes unnoticed”.
Different scientific disciplines such as industrial engineering, organizational psychology,
social psychology and socio-technical system theory, etc, have all developed approaches for
“work design” (Gladstein 1984; Hackman 1980, 1987; Hackman and Oldham 1980; Guzzo and
Shea 1992; Niepcel and Molleman 1998; Morgeson and Campion 2003; Morgeson et al 2006;

69

Campion et al 1993, 2005) but the studies conducted have been relatively independent of each
other. These studies provide a framework for “work/job/team/crew design” focusing on
work/team characteristics or indicators.
Campion et al (1993) suggested that psychological approaches to work design have been
consistent in achieving increased satisfaction for the teams/groups; however, this approach to
work design has always been in conflict with the traditional engineering approaches to work
design (Campion, 1988; Campion & McClelland, 1991; Campion & Thayer, 1985). So what is
the best way of designing work or a team? In the previous sections of this chapter, engineering
approaches such as of Lean Production and Lean Construction have been discussed.
In the following discussion, we investigate the concept of “crews” as “production work teams”
and provide a brief discussion of the distinctions among different types of “work teams” that
have emerged in the theoretical literature.
A team can be defined as (a) two or more individuals who (b) socially interact (face-to-face
or, increasingly, virtually); (c) possess one or more common goals; (d) are brought together to
perform organizationally relevant tasks; (e) exhibit interdependencies with respect to workflow,
goals, and outcomes; (f) have different roles and responsibilities; and (g) are together embedded
in an encompassing organizational system, with boundaries and linkages to the broader system
context and task environment (Alderfer, 1977; Argote & McGrath, 1993; Hackman, 1992).
Many classification schemes are present in the literature on how to define groups. Focusing
on what groups do, Carter et al. (1950) identified six primary group task types: (a) clerical, (b)
discussion, (c) intellectual construction, (d) reasoning, (e) mechanical assembly, and (f) motor
coordination. In the sixties, McGrath and Altman (1966) extended these ideas by identifying four
ways that small groups could be classified according to the (a) behaviors enacted by group
70

members, (b) behaviors required of group members for effective performance, (c) abilities
required of group members, and (d) characteristics of the group’s task per se.
Later, McGrath (1984) consolidated several features of existing classifications to produce an
eightfold typology of group tasks based on three dichotomized dimensions (i.e., conflict–
cooperation, conceptual–behavioral, and choose execute): (a) generating plans, (b) generating
ideas, (c) solving problems with correct answers, (d) deciding issues with no correct answers, (e)
resolving conflicts of viewpoints, (f) resolving conflicts of interest, (g) resolving conflicts of
power, and (h) executing performance tasks.
At about the same time, Hackman (1990) offered another influential classification of seven
work group types in the preface to an edited volume featuring case studies of team effectiveness;
he divided the teams into (a) top management groups, (b) task forces, (c) professional support,
(d) performing groups, (e) human service teams, (f) customer service teams, and finally (g)
production teams.
Present literature offers a conceptual distinction between “teams” and other types of
“groups”. McGrath (1984) concluded that in terms of small groups there are three attributes of
interest: (1) two or more individuals, (2) regular interaction among group members, and (3)
interdependence among them in some way. This was similar to the definition provided by Dyer
(1984) who suggested that teams “consisted of two or more people who have a common goal,
specific role assignment, and interdependence”.
This was further supported by Morgan et al (1986) who defined teams as "distinguishable
sets of more than two individuals who interact interdependently and adaptively to achieve
specified, shared, and valued objectives". Further McGrath and colleagues suggest that “crews

71

are formed when an organization has developed the technology to carry out a class of projects
(tasks) and then select personnel to staff the project”.
Guzzo and Dickson (1996) suggested that the use of term “group” has been largely replaced
by term “team” in the discipline of organizational psychology and that many labels are used for
“team” such as empowered teams, autonomous work groups, semi-autonomous work groups,
self-managing teams, self-determining teams, self-designing teams, crews, cross-functional
teams, quality circles, project teams, task forces, emergency response teams etc.
Further, a number of different types of work teams have been classified based on the team
autonomy. The classification provided by Banker et al (1996) has been shown in Figure 2.15.

Figure 2.15: Graphical representation of Work Team Autonomy (Banker et al 1996)
Traditional work groups perform core production activities and have no management
responsibility or control. Support activities such as quality control, inspection, maintenance, etc.
are the responsibility of other groups. In Quality circles, the work group has the responsibility for
making suggestions but do not have any decision making authority.
Along with execution, semi-autonomous work groups have the responsibility to also manage
the production activities. The difference between the autonomous and semi-autonomous work
groups is that “the scope of the production tasks managed and executed by autonomous groups is
narrower than that of the semi-autonomous groups”; these autonomous groups can self-regulate
72

their work and have complete control over the management (planning, execution and control) of
their tasks. Further, “self-designing teams” are similar to the self-managing or autonomous
groups along with a control over design of the team and the tasks involved.
Sundstrom et al. (1990) provided the following classification of work teams:
advice/involvement, production/service, project/development, and action/negotiation. Sundstrom
(1999) consolidated this classification where he distinguished six types of organizational teams:
(a) production teams, (b) service teams, (c) management teams, (d) project teams, (e)
action/performance teams and (f) parallel teams. Further, in a review of the research on work
groups published in 1990’s, Cohen and Bailey (1997) distinguished four team types: (a) work
teams, (b) parallel teams, (c) project teams, and (d) management teams.
“A work team is a structured set of employees who require coordinated interactions to pursue
collective performance objectives within the larger organizational system and successfully
accomplish relevant tasks; such employees possess complementary skills” Ogungbamila et al
(2010). Sundstrom et al. (1990) suggested that work teams have been extensively used to
generate products or services, such as in manufacturing, maintenance, construction, mining, and
commercial airlines, and they “usually consist of first-line employees working together full-time,
sometimes over protracted periods, with freedom to decide their division of labor”.
Work teams have been common both in manufacturing and service industry. Optimizing the
performance of crews (production work teams) is a very essential element as “work teams are
considered to be an integral tool aiding continuous improvement in work operations” (CutcherGershenfeld and Associates 1994).
For this research, the classification by Cohen and Bailey (1997) and Sundstrom (1990, 1999)
where work teams are defined as “continuing work units responsible for producing goods and
73

providing service” fits very well with our description of a “crew” for a production setting in
construction.
Work teams that exhibit the following characteristics could be “classified as crews: (1)
membership of expert specialists, (2) perform events that are closely synchronized with
counterparts and support units inside the organization (e.g., fire fighter crews), and (3) have brief
performance events repeated under new conditions” (Sundstrom et al. 1990). Also, a crew can be
defined as a “group of expert specialists each of whom have specific role positions, perform brief
events that are closely synchronized with each other, and repeat these events across different
environmental conditions” (Klimoski & Jones, 1995; Sundstrom et al., 1990).
Further, this study examines the several design indicators or characteristics of “work
design/team design” and how they relate to different performance outcomes based on approaches
other than engineering which has already been discussed in the previous section of this literature
review. This is accomplished by primarily reviewing research studies on “work design/team
design” that have well developed theoretical and empirical foundations in socio-technical system
theory, social and organizational psychology. This is conducted, to ensure that the proposed
crew design framework and guidelines in this study are strongly supported.
Work design or team design from a number of significant studies (Gladstein 1984; Hackman
1980, 1987; Guzzo and Shea 1992; Morgeson and Campion 2003; Morgeson et al 2005, 2006;
Morgeson and Humphrey 2008; Campion et al 1993, 2005; Medsker and Campion 2007;
Humphrey et al 2007; Kozlowski and Ilgen 2006; Sundstrom et al 1990) is reviewed to
determine the relevant design indicators or characteristics and the impact of these indicators on
various performance outcomes.

74

In the reviewed literature, as stated earlier, team/crew/group effectiveness has been defined
in a number of ways. Guzzo and Dickson (1998) defined performance effectiveness consistent
with Hackman (1987) and Sundstrom (1990) that “effectiveness in groups is indicated by (a)
group produced outputs (quantity or quality, speed, customer satisfaction, and so on), (b) the
consequences a group has for its members, or (c) the enhancement of a team’s capability to
perform effectively in the future”. Shea and Guzzo (1987) defined group effectiveness as
"production of designated products or services per specification".
Hackman and Morris (1975) suggested that effectiveness has three main components: group
performance, satisfaction of group-member needs, and the ability of the group to exist over time.
Campion et al (1993) defined work group effectiveness in terms of productivity, satisfaction, and
manager judgments. Medsker and Campion (2007) extended Campion et al (1993) measures of
performance outcomes into “productivity: productivity, efficiency, quality, reduced costs” and
“satisfaction: job satisfaction, satisfaction with team, commitment, motivation, customer
satisfaction”.
Campion et al (1993) suggested that almost all the job characteristics of Hackman and
colleagues (e.g., Hackman & Lawler, 1971; Hackman & Oldham, 1980) can be applied to groups
for work design. However, because of the scope of this research, significant work design
indicators or characteristics relevant to the production setting in construction were identified and
investigated. Further, several meta-analytic findings from the theoretical literature were
identified to investigate the work design indicators or characteristics.
Relative size is the number of individuals in a team/crew/group and has shown to be related
to effectiveness or performance (Steiner, 1972; Gladstein 1984; Nieva et al 1985; Campion et al.
1993; Guzzo and Shea 1992; Cohen and Bailey 1997; Hackman 2002; Morgeson et al 2005;
75

Leach et al 2005; Ogungbamila et al 2010). Campion et al. (1993) identified that group size was
positively related to both productivity and satisfaction; and suggested that the groups need to be
of feasible size in order to accomplish the work assigned to them. According to Huchman et al
(2009) increasing the size of a work team may lead to certain benefits such as more resources
available and increased capacity resulting in high overall productivity for a work team.
However, “an increase in resources (because of more team members) is often
counterbalanced by the team’s increased difficulty in arriving at a decision and a decrease in the
average personal and work space of each team member” (Ogungbamila et al 2010). In addition,
increase in size beyond a feasible limit might also lead to increased coordination problems and
unequal participation of group members (Hackman 20002). Too many or too few individuals in a
work team have been shown to reduce performance and that size of a work team has an inverted
U-shaped relation to effectiveness (Nieva et al 1985). However, depending on the nature of work
and its environment, the U-shaped relationship might not be true for all types of teams (Cohen
and Bailey 1997; Kozlowski and Bell 2003).
Task significance, task variety and task identity are significant characteristics affecting
performance or effectiveness (Cummings, 1978; Campion et al 1993; Morgeson and Humphrey
2008; Medsker and Campion 2007). Task significance means the extent to which the group's
work has significant consequences inside and outside the organization and the impact it has on
the lives of others. Task variety means the extent to which workers (same skill level) are
involved in a variety of tasks (Hackman 1987; Morgeson & Humphrey 2006; Campion et al
1993); it relates to the concept of job-enlargement from socio-technical theory allowing both
interesting and dull tasks to be shared among members (Davis and Wacker, 1987; Walton, 1972).
Task identity is the degree to which the group/individual completes an entire or separate piece of
76

work from start to finish (Campion et al 1993) and it may increase worker motivation by
instilling a feeling of pride (Hackman and Oldham 1976; Hackman 1987).
Meta-analytic findings such as (Humphrey et al., 2007) have shown that task significance
impacts job satisfaction (ρ = .41) and work motivation (ρ = .45); also task variety is positively
related to job satisfaction (ρ = .46) whereas, task identity is modestly related to worker
motivation (ρ = .26) and job satisfaction (ρ = .31).
Task complexity is the extent to which a given task is difficult to perform or the “extent to
which a task/job is multifaceted” (Humphrey et al 2007). Task/job complexity can be considered
as a mechanistic aspect of work (Morgeson and Humphrey 2008) and it can be well argued that
reducing the complexity or simplifying work should generally lead to efficiency gains. However,
Morgeson and Humphrey (2008) presented findings indicating that higher complexity is related
to higher subjective performance ratings (ρ = .37) (Humphrey et al 2007).
Task experience is the amount of time spent by a team/individual in performing a task or the
number of times a similar type of task has been performed. In addition, it can also refer to the
prior amount of knowledge an individual has in performing that task. Morgeson and Humphrey
(2008) suggested that “having higher task experience will help workers perform successfully in
jobs that have high task variety (i.e., breadth of knowledge) or specialization (i.e., depth of
knowledge)”. A higher task experience may also have causal effects in terms of higher costs,
improved performance, and technical support for the group/team members.
Interdependence primarily exists in the form of task interdependence, goal interdependence,
or outcome interdependence. Campion et al (1993) suggested that “Interdependence may
increase the motivational properties of work or the efficiencies with which the work is done, and
thus may be related to effectiveness”.
77

Task interdependence exists when a group of individuals get together and depend on one
another to perform a certain work consisting of several tasks; and either start or finish of one task
is dependent on the other tasks. Task interdependence exists in the form of pooled, sequential
and reciprocal interdependence (Thompson 1967). The manner it affects work performance has
been discussed in the previous sections of this literature review. Goal or outcome
interdependence exists in the form of clearly defined objective/mission or has also been defined
as the “degree to which the significant outcomes an individual receives depend on the
performance of others” (Campion et al 1993; Wageman 1995).

It is considered to be a

significant characteristic for group effectiveness (Davis and Wacker 1987; Gladstein 1984;
Guzzo and Shea 1992; Hackman, 1987). Campion et al (1993) suggested that “individual
members' goals must be linked to the groups' goals to be maximally effective”.
Autonomy means the “amount of freedom and independence an individual has in terms of
carrying out his or her work assignment (Morgeson and Campion 2003). Also, autonomy can be
of different types such as: work scheduling autonomy, work methods autonomy, decision-making
autonomy (Morgeson and Humphrey 2006, 2008; Humphrey et al 2007). Autonomy stems from
the research in organizational psychology (Hackman & Oldham 1980) and socio-technical theory
(Trist 1981) where it is recommended that increased autonomy in execution of work leads to
increased work satisfaction and motivation among the workers.
This view is also supported by Leach et al (2005), recommending that increasing autonomy
could be a worthwhile strategy from both team and individual perspective for enhancing team
performance (p < .05). Further, results from the meta-analytic findings (Humphrey et al., 2007)
shows that autonomy can be related to both objective (ρ = .17) and subjective (ρ = .23)
performance ratings. Also, autonomy was shown to reduce a number of well-being outcomes
78

such as anxiety (ρ = -.10), stress (ρ = -.23), and burnout (ρ = -.30). In addition, autonomy is
related to job satisfaction (ρ = .48), organizational commitment (ρ = .37), and internal work
motivation (ρ = .38). Among work scheduling autonomy, work methods autonomy and decision
making autonomy has a large impact (ρ = .58) on job satisfaction (Humphrey et al., 2007).
Self-management is the group level analogy to Autonomy and is presumed to enhance
performance or effectiveness by increasing the workers sense of responsibility and ownership of
work (Campion et al 1993; Manz & Sims 1980; Morgeson and Humphrey 2008; Medsker and
Campion 2007).
Feedback is considered to be a significant characteristic and is recognized in several research
studies (Guzzo and Shea 1992; Gladstein, 1984; Hackman, 1987; Steiner, 1972; Sundstrom et al
1990). In the theoretical literature, although feedback has generally been considered to be
originating from a single source; Morgeson and Humphrey (2008) made a distinction between
the type of feedback present into feedback from others and feedback from the job. Feedback from
others is the “extent to which members of the organization provide information about job
performance whereas feedback from the job is the extent to which a job imparts information
about an individuals’ performance” (Humphrey et al 2007; Morgeson & Humphrey, 2006).
Meta-analytic findings from Humphrey et al (2007) show that feedback from the job relates
positively with work outcomes such as work motivation (ρ = .42) and job satisfaction (ρ = .43).
Further, results also demonstrate that a timely and effective feedback leads to reduction in role
ambiguity (ρ = -.43) and role conflict (ρ = -.32).
Social support is another characteristic recognized in research studies (Gladstein 1984;
Morgeson and Humphrey 2006, 2008; Humphrey et al 2007; Campion et al 1993; Medsker and
Campion 2007). It can be defined as the “extent to which there are opportunities for gathering
79

assistance and advice from supervisors and coworkers,” (Morgeson & Humphrey, 2006,
2008).Campion et al (1993) suggested that “effectiveness may be enhanced when members help
each other and have positive social interactions.”
From the meta-analytic findings (Humphrey et al., 2007) it has been found that social support
is strongly related to job satisfaction (ρ = .56). Shea and Guzzo (1987) argued that given that
resources such as material and information are controlled and allocated by the management, then
managerial support is the key to group functioning. Campion et al (1993) built on this and
suggested managerial support is logically related to group effectiveness.
2.7

Chapter Summary

This chapter provided an overview of the existing construction models and crew and work
analysis methods followed by the review of production theories such as Lean Production, Lean
Construction and Socio-Technical System theory. Production theories reviewed such as SocioTechnical, Lean Production, Lean Construction and Social and Organizational formed the basis
for developing the guidelines for lean crew design.

80

3 MODELING AND SIMULATION IN CONSTRUCTION

3.1

Introduction to Modeling and Simulation

The purpose of this chapter is to review the concepts and techniques regarding the modeling and
simulation to gain better understanding of the use of simulation in a construction system. A
detailed view on simulation approaches such as discrete event simulation and system dynamics
modeling is presented. Further, advantages and limitations of each of the modeling techniques
are reviewed to gain better understanding of the process of building simulation models.
According to Bellinger (2004) Modeling and Simulation “is a discipline for developing a
level of understanding of the interaction of the parts of a system, and of the system as a whole”.
Although in general, Modeling and Simulation have been used interchangeably, they have very
specific meanings in the modeling and simulation scientific discipline.
According to the Department of Defense (1998) Modeling is defined as “application of a
standard, rigorous, structured methodology to create and validate a physical, mathematical, or
otherwise logical representation of a system, entity, phenomenon, or process”; whereas,
Simulation is the “method for implementing a model over time” (DoD 1998) or it can also be
defined as the operation of a model, which is a representation of that particular system itself.
In addition, according to Borshchev and Filippov (2004), Modeling is a “way of solving
problems that occur in the real world”. In general, Modeling is carried on in order to reduce the
costs associated with analyzing, experimenting or implementing any changes to the real system;
this way performance of real systems could be optimized before implementing any change to it.

81

However, Simulation is the “process of model execution that takes the model through
(discrete or continuous) state changes over time” (Borshchev and Filippov 2004). Also,
Simulation is the “process of designing a model of a real system and conducting experiments
with this model for the purpose of either understanding the behavior of the system or evaluating
various strategies (within the limits imposed by a criterion or set of criteria) for the operation of
the system” (Shanon 1975).With advances in computer technologies, simulation models have
been used and developed for decades to help analyze complex systems. Simulation is one of the
most widely used operations research and management-science techniques (Law and Kelton
2000).
There are several key concepts underlying modeling and simulation. Discussion of these
concepts is essential in understanding and performing simulation. Some of the key concepts are
system, model, and event. These concepts have been explained in detail in Carson (1993), Banks
et al (2000), and Law and Kelton (2000).
A system can be defined as a group of objects coexisting and mutually interacting towards
the accomplishment of a common purpose. According to Shannon (1975) a model is a
representation of an object, a system, or an idea in some form other than that of the entity itself.
A model is developed in order to have a simplified representation and gain better understanding
of a system, entity, or process. A model should always be similar and a close approximation to
the system but at the same time it should also be able to represent a complex system in much
simpler form in order to have easy experimentation and also to have a better understanding of the
system. An event can be defined as an occurrence that changes the state of the system it takes
place in. And these events can be internal and external events also known as endogenous and
exogenous depending on whether the event occurs within the system or outside the system.
82

3.2

Classification of Simulation Models

In general, Modeling and Simulation is the use of models either Physical (emulators, prototypes,
scale models, etc.) or Mathematical constructs to analyze and evaluate the system’s performance
and help in decision making. Mathematical models are also known as logical models which can
be either:
1. Analytical models (linear programming, queuing theory, differential equations, etc.) or
2. Simulation models (discrete event simulation models, system dynamics models, agent
based models, etc.).
In addition, due to the complex interaction among units in a system’s environment,
Analytical models can be applied to only a limited number of cases; whereas Simulation models
are more flexible and can be easily developed, analyzed, and experimented with for a majority of
actual processes, phenomenon or systems to optimize performance and help in decision making.
A number of system/process problems ranging from strategic (macro) level to operational
(micro) level have been successfully addressed using modeling and simulation. For example, it
has been used for designing, analyzing and optimizing manufacturing systems, military weapons
systems or tactics, policies for business systems, construction systems, transportation facilities,
financial or economic systems, supply chain systems, health systems, etc.
Further, Simulation models can be classified as:
1. Deterministic and Stochastic,
2. Static and Dynamic,
3. Discrete and Continuous.

83

In all these models, there is a set of input variables driving the simulation and producing a set
of outputs to be observed.
3.2.1

Deterministic and Stochastic Simulation Models

Deterministic models can be defined as those models that do not have any random or
probabilistic element present; these models do not account for any uncertainty that may occur in
a real system. Whereas in Stochastic models, randomness and probabilistic element are present
and they account for the uncertainty present in a real system by the use of probability
distributions.
According to Banks and Carson (1984) Deterministic models of a known (deterministic) set
of input variables which result in a unique set of outputs whereas Stochastic models have one or
more random variables as inputs leading to random outputs. Also as the outputs are random, they
represent only estimates of the real system and in general require a number of runs of the model
to get an effective estimate of the real system.
3.2.2

Static and Dynamic Simulation Models

In addition, both the Deterministic and Stochastic models can be static or dynamic in nature.
Static models do not account for the element of time and they represent a system at a particular
point of time; whereas, dynamic models account for the time element and represent a system as it
changes over time.
3.2.3

Discrete and Continuous Simulation Models

Simulation models can be Discrete or Continuous depending on if the variables present in the
simulation model changes continuously over time or at discrete points of time; for example, "If

84

alterations to the model occur continuously as time varies, the model is continuous; if such
changes occur only at discrete instants of time, the model is discrete" (Hoover and Perry 1989).
Figure 3.1 compares how continuous/discrete simulation models are time or state models
depending on if the state is defined continuously at all times or only at some discrete time and if
the state variables are either continuous or discrete (Jain 1991).
(a) Discrete State

(b) Continuous State

Time
Spent
by
Students

Number
of Jobs
in the
Queue

TIME

TIME

(a) Continuous Time

(b) Discrete Time

Number
of Jobs
in the
Queue

Number
of
Students
in a
Course
TIME

TIME

Figure 3.1: Graphical representation of discrete and continuous models (Jain 1991)
In addition, according to Law and Kelton (2000), simulation models can be discrete in two
ways, namely, time-stepped or event-stepped. In time-stepped models the state of the system

85

changes at each time step whereas in event-stepped models the state of the system changes at the
occurrence of some event in the system.
Both Discrete and Continuous simulation approaches have been used in construction industry
to predict and improve the performance of the construction projects. They have been successfully
used by managers, engineers and researchers to design, understand and improve the construction
process. Simulation makes it possible to analyze different construction processes and predict the
best possible methods without actually implementing them in real practice.
For example, simulation has been used to determine construction alternatives (Ioannou and
Martinez 1996); map and improve the production process of manufactured housing factory
(Senghore et al 2004); to minimize the production time and cost of concrete batch plant
production process (Zayed and Halpin 2001); to explore the interaction between project and
business-level processes involved in bidding (Bayer and Gann 2006); to understand change and
rework in construction project management systems (Love et al 2002); to explore performance
enhancement in a construction organization (Ogunlana et al 2003); to investigate the effects of
resource allocation policies on project durations (Lee et al 2004); and to investigate crew design
alternatives (Naveen and Abdelhamid 2010).
The reason simulation has been so useful is because it provides for a faster, less expensive
and simple method of analyzing a real process consisting of multiple interdependent components
involving multiple feedback and nonlinear relationships present at the strategic (macro) and
production (micro) level. Discrete (Discrete event simulation) and continuous (system dynamic
modeling) simulations contribute in evaluating and improving both the macro and micro level
construction processes.

86

3.3

Discrete Event Simulation (DES)

3.3.1

DES Introduction

Discrete event simulation (DES) is a simulation technique based on statistical analysis and
queuing theory where state changes occur at discrete points of time whereas system dynamics
(SD) is a modeling technique consisting of a system of stocks and flows where continuous state
changes occur over time (Brailsford and Hilton, 2001). The main objective in both simulation
approaches is to model the real system and understand how systems behave over time and to
compare their performance under different conditions (Sweetser, 1999).
Discrete event simulation (DES) has been widely recognized as an effective technique to
design and analyze construction operations for over three decades by virtue of its ability to
handle complexity and uncertainty (Halpin 1973; Martinez and Ioannou 1999, Halpin et al
2003). Discrete event simulation is a useful technique to imitate the real world systems over time
by modeling repetitive processes where variables representing the state of the system change at
discrete points of time along with many resource interactions (Law and Kelton 2000). In
discrete-event simulation, the operation of a system is represented as a chronological sequence of
events where each event occurs at an instant in time and marks a change of state in the system
(Robinson, 2004).
3.3.2

DES Modeling Concepts

Page (1994) suggested that in discrete event simulation two concepts of time and state are of
major importance and defined the relationship between them in the terms of event, activity and
process shown in Figure 3.2.

87

1. Instant: value of system time at which the value of at least one attribute of an object can
be altered.
2. Interval: duration between two successive instants.
3. Span: contiguous succession of one or more intervals.
4. State: enumeration of all attribute values of that object at a particular instant.
5. Activity: state of an object over an interval.
6. Event: change in an object state, occurring at an instant, initiating an activity.
7. Process: succession of states of an object over a span.

Figure 3.2: Graphical representation of event, activity and process in DES (Page 1994)
In addition to the terms defined by Page (1994), other important terms in DES are entities known
as the objects in a system, attributes representing the properties of the entities in a system and
simulation clock representing the simulation time or the variable that keeps the current value of
(simulated) time in the model (Law, 2007). Simulation time may not be equal to the real time.

88

3.3.3

DES World Views

In DES the conceptual framework (world view) or the simulation strategy guides model
development (Page 1994; Pritsker et al 1997; Pidd 1998; Martinez and Ioannou 1999; Martinez
1996) in the following three ways:
1. Event scheduling view (ES) is governed by the occurrence of the events in a system. In
this view the occurrence of events that change the state of the system are determined by
the model developer. Time is advanced to the next scheduled event that is the end of an
activity and next scheduled activities are examined to determine if they can start or not
(Abu-Taieh and El Sheikh 2008). “The event based simulation model is driven by the
scheduling and execution of the events which in turn schedule and execute other events,”
(Martinez 1996). The event scheduling paradigm/view is shown in Figure 3.3.

Figure 3.3: Graphical representation of event scheduling paradigm/view Pidd (1998)

89

2. Activity scanning view (AS) is governed by the occurrence of activities (generally
represented as pairs of start and end events) in a system. In this view a model developer
determines which activities should be present and under what conditions they can take
place. It is also known as a two phase approach (Figure 3.4); where in the first phase,
time is scanned and advanced if the activities meet the conditions specified, and then
executed; and in the second phase, the system is updated if an event occurs (Abu-Taieh
and El Sheikh 2008).

Figure 3.4: Graphical representation of activity scanning paradigm/view Pidd (1998)
3. Process interaction view (PI) is governed by the flow of entities through a system. “These
entities undergo a process in which they attempt to acquire, take hold of, and release
scarce resources” (Martinez 1996). Also, the flow or movement describes all of the states
that an entity can attain in the system (Abu-Taieh and El Sheikh 2008).
90

In addition to the three world views, there is another approach known as the Three-Phase
(AS) that combines both the activity scanning (AS) and event scheduling (ES) views (Martinez
1996; Balci 1988). This three-phase (AS) approach was first developed by Tocher (1963) in his
book, The Art of Simulation.
In this approach there are two different types of system activities; one which are bound to
happen known as bound activities (B activities) and other as conditional activities (C activities)
that would take place only when the activity conditions are satisfied.

Figure 3.5: Graphical representation of Three-Phase (AS) paradigm/view Pidd (1998)

91

3.3.4

DES Software Tools

In general, discrete event simulations (DES) can be executed through the use of either “specific
purpose simulation (SPS) tools” or “general purpose simulation (GPS) tools”.
Specific purpose simulation (SPS) tools target a very narrow range of systems. They are
designed for specific tasks or operations and can also be classified as simulators (Law and
Kelton 1991). AbouRizk and Hajjar (1998) define SPS as “a computer-based environment built
to enable a practitioner who is knowledgeable in a given domain, but not necessarily in
simulation, to model a project within that domain in a manner where symbolic representations,
navigation schemes within the framework, creation of model specifications, and reporting are
completed in a format native to the domain itself.”
An example of specific purpose simulation tools is SIMPHONY, which is a Microsoft
Windows based system for the development as well as the utilization of construction SPS tools
(Hajjar and AbouRizk 1999). Other SPS tools developed for construction include AP2-Earth,
CRUISER and CSD, all targeting specific construction operations such as large earth moving
projects, aggregate production plants, and construction site dewatering operations (Hajjar and
AbouRizk 1996, 1998; Hajjar and AbouRizk and Xu 1998).
On the other hand, general purpose simulation (GPS) tools can model wide variety of real
systems. GPS tools are more flexible and can be designed to represent a variety of construction
operations. These tools make use of general purpose programming languages (C, C++,
FORTRAN, etc) to describe the model by providing a set of instructions to represent the real
system (Fathi et al 2003). For example, CYCLONE, RESQUE, COOPS, CIPROS, STEPS,
STROBOSCOPE etc are few of the general purpose simulation tools present for modeling and
representing general systems.
92

Most of the general purpose simulation tools are based on activity cycle diagrams (ACDs)
making them more suitable for modeling construction operations. One of the earliest and most
used GPS tool based on ACDs of construction operations is CYCLONE (CYClic Operation
Network). CYCLONE was developed by Dr. Halpin at the University of Maryland in the early
1970’s and is one of the best known DES tool used for modeling construction operations. Since
the development of CYCLONE (Halpin and Woodhead 1976), there have been several proposed
improvements to extend its merits, including INSIGHT (Kalk 1980), UM-CYCLONE (Ioannou
1989), Micro-CYCLONE (Halpin 1990) and STROBOSCOPE (Martinez 1996).
STROBOSCOPE (STate and ResOurce Based Simulation of COnstruction ProcEsses) is one of
the more recent general purpose simulation tool developed for modeling construction operations
(however it is also possible to model other type of systems apart from construction systems).
STROBOSCOPE is programmable and can be extended for modeling complex operations
(Martinez 1996). It is based on the concept of Activity Cycle Diagrams (ACDs) and supports the
Three Phase Activity scanning approach.
STROBOSCOPE (and a Visio-based version called EZStrobe) has been extensively used in
modeling the construction process at the operation level. A construction project or operation can
be defined in terms of processes (Halpin and Riggs 1992). And the model developed in
STROBOSCOPE can represent these processes in the form of collection of activities involved in
that particular operation. In STROBOSCOPE, heterogeneous types of labor, equipment or
materials can be included in the model where resources of the same type can be distinguished
from one another and each can have individual properties (Martinez and Ioannou 1999).

93

The STROBOSCOPE/EZStrobe (based on the ACDs and the Three Phase Activity scanning
approach) consists of the following modeling elements (Martinez 2001) shown in Figure 3.6:
Modeling Element

Definition/Function

CombiName

Conditional Activity - A named element that represents tasks that can start
whenever the resources that are available in the Queues that precede it are
sufficient to support the task. In a Conditional Activity start priorities and
probabilistic duration can also be included.

NormalName

Bound Activity - A named element that represents tasks that starts
whenever an instance of any preceding Activity ends. A Bound Activity
can follow any node except a Queue, and can precede any node except a
Conditional Activity.

Queue
Name

Queue - A named element that holds idle resources. The number of
resources it holds is usually shown at the beginning of the simulation. A
Queue can follow any other node except another Queue and can only
precede a Conditional Activity (Combi).

Fusion

Fusion Queue - A named element that similar to Queue used to hold
resources. Its main function is to connect parts of the model that cannot be
modeled together due to space limitations by representing and connecting
with an earlier Queue used in the model.

>0, 1

Draw Link – It connects a Queue to a Conditional Activity. A Draw Link
shows two pieces of information separated by a comma. The first part is
the condition necessary for the successor and the second part is the
amount of resource that the Conditional Activity will attempt to remove
from the predecessor Queue in the event that the Conditional Activity
does start.

1

Release Link - It connects an Activity to any other node except a
Conditional Activity. It indicates the amount of resource that will be
released after each instance of the predecessor activity ends.

P:1

Branch Link – It connects a Fork to any other node except a Conditional
Activity.

Figure 3.6: Graphical representation of the modeling elements in EZStrobe (Martinez 2001)

94

3.3.5

DES Application

DES has been the most widely used operations research tool (Banks and Carson 1984).
Successful application of DES has been seen in numerous construction management areas such
as simulation of earth-moving operations (Smith et al 1995), investigation of lead time buffering
(Srisuwanrat and Ioannou 2007); resource management, concrete batch plant production process,
pile productivity assessment (Zayed and Halpin 2001, 2002, 2004); simulation of large precast
operations (Balbontin-Bravo 1998); pull-driven scheduling for pipe-spool installation
(Tommelein, 1998).
3.3.6

DES Advantages and Limitations

In general, DES has overcome the “inability of the project-level models commonly used in
industry practice (critical-path method or CPM scheduling tools, cost control systems, materials
management databases, etc.) to model the construction process” (Tommelein 1998). It has gained
wide popularity in modeling the construction process at the operation level because of its
capabilities of conducting detailed analysis of complex systems where state of the system
changes occur at specific points of time in a very linear fashion. Further, use of DES permits
analysis of construction operations providing information about expected cost, duration, and
other process level details such as resource utilizations, rate of production, idleness of resources,
waiting/delays etc. In addition, it is a suitable approach to perform bottleneck analysis at the
process level to determine the cause of the delays in work-in-process, information, materials, or
other processes (Banks 1999).
The limitation in DES is that it is more suited for modeling at the operation level where the
entities are passive objects, representing a low level of abstraction. When the systems get more
complex and integrated, discrete event simulation faces serious challenges to keep up with the
95

high level of abstraction present in the systems. In discrete event simulation, it is also difficult to
model "softer" aspects of projects such as learning, pride or morale, trust, reliability, and
incorporate managerial actions which are important in determining the project behavior
(Williams 2002).
3.4

Continuous Simulation - System Dynamics (SD)

As discussed above, DES has proved a suitable approach for addressing operational level issues
and is more focused towards cost, duration, and other process level state variables (resource
utilizations, rate of production, idleness of resources, waiting/delays, etc.) at low level of
abstraction that change at discrete points of time.
To model various nonlinear cause-and-effect relationships, feedback, and time delays
existing in the complex construction processes, continuous simulation models, where state
variables change continuously over time, may be more suitable. The cause-and-effect
relationships and interactions among the state variables can be analyzed more effectively using
continuous simulation.
3.4.1

System Dynamics (SD) Introduction

System Dynamics (SD) is a systems thinking approach for continuous simulation and an aspect of
systems theory that aims at understanding the dynamic behavior of complex systems.
A systems thinking approach incorporates several tenets such as “Hierarchy” (where complex
wholes are made up of smaller subsystems), “Interdependence” (independent elements can never
form a system, but interdependence of objects and their attributes is inherent in a real system),
“Regulation” (a method of feedback is necessary for the system to operate predictably) (Skyttner
2006). Kirkwood (1998) suggested that a systems thinking approach emphasizes that a system is

96

made up of interacting parts; hence it requires a shift from focusing on event orientation (i.e.,
focusing on isolated events and their causes) to focus more on the internal system structure
(interacting parts in a system) in order to avoid resurfacing of problems in the system. Systems’
thinking focuses on cyclical rather than linear cause and effect relationship among parts of the
system.
System Dynamics was developed in the mid-1950’s by Professor Jay W. Forrester of the
Massachusetts Institute of Technology. It was originally developed by applying principles of
systems theory to industrial systems to improve understanding of the industrial process. And
since then it has been continuously used for policy analysis and design and has been broadened
to include social and economic systems other than industrial systems.
“System dynamics is a powerful methodology and computer simulation modeling technique
for framing, understanding, and discussing complex issues and problems” (Radzicki and Taylor
1997). SD follows an integrative perspective in modeling systems; it recognizes the “information
feedback characteristics so as to show how organizational structure (in policies), and time delays
(in decisions and actions) interact to influence the behavior of the system” (Forrester, 1965).
In this context, it is an appropriate modeling approach to simulate complex systems by
capturing (or sometimes just exploring) the cause-and-effect relationships of the various factors
affecting the behavior of the system. In system dynamics both qualitative and quantitative
modeling approaches are adopted to analyze complex system problems (Morecroft and Sterman
2000). For example, in the qualitative modeling approach, the feedback loop structure is
recognized to develop casual loop diagrams (influence diagrams) by analyzing the cause-andeffect relationships of the various factors affecting the behavior of the system. And then in the

97

quantitative modeling approach, stock and flows are used in the form of mathematical equations
that represent the rate of change of the state variables over time.
3.4.2

SD Modeling Concepts

The key modeling elements and concepts such as dynamic and complex systems, casual and
feedback loop diagrams, stocks and flows etc., are reviewed in more detail in the following
discussion.
As defined in the previous section, a system can be defined as a group of objects coexisting
and mutually interacting towards the accomplishment of a common purpose. A system has also
been defined as “an entity that maintains its existence through the mutual interaction of its parts"
(Gene Bellinger 2004); or as “a combination of parts or elements or components intended to act
together to accomplish an objective” (Dukkipati 2007).
According to Berard (2010), a system is said to be “complex due to the multiplicity of its
elements (natural, technical, economic and social) and of their interactions, but also because of
the diversity of behaviors and properties it can exhibit (dynamic, emergent, etc.)”. Also, Sterman
(2000) suggested that in a system, three main aspects determine the amount of complexity such as
number of elements, number of interrelationships and connection between the elements. And
complex systems are dynamic due to the presence of interactions, dependency on time, complex
internal structure, and delayed behavioral reactions (Sterman, 2000).
Kirkwood (1998) used the term system “to mean an interdependent group of items forming a
unified pattern”. The main part in a systems’ definition in a system dynamics paradigm is the
“interactions of parts” and “interrelationships among elements involved”; these interactions and
interrelationships results in important patterns of behavior of a system. It is important to identify

98

the patterns of behavior in order to build a model that can represent the system characteristics.
The main patterns of behavior that elements/variables present in a system display (Kirkwood
1998; Senge 1990; Senge et al 1994) are shown in Figure 3.7. It represents how a system
variable (such as performance) changes over time showing different patterns of behavior.

Figure 3.7: Graphical representation of patterns of behavior of systems (Kirkwood 1998)
In system dynamics, “causal thinking” is used to better understand the internal system
structures and the reasons that cause the patterns of behavior. A cause, affect, or influences are
99

the key elements used to describe the interacting elements in a system in terms of feedback or
causal loops diagrams. “Feedback is defined as the transmission and return of information” and
a “feedback loop (causal loop) is a closed sequence of causes and effects, that is, a closed path of
action and information,” (Richardson and Pugh 1981). In system dynamics, a system can be
represented either as open or closed, depending on whether output elements influence the input
elements or not; open systems do not take into account feedback whereas closed systems involve
feedback and influence their past system variables (Forrester 1968).
The causal loop diagram consists of system elements and arrows (known as causal links)
with a positive or negative sign on each link to represent the positive or negative influence on the
succeeding element. In a closed system, there are two types of feedback loops: positive loops
(reinforcing) and negative loops (balancing or counteracting loops). Figure 3.8 represents the
positive and negative feedback or causal loop diagram.

+ A
D

+ A

+

(+)

D

B

+

−

(−)

B

+
C

C

+

+

Figure 3.8: Graphical notation of positive and negative feedback or causal loops diagram.
In the above Figure 3.8, the sign at the arrow head represents the influence that a preceding
element would have on the next element. A positive (+) sign indicates that a change in element A
would produce a positive change in element B or a change in the same direction. And a negative
(-) sign indicates that a change in element A would produce a negative change in element B or a
100

change in the opposite direction. And the positive or negative sign in the middle denotes whether
the causal loop is reinforcing or balancing. If the causal loop contains an “even” number of
negative signs then it is a positive (reinforcing) causal loop and if it contains “odd” number of
negative signs then it is a negative (balancing) causal loop.
For example, Figure 3.9 shows an example for sales growth having a combination of positive
and negative casual loops (Kirkwood 1998).

Figure 3.9: Sales Growth example for combination of positive and negative causal loops
(Kirkwood 1998).
A positive (reinforcing) causal loop reinforces change or amplifies the behavior of the system
with more change in the same direction. This usually leads to rapid growth or decline at an ever
increasing rate leading to destabilizing the systems. And thus positive loops are often known as
vicious or virtuous cycles/loops. A negative (balancing) causal loop tends to show to a goalseeking behavior aimed at moving the system towards a desired state. These loop structures aim
at either increasing or decreasing the value of the elements present in order to stabilize the

101

system and reach equilibrium. When both the positive and negative causal loops are present in a
feedback diagram, interactions of both of positive and negative loops may results in a
combination of system behaviors of linear or exponential growth, decline, oscillations, and other
patterns of behavior (Sterman, 2000).
In system dynamics modeling, apart from the patterns of behavior, feedback and casual loop
diagrams, other main building blocks are stocks and flows representing the structure of the
system. Radzicki and Taylor (1997) suggest that the dynamic behavior in system dynamic
modeling “is thought to arise due to the Principle of Accumulation; according to which all
dynamic behavior in the world occurs when flows accumulate in stocks”. The causal loop
diagrams are converted into stock and flow diagrams to represent the dynamic behavior of a
system with the help of mathematical equations that cannot be accomplished with only feedback
or casual loop diagrams.
The stocks (also known as levels) are used to represent the state variables that undergo a
change in a system. Stocks are defined at a particular point of time and have a definite value.
They can be a measure of the quantity accumulated in them by inflows and depleted by outflows.
The flows (also known as rates) are used to represent the actions that produce or control the
changes in the stocks. They are usually measured over a period of time and represent both the
inflows and outflows from the stocks. Stocks can be represented by rectangles and flows by
valves.
3.4.3

SD Modeling Tools

In system dynamics, modeling and analysis of complex dynamic system is done with the use of
computer simulation tools. These simulation tools help to visually depict and communicate the

102

mental models of complex dynamic systems with much ease and better understanding of the real
system.
Sterman (1992) provided the following advantages of using computer models over mental
models (Sterman 1992):
1. Mental models are “dynamically deficient, omitting feedbacks, time delays,
accumulations, and nonlinearities”
2. Computer models are explicit and their assumptions are open to all for review.
3. They can accurately compute the logical consequences of the modeler’s assumptions.
4. They can be simulated under controlled conditions, allowing analysts to conduct
experiments which are not feasible or ethical in the real system.
In the late 1950s and early 1960s, System Dynamics evolved from hand-simulation of mental
models to computer modeling and simulation due to the efforts of Forrester and a team of
graduate students (Radzicki and Taylor 1997). SIMPLE (Simulation of Industrial Management
Problems with Lots of Equations) was the first system dynamics computer modeling language
developed by Richard Bennett in 1958 and a year later in 1959, an improved version of SIMPLE
known as DYNAMO (DYNAmic MOdels) was developed by Phyllis Fox and Alexander Pugh
that was used in the industry for several decades (Richardson and Pugh 1981; Forrester 1995).
DYNAMO is based on “writing source equations and lacks a modern graphical user interface”
(Madachy 2008). After DYNAMO, the system dynamics simulation tools have evolved and
improved to a great extent and now include visual interface along with custom programming and
extensive numerical analysis features. Some of the present system dynamics simulation tools are
Extend, iThink/Stella, Powersim, and Vensim.

103

3.4.4

SD Application

System dynamics can be applied to a wide range of systems where state variables change over
time. It was originally developed to analyze business management systems but since its origin it
has been successfully applied to a number of social-economic systems (Forrester, 1961),
business systems (Sterman 2000), environmental systems (Ford 1999), healthcare systems
(Homer and Hirsch 2006), agricultural systems (Saysel et al 2002), urban and global systems
(Forrester and Collins 1969, Forrester 1971) etc.
System dynamics has proven to be an effective tool for modeling and aiding the decision making
process for engineering and construction project management. Several issues ranging from high
number of construction changes, rework, resource allocation, quality issues, and risk analysis
have been successfully addressed with the help of system dynamics.
For example, Love et al (2002) used system dynamics to better understand change and rework in
construction project management systems. Quality issues facing the construction industry were
identified by modeling the internal and external dynamics in a project using the feedback loop
diagrams. Change and rework issues were also addressed by Park and Pena-Mora (2003) by
providing a dynamic planning and control methodology (DPM) model to predict the changes in
construction affecting the project performance. The model identified the feedback process
resulting from construction changes by determining the behavior patterns of construction change
and analyzed change impact on the construction performance using system dynamics. However,
uncertainties resulting from changes were not explicitly addressed by the DPM model. Lee et al
(2005) developed an enhanced DPM model, aimed at protecting the planned performance against
resulting from errors and changes. Feedback processes on quality and change management

104

process were identified along with development of construction process model with the help of
system dynamics.
In addition, system dynamics has been successfully applied in construction systems to explore
performance enhancement in construction organizations (Ogunlana et al 2003), construction risk
analysis (Nasirzadeh et al 2008), dynamic resource management for construction projects (Park
2005), for managing constructibility reviews to reduce highway project durations (Damron et al
2004), to explore the effects of resource allocation policies on project durations (Lee et al 2004),
for managing the dynamics of projects and changes (Cooper and Lee 2009), Lean Construction
(El-Gafy and Abdelhamid 2010),evaluating the alternative of type in construction and demolition
waste (Zhao et al 2011).
3.4.5

SD Advantages and Limitations

Sterman (1992) suggested that large scale projects belong to a class of complex dynamic systems
and are characterized by the following:
1. High complexity, involving multiple interdependent components
2. Highly dynamic
3. Multiple feedback processes
4. Non-linear relationships
5. Both hard (direct) and soft (indirect) data
System dynamics has proven useful in representing the construction system involving these
above characteristics that are difficult to model in the DES. However, SD is unable to represent
the process details involved in construction operations that can be easily modeled using DES

105

(Williams 2002). DES models have mainly addressed the operational issues whereas SD models
have been more successful in addressing the management and strategic issues.
3.5

Chapter Summary

This chapter provides an overview of the simulation approaches such as Discrete Event
Simulation and System Dynamics modeling in detail along with discussing the applications,
advantages and limitations of each of the modeling techniques in order to gain better
understanding of the process of building simulation models.

106

4 LEAN BASED CREW DESIGN

4.1

Introduction

Construction work is executed by a variety of speciality trades, using group of workers
commonly known as crews. Yet decisions regarding how to design crew make-up and improve
their performance are seldom formally addressed. Designing a crew to carry out an operation in
a dynamically changing construction environment is a complex task; still it has usually been
approached with assumptions or self-discretions of the supervisor/foreman that rarely materialize
and at times lead to low performance.
Performance of these crews is very crucial to the work being executed and consequently for
the overall project performance. Any change in the crew size, composition, and other important
crew design parameters might result in irreversible spiralling consequences to the cost, duration,
and other performance outcomes. However, despite the clear significance of crew design, only a
modest scholarly interest in this area has been seen in the last two decades.
The aim of this research was to develop a lean-based crew design model that integrate both
existing crew design and work design research and lean principles for designing and improving
the performance of crews. This was achieved by conducting an extensive literature review and
investigating and synthesizing research studies on “work design/team design” that have well
developed theoretical and empirical foundations in lean production, lean construction, sociotechnical system theory, social and organizational psychology.

107

This chapter presents the

resulting model, parts of which will be tested in the following chapter between Discrete Event
Simulation and Systems Dynamics.
4.2

Crew and Work Design Concept

Crew design is often confused with construction process design (method or operation design).
However, in this research we use crew design as an integral part of any production process that
involves the analysis and redesign of crew make-up. As discussed in the previous sections, crew
design involves both the design of “group of workers/people” and the “work/job” that needs to
be executed by them.
There are numerous decisions to be made regarding how the work is to be executed, how
many workers need to be in a crew/team, what should be the composition of workers in a crew,
sequencing of tasks, how the work should flow among the workers in a crew, and so on. Design
of only one aspect would not lead to a joint optimization of both work and team/crew. Work
teams have been common both in manufacturing and service industry. Optimizing the
performance of crews (production work teams) is a very essential element as “work teams are
considered to be an integral tool aiding continuous improvement in work operations” (CutcherGershenfeld and Associates 1994).
For this research, the Cohen and Bailey (1997) and Sundstrom (1990, 1999) classification is
used where work teams are defined as “continuing work units responsible for producing goods
and providing service,” as it fits very well with the description of a “crew” for a production
setting in construction.

108

4.3

Lean Crew Design Characteristics

The lean crew design characteristics were developed by synthesizing the existing crew and work
design approaches from lean production, lean construction, socio-technical system theory, and
social and organizational psychology (as conceptually depicted in Figure 4.1).
The crew design characteristics are those that should be considered during the crew design
process.

Socio-Tech
Social &
Organization

Lean
Construction

Lean
Production

Lean Crew Design Characteristics
Figure 4.1: Synthesis of existing crew and work design approaches.
In general, performance is the ability to meet preset targets or standards. In this research, targets
for a crew are set based on: (1) satisfaction; (2) motivation; (3) learning; (4) pride in work
performed; (5) cost of work performed; (6) duration of work performed; (7) quality for work
performed; (8) productivity. The lean crew design characteristics that would lead to high crew

109

performance are shown in Figure 4.2. In Figure 4.2, lean crew design characteristics are
comprised of “primary crew design characteristics”, “intermediate crew design characteristics
and outcomes” and “final crew performance outcomes”. The primary crew design characteristics
are divided into two groups (1) work characteristics and (2) flow characteristics. These are
discussed in the following sections.

110

Figure 4.2: Lean Crew Design Characteristics
111

4.3.1

Flow Characteristics

The flow characteristics have been derived from Lean Production (Shingo 1989; Ohno 1988;
Womack et al. 1990; Womack and Jones 1996) and Lean Construction (Howell and Ballard
1999; Koskela 1992, 2002). The primary flow characteristics are aimed at the elimination of
waste in order to achieve continuous flow for the operations executed by the crew.
The flow characteristics are grouped into 7 types of waste characteristics: (1) Motion; (2)
Transport; (3) Over-processing; (4) Over-production; (5) Inventory; (6) Waiting; (7) Defects.
These primary flow characteristics are further divided into intermediate crew design
characteristics and outcomes. The causal relationships are provided in Figure 4.3 to better
understand and develop the feedback or casual loop diagrams in the next chapter of simulation
model building. In Figure 4.3, the plus sign denotes an increase (positive effect) and a minus
sign denotes a decrease (negative effect) on the characteristic or outcome.
The crew should be designed in a way that reduces or eliminates all the seven types of
wastes. Excessive Motion (Figure 4. 3) is one of the seven wastes in the flow characteristics that
should be kept to minimum in order to enhance crew performance in terms of duration and cost
of operations. Excessive motion may lead to more moves than needed for the crew. Any increase
in the number of moves or the complexity of moves would increase the processing time a crew
takes to execute an operation. In turn, this increases the processing time, leading to an increase in
cycle time and the processing cost associated with an increase in the labor cost. Hence, in order
to improve the crew performance, excessive motion should be reduced and eliminated.
Excessive Transport (Figure 4. 3) has also been found to have a direct correlation with cycle
time and operating costs. More transport would mean that both the labor and equipment are

112

subjected to travel and operate over long distances that would increase the operating time leading to an increase in cycle time and
operating costs for both labor and equipment. Hence in order to improve crew performance, excessive transport should be reduced or
eliminated.

Figure 4.3: Causal relationships between primary & intermediate crew design (flow) characteristics and the final crew performance
outcomes (Part A).
113

Figure 4.4: Causal relationships between primary & intermediate crew design (flow) characteristics and the final crew performance
outcomes (Part B).
114

Another waste that has direct correlation with increasing the cycle time and processing cost is
Over-processing (Figure 4. 3) increases the numbers of steps that are employed to carry out an
operation by a crew than that are actually required to deliver value. Increased number of steps
then in turn increases the processing cost and cycle time of the operation. Hence in order to
improve crew performance, over-processing should be reduced or eliminated.
Over-production (Figure 4.4) is considered to be the most serious waste, and should be
avoided whenever possible. Over producing products/work that cannot be completed or finished
leads to work-in-progress in the intermediate steps involved in an operation. This is a major
cause for overburden among the workers executing the operation. This overburden among the
workers in turn may lead to variation in the operation leading to non-uniform flow that in turn
leads to increase in the cycle time of operations. Hence, a crew, and their work, should be
designed in way that it is not over-producing
Another waste that leads to work-in-progress is excessive Inventory (Figure 4.4). More
inventory than necessary at times leads to an increase in the batch-size. And increasing the batch
size without increasing the production capacity of the crew would lead to more work-in-progress
along with other kinds of wastes (e.g., transportation). Hence in order to improve crew
performance, minimum inventory should be kept that can be finished/completed by the crew
without leading to any work-in-progress.
In addition, a crew should be designed in way that minimizes the waste of Waiting. Waittime among the crew would mean a decrease in the resource utilization of the crew leading to an
increase in the cycle time. Also if the crew is not utilized close to, but not at, its maximum
capacity, that would mean a decrease in the overall productivity of the crew.

115

And finally, a crew should be designed in such a way that they execute an operation with
minimum Defects (Figure 4.4). An increase in the number of defects would lead to an increase in
amount of inspection and rework that needs to be done. Both rework and inspection would lead
to an increase in the processing time that would in turn increase the cycle time and processing
cost.
4.3.2

Work Characteristics

The “Work Characteristics” (derived from socio-technical system theory, social and
organizational psychology) have been based on significant studies conducted by other authors
(Gladstein 1984; Hackman 1980, 1987; Hackman and Oldham 1980; Guzzo and Shea 1992;
Niepcel and Molleman 1998; Steiner, 1972; Davis and Wacker, 1987; Campion et al 1993, 2005;
Morgeson and Campion 2003; Morgeson et al 2005, 2006, 2008; Medsker and Campion 2007;
Cohen and Bailey 1997; Leach et al 2005; Kozlowski and Bell 2003).
Crew, Task, and Social, are the further division of “Work Characteristics” (Figure 4.2). For
‘Crew’, the factors are size, flexibility and heterogeneity.
Crew size (Figure 4.5) is the number of individuals in a crew/team. Increasing the size of
crew does increase the production capacity of crew and thereby may increase the overall output
of crew for that operation. But at the same time, increasing size also increases the cost of labor
increasing the total cost to execute that operation. Increasing the crew size may also lead to overcrowding if sufficient space is not available. And over-crowding may lead to more conflict and
less coordination among the crew members that may finally lead to less motivation, satisfaction
and learning among the crew members to work together to execute an operation.

116

Crew flexibility (Figure 4.5) is the ability of the crew members to perform each other’s’
task/work; it could also be known as synonym for multi-tasking. If there is more flexibility
among the crew members, it would mean that apart from performing their own tasks, they would
be able to help their crew members in their tasks/work if required. This would save time taken to
execute an operation by a crew as idle crew members would be able to perform others’ tasks and
shorten the operation duration. Also this would lead to direct savings in the number of workers
needed in a crew reducing the labor costs, as the need for specialized labor would be reduced
having crew members which can multi-task. Finally, having more crew flexibility would also
lead to satisfaction and learning among the crew members arising from the feeling of helping
their team/crew members and learning from them as a result of working together.
Crew heterogeneity relates to how a crew is composed in terms of its member’s experience,
skills, etc. A more heterogeneous crew would mean crew members have a more variety of skills,
knowledge, and experience level. If there is variety in the crew, then the more experienced crew
members would be able to provide support (technical) in terms of sharing their expertise and
knowledge with other crew members and encourage learning of other crew members. Having a
more heterogeneous crew may also lead to more motivation among the crew either to provide
support or to learn from others crew members.
The ‘work characteristic’ Task

is itself made of significance, variety, identity,

interdependence, complexity, and experience, all are significant characteristics affecting the
crew performance.
Task significance means the extent to which the crews’ work has significant consequences
and impact on the project success. It would be logical to say that if a crew is involved in a
significant task/work, they would be more motivated towards executing the task/work with the
117

best of their abilities compared to if they are involved in a mundane or not so significant
task/work. Consequently, with an increase in task/work significance, a crew may be more
satisfied of executing it than if the task/work was simple and not so significant.
Task variety means the extent to which workers are involved in a variety of tasks; it relates to
the concept of job-enlargement from socio-technical theory allowing both interesting and dull
tasks to be shared among members. Having more task variety would mean crew members would
be more satisfied while executing an operation as they would not be involved with doing the
same task/work over and over again. This would also mean more learning among the crew as
they would be involved in a variety of tasks.
Task identity is the degree to which the group/individual completes an entire or separate
piece of work from start to finish. And more the crew members can relate themselves with a
particular task/work and be able to take ownership of the work performed, the logical it seems
that it may increase worker motivation by instilling a feeling of pride. And also completing a
task from starting to end may also lead to feeling of satisfaction among the crew members.
Task complexity can be defined as the extent to which a task/work is difficult to perform.
Increase in the complexity of tasks/work performed by the crew may lead to an increase in the
processing time, and may also reduce the quality of work performed due to the complex nature
of work. And as discussed earlier, any increase in the processing time would lead to an increase
in the processing cost and cycle time. Further, it is also argued that reducing the complexity or
simplifying work should generally lead to productivity gains for the crew.
Task experience is the amount of time spent by a team/individual in performing a task; the
number of times a similar type of task has been performed; or the prior amount of knowledge an
individual has in performing that task. More task experience would mean having more expertise
118

and knowledge in carrying out a particular task. It may lead to higher productivity of crew
members having higher experience and consequently may lead to higher labor costs of crew
members having higher experience. Another benefit of having crew members with high task
experience would be in terms of support (technical) they can provide to other members of the
crew and also lead to learning among the crew members.
Task Interdependence exists when a group of individuals get together and depend on one
another to perform a certain work consisting of several tasks; and either start or finish of one task
is dependent on the other tasks. It is related to cycle time, work-in-progress or in-process
inventory, overburden, and work flow disruptions. High interdependence may lead to a
continuous overburden leading to work flow disruptions (high task interdependence makes it
essential to finish the task on or before time so that the tasks down the line are not stopped or
waiting because of the incomplete tasks preceding them). And as discussed earlier high
overburden is directly correlated with variability, low quality, and non-uniform flow leading to
an increase in cycle time.
The ‘work characteristic’ Social is itself made of autonomy, self-management, feedback, and
support, all are also significant characteristics affecting the crew performance.
Autonomy means the “amount of freedom and independence an individual has in terms of
carrying out his or her work assignment. Increased autonomy in execution of work leads to
increased work satisfaction and motivation among the workers.
Self-management is the group level analogy to Autonomy and is presumed to enhance
performance or effectiveness. Providing increased autonomy to the crew would make them more
responsible towards the outcomes of their work by increasing their sense of ownership of the
work. However, the literature indicates that moderate amount of autonomy works better than no
119

or full autonomy to teams/crews to execute their work. Usually a top-down approach is adopted
by management to improve the performance of the crews/teams; but giving a team/crew
moderate level of autonomy would only help them to identify and suggest solutions for root
cause problems occurring at the production level in the tasks/work executed by them. Also selfmanagement of work by the team/crew increases the sense of pride in the crew as they are able to
manage their work by themselves by increasing the worker’s sense of ownership of the work.
Further, both Feedback and Support are significant characteristics in the design of
crews/teams. Proper feedback from the supervisors or management about the work executed by
the crew/team would lead to less conflict among the crew members. Having timely feedback
would also provide a feeling of satisfaction among the crew and also motivate them to improve
their performance if required. This would finally lead to learning among the crew as they would
be aware of their performance about the work executed by them.
Support is ability to get assistance, help, and advice from crew/team members, supervisors
and management in both the technical and social aspects. Opportunities for either technical or
social support would logically lead to a higher team/crew performance by increasing the
motivation, satisfaction and learning among the crew.
As a summary of the causal relationships among the primary and intermediate crew design
characteristics and final crew performance outcomes, Figure 4.5 was constructed. The
knowledge of these cause and effect relationships is of value in determining the optimal crew
design. In this graphical representation, the plus sign denotes an increase (positive effect) and a
minus sign denotes a decrease (negative effect) on the characteristic or outcome.

120

Figure 4.5: Causal relationships between primary & intermediate crew design (work) characteristics and the final crew performance
outcomes (Part A).

121

Figure 4.6: Causal relationships between primary & intermediate crew design (work) characteristics and the final crew performance
outcomes (Part B).

122

Figure 4.7: Causal relationships between primary & intermediate crew design (work) characteristics and the final crew performance
outcomes (Part C).

123

The primary team/crew design (flow and work) characteristics and intermediate crew design
characteristics and performance outcomes can be defined in terms of hard (direct factors
affecting the crew performance) and soft (indirect factors affecting the crew performance) factors
(Figure 4.5 and 4.6).

HARD FACTORS AFFECTING THE CREW PERFORMANCE
CREW SIZE

PROCESSING

OPERATING

PRODUCTION

RESOURCE

# OF MOVES

OPERATING

OPERATING

CYCLE TIME

RESOURCE

# OF STEPS

PROCESSING

WORK-IN-

BATCH SIZE

Figure 4.8: Hard factors affecting the crew performance.

SOFT FACTORS AFFECTING THE CREW PERFORMANCE
CREW SIZE

TASK

AUTONOMY

COMPLEX

CONFLICT

CREW

TASK VARIETY

SELF-

REWORK

COORDINATIO

CREW

TASK IDENTITY

FEEDBACK

INSPECTION

OVER-

TASK

SUPPORT

DEFECTS

OVERBURDEN

TASK

TASK

VARIABILITY

Figure 4.9: Soft factors affecting the crew performance.
4.4

Lean Crew Design Guidelines

The objective in this stage is to make sure that the crew has a design that would lead to
maximizing their performance. The guidelines for Lean Crew Design are proposed by primarily
reviewing research studies on “work design/team design” that have well developed theoretical

124

and empirical foundations in lean production, lean construction, socio-technical system theory,
and social and organizational psychology.
The proposed guidelines lead to the joint optimization of both the crew and the work
performed, resulting in better overall crew performance. In this research, crew performance is
defined in terms of: (1) duration of work performed; (2) cost of work performed; (3)
productivity; (4) quality of work performed; (5) worker satisfaction; (6) worker learning; (7)
worker motivation; (8) pride in work performed.
The crew performance is optimized by (1) “Minimizing Waste” in the construction
operations and (2) “Maximizing Value” generation in the construction operations. “Minimizing
Waste” in the work performed by the crew is achieved by eliminating or reducing the seven
types of wastes discussed in the previous section (integrated lean crew design framework)
namely: (1) Transport; (2) Inventory; (3) Motion; (4) Waiting; (5) Overproduction; (6) Over
Processing; (7) Defects. And “Value” in the design of crews is maximized by optimizing the (1)
duration of work performed; (2) cost of work performed; (3) productivity; (4) quality of work
performed; (5) worker satisfaction; (6) worker learning; (7) worker motivation; (8) pride in work
performed. Maximum “Value” in the design of crews is delivered by optimizing all the eight
performance outcomes; however the crew could also be designed based on the either of the
targeted performance criterion. Guidelines for optimizing each of the eight performance
outcomes are presented in the following section:
(1) The “Duration of Work performed” can be optimized by improving and keeping the cycle
time of the operation to the minimum. Steps involved in reducing the cycle-time of
operations are as follows:

125

1) Simplification of process and reduction of processing time
i.

Reduce the number of unnecessary steps in a process
A construction operation consists of a number of processes and each process has
number of steps known as activities. The design of the crew in terms of crew size
and crew composition depends on the number of steps required to execute a
particular process. More number of steps in a process would increase the total
processing time resulting in higher cycle time and increasing the duration of work
performed by a crew. Hence, crew design (design of “group of workers/people” and
the “work/job” that needs to be executed by them – see section 4.2) should target a
minimum number of steps in a process.

ii.

Reduce the number of moves required to execute each step in a process
A number of moves (tasks) are required to execute each step (activity) in a process.
More number of moves to execute each step in a process would increase the total
processing time resulting in higher cycle time and increasing the duration of work
performed by a crew. Hence, crew design should target a minimum number of
moves required to execute each step in a process.

iii.

Reduce the task complexity/number of complex moves in each step in a process
Task complexity or the number of complex moves in each step of the process is
directly proportional to the processing time of the work performed by the crew.
Higher task complexity or more number of complex moves in a process would
increase the total processing time resulting in higher cycle time, leading to longer
overall work duration. Hence, crew design should focus on simplifying tasks and
moves involved in the process.

126

iv.

Reduce the operating distance to execute each step in a process
The operating distance required to execute each step in a process is directly
proportional to the processing time of the work performed by the crew. Higher
operating distance to execute the steps in a process would increase the total
processing time resulting in higher cycle time, leading to longr duration of work.
Hence, crew design should consider minimizing the operating distances that crew
has to traverse.

v.

Increase the movement speed for the operating distances
Increasing the movement speed for the operating distances to execute the steps
involved in a process would decrease the processing time resulting in reduced cycle
time.

Crew design should therefore increase the movement speed over the

operating distances.
2) Reduction of non-processing time
i.

Reduce transportation/motion time
Non-processing time such as excess transportation/motion time does not add value
and increases the cycle time of the process and correspondingly the duration of the
work performed. Any change in either the size or composition of the crew affects
the amount of transportation/motion required to execute a process. Hence, crew
design should target a minimum transportation/motion involved in the work
performed by a crew.

ii.

Reduce wait time
Wait time in a process increases the cycle time and correspondingly results in the
increase in duration of the work. Wait time in a process does not add any value.

127

Any change in the crew size and composition would also change the wait time in
the process being executed by the crew. Crew design should therefore minimize
wait times where possible.
iii.

Reduce set-up time
Set-up time in a process increases the cycle time and correspondingly results in the
increase in duration of the work. Set-up time does not have any direct value to the
process. A crew design should consider reductions in set-up times using
administrative or engineering interventions.

iv.

Reduce inspection time
Inspection time in a process increases the cycle time and correspondingly results in
the increase in duration of the work. Inspection time does not add any direct value
to the process. Hence, crew design should incorporate in-line or built-in inspections
(manual or automated) to reduce the overall inspection times.

v.

Reduce rework
Rework in a process takes up considerable amount of time in delivering the finished
product while increasing the cycle time and correspondingly resulting in the
increase in duration of the work. Rework time does not add any direct value to the
process. It could be reduced by doing the work correct the first time, which is a
critical component of a crew design process often known as built-in quality (job
done right the very first time leading to reduced rework in the process).

128

3) Reduction of Variation
i.

Level the work load among each step of the process
Variation in a process leads to an increase in the cycle time and correspondingly
results in the increase in duration of the work. Variable workload may lead to
increased wait times in a process and uneven flow of work in a process. It can be
reduced by leveling the workload among the different production units (production
steps) in a process. Hence, crew design should target making the operation
production rate of each production unit (production step) to be equal to the
maximum production rate in that process leading to continuous and uniform flow of
work.

ii.

Reduce overburden (muri)
Overburden (muri) results in variable work rate resulting in increased variation in a
process; that may again increase the cycle time and the duration of the work. Hence,
a crew design that eliminates or reduces overburden is critical.

iii.

Reduce work-in-process (WIP)
Work-in-process occurs when more products are produced than can be processed or
installed completely; it is the work that has been started but not completed and is
still in process. It leads to overburden, variation and finally increases the cycle time
to finish each product and correspondingly results in the increase in duration of the
work. Hence, crew design is to reduce the work-in-process as much as possible.

iv.

Reduce the batch size
The batch size has significant effect on the cycle time and duration of work.
Increased batch size leads to increase in wait times, work-in-process, overburden,

129

and more variability in a process. Hence, crew design should consider a set-up that
leads to minimum batch sizes in a process.
v.

Apply pull system
Pull systems are very effective in reducing over-production leading to the better
cycle time and duration of work. Hence, a crew design should integrate pull systems
such that that the products are pulled by the production units (crew members)
whenever needed.

(2) The “Cost of Work performed” can be optimized by the following steps:
1) Keep the crew size to minimum
2) Increase crew flexibility
3) Reduce processing cost
4) Reduce non-processing cost
5) Reduce defective products
1) Keep the crew size to minimum
The crew size has a direct correlation with the cost of labor and correspondingly with the
cost of work performed. Increasing the crew size more than is necessary to maintain the
flow of work leads to increasing the cost of work performed. Hence, crew design should
target the minimum crew size that will accomplish the intended operation. .
2) Increase crew flexibility
The crew flexibility is used as a synonym for multi-tasking. It is the ability of the crew
members to perform each other’s task/work. Increasing crew flexibility enables the crew
members to help each other in executing the tasks of a process other than their own tasks.
Higher crew flexibility saves times as idle crew members are better utilized leading to

130

shortening the duration of a process and as a result save on labor costs. Also, with
increased crew flexibility, and crew members involved in multi-tasking, the need for
specialized labor for each task would be reduced leading to further optimizing the cost of
work performed. Hence, a crew design should consider opportunities for crew flexibility.
3) Reduce processing cost
The cost associated with processing and performing the work should be reduced in order
to optimize the total cost of work performed. Guidelines on designing a crew leading to
reduced processing cost such as reducing the number of unnecessary steps in a process,
reducing the number of moves required to execute each step in a process, reducing the
task complexity/number of complex moves in each step in a process, reducing the
operating distance to execute each step in a process, increasing the movement speed for
the operating distances have been discussed in detail in the previous section.
4) Reduce non-processing cost
The non-processing cost associated with unnecessary or excessive transportation/motion
time, wait time, set-up time, inspection time, rework should be reduced in order to
optimize the total cost of work performed. Guidelines on designing a crew leading to
reduced non-processing cost have been discussed in detail in the previous section.
5) Reduce defective products
The cost associated with the defective products increases the total cost of the work
performed. More number of defective products leads to more rework and inspection
required, leading to increase in the processing time, cycle time and correspondingly
resulting in increased labor and processing costs. As discussed in detail earlier, crew

131

design should target least amount of rework and inspection involved to complete a
process.
(3) The “overall crew productivity” can be optimized by the following steps:
1) Increase task experience of crew
2) Reduce overburden
3) Apply advanced technology to labor and equipment
4) Increase resource utilization
5) Increase crew flexibility
6) Increase crew heterogeneity
7) Reduce task complexity
8) Reduce task interdependence
1) Increase task experience of crew
Task experience is directly correlated to the crew productivity to execute the work.
Higher task experience or inclusion of more experienced workers is associated with
higher labor productivity and correspondingly higher overall crew productivity. Hence,
crew design should enable improvement of task experience and familiarity to execute
more number of tasks in process than a single fixed task. Also, the crew design should
include more experienced crew members in order to enhance the overall crew
productivity.
2) Reduce overburden
Reducing overburden was discussed as a way to reduce variation in crew performance.
The same applies here, wherein overburden is also inversely proportional to the

132

productivity. A crew design that eliminates or reduces overburden is critical for overall
crew productivity. Apply advanced technology to labor and equipment
Applying advanced technology to labor and equipment is associated with higher
productivity for work to be performed. Hence a crew design should incorporate advanced
technology where suitable and economical.
3) Increase resource utilization
Resource utilization is directly proportional to labor productivity. Increasing resource
utilization is associated with higher crew productivity. However, in Lean Construction
there is caution in making utilization a focus because it is possible to slip into local
optimization and not system optimization. In other words, making one crew faster by
being more productive will only benefit that crew and not the project as whole.
Therefore, while a crew design should target higher resource utilization, this is to be
balanced with the productivity of other crews that is possible through the use of the Last
Planner System® – a topic outside the scope of this thesis. .
4) Increase crew flexibility
Crew flexibility was discussed earlier, and concluded that a crew design should consider
opportunities for multi-tasking.
5) Increase crew heterogeneity
Crew heterogeneity as discussed earlier is positively associated with crew productivity.
Increasing crew heterogeneity leads to higher variety of skills, knowledge, and
experience among crew members that correspondingly leads to higher overall crew
productivity. Hence, a crew design should target higher crew heterogeneity leading to
higher overall productivity.

133

6) Reduce task complexity
As discussed earlier, task complexity has a major impact on work duration, making it
directly proportional to productivity. Hence, crew design should focus on simplifying
tasks and moves involved in the process..
7) Reduce task interdependence
Task interdependence refers to the relation between the series of tasks that a crew has to
carry out in order to complete a construction operation – it can also be extended to mean
interdependence between construction operations, which is not the focus of this thesis.
As discussed in Chapter 2, task interdependence among the crew members can exist in
the form of pooled, sequential or reciprocal, pooled being least and reciprocal being the
highest form of interdependence. Lower task interdependence is associated with higher
work-flow reliability and correspondingly higher overall crew productivity. Hence, crew
design should target the least amount of task interdependence (pooled interdependence)
among crew members.
(4) The “quality of work performed” can be optimized by the following steps:
1) Reduce defective products
(Proceeding discussion applies)
2) Increase task experience
Task experience is positively associated with the quality of work performed. Higher task
experience of crew members’ results in better quality of work performed, since the crew
members are equipped with better knowledge and skills to execute the tasks they are
more familiar or have experience in. Hence, crew design should enable improvement of

134

task experience among the workers and inclusion of more experienced workers leading to
better quality of work performed.
3) Reduce task complexity
(Proceeding discussion applies)
4) Reduce overburden
(Proceeding discussion applies)
5) Reduce task interdependence
(Proceeding discussion applies)
(5) The “worker satisfaction” can be optimized by the following steps:
1) Reduce conflict
2) Increase coordination
3) Increase crew flexibility
4) Increase crew heterogeneity
5) Increase support
6) Increase task significance
7) Increase task variety
8) Increase task identity
9) Increase self-management
10) Provide reasonable autonomy
11) Increase feedback
1) Increase coordination
Crew coordination is associated with crew satisfaction. Higher coordination is likely to
produce more satisfaction among the crew to work together. Further coordination among

135

crew members is influenced by over-crowding/congestion that is influenced by the
change in crew size. Increase in over-crowding/congestion provides fewer opportunities
for coordination and therefore lower satisfaction among crew members. Hence, the crew
size should be selected in such a way that enables maximum coordination leading to
higher satisfaction among crew members.
2) Reduce conflict
Conflict among crew members is also associated with crew satisfaction. Higher conflict
is likely to produce less satisfaction among the crew to work together. Further conflict
among crew members is influenced by over-crowding/congestion that is influenced by
the change in crew size. Increase in over-crowding/congestion provides more chances for
conflict and therefore lower satisfaction among crew members. Hence, the crew size
should be selected in such a way that enables least conflict leading to higher satisfaction
among crew members.
3) Increase crew flexibility
(Proceeding discussion applies)
4) Increase crew heterogeneity
(Proceeding discussion applies)
5) Increase support
Support in terms of both social and technical as discussed in Chapter 2 is associated with
crew satisfaction. More support (social or technical) is likely to produce higher
satisfaction among crew members. Hence, a crew should be designed in such a way that
enables support in the form of assistance, help, and advice from crew members,
supervisors and management leading to higher satisfaction among crew members.

136

6) Increase task significance
(Proceeding discussion applies)
7) Increase task variety
Task variety is associated with crew satisfaction. Higher task variety leads to higher
satisfaction among crew members, since with increased task variety crew members would
be able to execute and perform a variety of tasks and not restricted to only a fixed task
over and over again. Hence, a crew design that enables inclusion of a variety of tasks in a
process and also provide equal opportunities to all crew members to perform them is
highly desirable.
8) Increase task identity
Task identity is also associated with crew satisfaction. Higher task identity leads to higher
satisfaction among crew members, since the crew members can relate themselves with a
particular task and be satisfied in completing it from start to finish. Hence, crew design
should target high task identity leading to higher satisfaction among crew members.
9) Increase self-management
Self-management is also positively related to crew satisfaction. It is the group level
analogy to autonomy and means how well a crew is able to manage and execute a process
on its own without much external support (social/technical). Higher self-management in
the crew would lead to higher satisfaction among the crew members, since they would be
able to manage and execute a process on their own with the need of external support.
Hence, crew design should enable higher self-management in the crew leading to higher
crew satisfaction.

137

10) Provide reasonable autonomy
Autonomy is the amount of freedom and independence a crew member has in performing
his tasks/work. Reasonable autonomy is positively associated with higher satisfaction
among crew members, since it is more likely that if certain autonomy is given to the crew
members to decide about how to execute a given task and if top down management
approach is avoided, reasonable autonomy will lead to higher crew member satisfaction.
Hence, crew design should afford reasonable crew autonomy in order to optimize crew
satisfaction.
11) Increase feedback
Feedback is also associated with crew satisfaction. Proper feedback within the crew
members and from the supervisors and higher management leads to less conflict in the
crew as the crew is likely to be more aware of their performance of work after a timely
feedback, and correspondingly be more satisfied. A crew design should therefore provide
for proper and timely feedback, leading to higher crew satisfaction.
(6) The “worker learning” can be optimized by the following steps:
1) Increase coordination
Crew coordination is also associated with crew learning. Higher coordination leads to
more interaction among the crew members; and correspondingly results in learning
induced in the crew while coordinating and working together towards a common goal.
Hence, a crew should be designed in such a way that enables maximum coordination
leading to higher learning among crew members.
2) Increase crew flexibility
(Proceeding discussion applies)

138

3) Increase crew heterogeneity
(Proceeding discussion applies)
4) Increase task experience
(Proceeding discussion applies)
5) Increase task variety
(Proceeding discussion applies)
6) Increase feedback
(Proceeding discussion applies)
(7) The “worker motivation” can be optimized by the following steps:
1) Increase coordination
(Proceeding discussion applies)
2) Reduce conflict
(Proceeding discussion applies)
3) Increase crew heterogeneity
(Proceeding discussion applies)
4) Increase task significance
(Proceeding discussion applies)
5) Increase task identity
(Proceeding discussion applies)
6) Provide reasonable autonomy
(Proceeding discussion applies)

139

(8) The “pride in work performed” can be optimized by the following steps:
1) Increase task identity
2) (Proceeding discussion applies)Increase self-management
(Proceeding discussion applies)
4.5

LBCD Model Development

In this research, the structure of the Lean-based Crew Design (LbCD) model is patterned after a
transformation process structure where in inputs are converted to outputs by means of a
conversion process. The proposed LBCD model is shown in the Figure 5.1 below.

INPUT

PROCESS

OUTPUT

Lean Crew Design
Characteristics

Crew Design
Process

Lean-based Crew
Design

FORMATION

EXECUTION

REFINING

Figure 4.10: Lean-based Crew Design (LbCD) Model

As shown in Figure 4.1, the model input consists of the lean crew design characteristics,
developed based on existing crew and work design approaches. The characteristics include both
soft and hard factors from several scientific disciplines. The characteristics are used in the design
process to develop the final crew design that will achieve the joint optimization of both the crew
and work performed. The crew design process itself consists of three main phases, namely,
formation, execution, and refining.
140

4.6

Lean Crew Design Process

The lean crew design process consists of three phases namely: (1) formation, (2) execution, and
(3) refining. The three phases are proposed to guide the crew design process in a production
setting. It consists of the initial formation of the crew and structuring of work involved, followed
by execution and refining of the crew.
In the first phase of formation, the initial formation of the crew is done along with structuring
of the work. In formation of crew, the estimation and scheduling teams determine the standard
crew, number of activities required to perform the particular work, list of tasks in a activity and
activity/task durations. The initial size and composition of the crew is determined based on the
amount of work/quantity that needs to be done generally by the using the standard construction
industry data (example RS Means) or the contractors’ own data and other work/job specific data.
From these data, standard crew productivity is determined, which is used to calculate the
activity/task durations or the standard unit crew days/time required to execute the work.
After the initial standard crew size and makeup, list of activity and tasks, and activity/task
durations are determined, the structuring of work should be done by determining the sequence of
tasks involved in an activity. Once the task sequence is determined, amount of work that needs to
be executed by each crew member needs to be determined (including the amount assigned and
released from each crew member). In the second phase, execution of work is done by the initial
crew determined in the formation phase. In addition, data collection of primary and intermediate
crew design indicators is conducted. Then these design indicators are analyzed in the third phase
to refine the initial crew design and maximize the crew performance based on the lean crew
design guidelines proposed in the previous section.

141

Phase I

Phase II

Phase III

Formation

Execution

Refining

Formation of Crew

Structuring of Work
Execution of
Work

Estimation
Amount of
Work/Quantity

Transf
ormati
on

Amount of Work
Assigned to
Each Crew Member

Over production

Waiting

Over processing

Defects

Motion

Flow

Standard Crew
Productivity

Inventory

Transport
Sequence of
Tasks

List of Tasks in an
Activity

Activities Required to
Perform Work

Crew Performance
Outcomes

Achieved by Elimination of all Waste in
the work executed by the crew

Standard Crew Size &
Composition

Activity/Task Duration

Crew Design
Characteristics

Amount of Work
Released from
Each Crew Member

Data
Collection of
Primary and
Intermediate
Crew Design
Characteristics
and Outcomes

Value

Value is delivered by optimizing both
Crew & Work
Flow
Work
Reducing Cost

Motivation

Reducing Time

Learning

Improving Quality

Satisfaction

Scheduling
Figure 4.11: Lean-based crew design (LbCD) process
141

Pride

4.7

Chapter Summary

This chapter proposes lean crew design characteristics and guidelines by synthesizing research
studies on “work design/team design” that have well developed theoretical and empirical
foundations in lean production, lean construction, socio-technical system theory, social and
organizational psychology.
As, discussed earlier, performance of crews is very crucial to the work being executed and
consequently for the overall project performance. Hence, in this chapter all the significant crew
design parameters affecting crew performance are synthesized from literature into lean crew
design characteristics to guide the design of crews and maximize their performance. The crew
design characteristics are classified into two main types: flow and work characteristics that are
further divided into four characteristics such as waste, task, social, and crew comprising of both
the hard and soft crew design factors.
The proposed guidelines lead to the joint optimization of both the crew and the work
performed, resulting in better overall crew performance by optimizing crew performance. The
guidelines are proposed for crew performance outcomes, such as duration of work performed,
crew productivity, worker learning, worker pride etc.

142

5 CASE STUDY - MODEL BUILDING, SIMULATION &
EXPERIMENTATION

5.1

Introduction

In the previous chapter 4, a lean-based crew design model (LbCD) is proposed that integrates
both the existing crew design and work design research providing a crew design framework (that
includes important crew design characteristics) and guidelines for designing and improving the
performance of crews.
The purpose of this section is to develop simulation models in order to demonstrate how the
crew design characteristics affect the performance of the crew. To develop a simulation model
that can effectively identify and demonstrate both the hard and soft crew design characteristics, a
combination of Discrete Event Simulation (DES) and System Dynamics (SD) is used. The
simulation models are developed first using Discrete Event Simulation (DES) and then moving
to System Dynamics (SD) providing the qualitative and quantitative crew design models. The
simulation models are developed for a construction operation involving installation of light
fixtures from a previous research work conducted by Howell et al (1993). The background of the
case study is discussed in detail in the following section.
5.2

Case Study Background

For the purpose of this research, a case study involving a well-documented construction
operation from a prior research work conducted by Howell et al (1993) was selected. The case
study consists of a real time construction operation focusing on the installation of light fixtures
(Figure 5.1) by a construction crew. The initial raw data for this case study was collected by the
143

previous research work by conducting interviews with the construction staff prior to start of the
operation, as described in Howell et al (1993). The operation was monitored carefully to keep a
record of the initial durations, crew make-up, equipment used, design process of operation and
other significant parameters involved during carrying out the operation.

Figure 5.1: Type of light fixture installed (Figure 9 in Howell et al. 1993)
The construction operation involved installation of 314 light fixtures. The initial crew make-up
for the installation operation consisted of 1 journeyman, 1 apprentice, and 1 scissor lift. The
estimated productivity planned by the construction staff to carry out this operation was 2.5
fixtures per crew hour, or a total of 48 worker-minutes per fixture. The total time of operation
was estimated to be 125.6 hours at the rate of 2.5 fixtures per crew hour. The hourly wages of the
crew were $13.40 per hour for 1 journeyman, $6.00 per hour for 1 apprentice, and $1.50 per hour
for the scissor lift.
The original light fixture installation operation was divided into two main cycles: (1)
Assembling Cycle and (2) Installation Cycle. The assembling cycle involved one task of
assemble fixture (9minutes) executed by an apprentice and the installation cycle involved three
tasks of install fixture (7 minutes), wire & clip (8.5 minutes), and lamp & finish (7.5 minutes)
executed by one journeyman. For the purpose of modeling the exact process of installation of
144

light fixtures, the original construction operation was modified from the original two cycles
(assembling and installation) to three cycles (assembling, restocking, and installation) shown in
the Figure below. The restocking cycle included 3 tasks of restocking (1 minute), lift going up
(1minute) and lift going down (1minute).

ASSEMBLING
CYCLE

ASSEMBLE
9 MIN

INSTALLATION
CYCLE

ORIGINAL PROCESS

INSTALL
7 MIN

WIRE&CLIP
8.5 MIN

LAMP&FINISH
7.5 MIN

ASSEMBLING
CYCLE

ASSEMBLE
9 MIN

RESTOCKING
CYCLE

RESTOCKING
1 MIN

LIFT GOING UP
1 MIN

LIFT GOING DOWN
1 MIN

INSTALLATION
CYCLE

MODIFIED PROCESS

INSTALL
7 MIN

WIRE&CLIP
8.5 MIN

LAMP&FINISH
7.5 MIN

Figure 5.2: Original vs. modified process of installation of light fixtures

145

5.3

Discrete Event Simulation Model

As discussed earlier in chapter 3, Discrete Event Simulation (DES) has been widely recognized
as an effective technique to design and analyze construction operations by virtue of its ability to
model repetitive processes. In this research, EZStrobe (Martinez 1996) is used to model the
repetitive process of installation of light fixtures.
Three DES models are developed with different crew designs to test the performance of the crew
for executing this construction operation involving installation of light fixtures. Having the right
crew design is very important, as any change in the design of the crew would also mean a change
in how the operation is executed by them. A poor crew design could lead to an increase in the
duration and cost of the operation, a decrease in the productivity and quality of the work
performed along with lower levels of satisfaction, motivation, learning and pride among the
crew.
5.3.1

DES Base Crew Model

In the “Base” crew case, the amount of work to be performed is the installation of 300 light
fixtures. The size of the crew is 2 workers with crew composition of 1 apprentice, 1 journeyman,
and 1 lift. Activities required to perform the operation of installation of light fixtures are
assemble, restocking, lift going up, install, wire & clip, lamp & finish, and lift going down.
Depending on the standard crew productivity activity duration determined by the construction
staff is shown in the Table 5.1. Once the initial crew is formed, structuring of the work is done;
the apprentice is assigned the responsibility for assembling the fixtures which are then restocked
and installed (install, wire & clip, and lamp & finish) by one journeyman. The journeyman uses
one scissor lift to restock the light fixtures, complete the installation and then again restock. The
146

crew assembles, restocks and install one light fixture at a time. The sequence of activities for this
operation is shown in the Figure 5.3 and the activity, entity, resource, crew make-up and time
data used to develop the base model is shown in the Table 5.1.

ASSEMBLING
CYCLE

RESTOCKING
1 MIN

LIFT GOING UP
1 MIN

WIRE&CLIP
8.5 MIN

INSTALL
7 MIN

RESTOCKING
CYCLE

INSTALLATION
CYCLE

ASSEMBLE
9 MIN

RESTOCKING
CYCLE

BASE CASE - ACTIVTY SEQUENCE

LAMP&FINISH
7.5 MIN

LIFT GOING DOWN
1 MIN

Figure 5.3: Sequence of activities in the “Base” crew case of installation of light fixtures
Table 5.1 shows the activities and their types involved in this process along with the entity and
resource types that are needed to execute each type of activity. The two activity types used to
model this process are combi (conditional) and normal along with queues where entity and
resource types accumulate. As discussed earlier in chapter 3, the Combi activity requires certain
condition set by the draw link to satisfy in order to start and produce an outcome by use of
release link. For example, in this model for activity assemble (combi) to start; the conditions set
by the draw link should be met for both the light fixture and the apprentice. Similarly, for
147

activities restocking, lift going up, install to start, the conditions set by the draw link should be
met for the journeyman, lift, and the light fixture. The Figures 5.4, 5.5 and 5.6 shows the
assembling cycle, restocking cycle and installation cycle of simulation model developed for the
“Base” crew case of installation of light fixtures.
Table 5.1: DES "Base" crew model data
DES “BASE” MODEL DATA
S.NO
.
1
2
3
4
5
6
7

ACTIVITY
Assemble
Restocking
Lift Going

TYPE

Combi
Combi
Combi
Install
Combi
Wire & Clip Normal
Lamp &
Normal
Lift Going
Combi
SIZE/NUMBER

ENTITY
TIME
(min) Light Fixture
9
√
1
√
1
√
7
√
8.5
√
7.5
√
1
X
300

RESOURCE
Apprentice Journeyma
√
X
X
X
X
X
X
1

X

Lift
X

√
√
√
√
√
√
1

√
√
√
√
√
√
1

In this model, initially 300 light fixtures are waiting in the LitFix queue to be assembled. For the
conditional activity assemble to start, both apprentice and light fixtures should be available (i.e.
AprntcWt and LitFix >0); so once both these are available assembling takes place for 9 minutes
and releases one each of apprentice and assembled light fixture that waits in the apprentice wait
queue (AprntcWt) and assembled light fixture queue (AsmbdLitFixA) respectively. For
conditional activity restocking to start, there should be at least one light fixture assembled, one
journeyman and one lift waiting in the AsmbdLitFixA, JrnymnWt, LftDwnWt queue and
restocking count (RestockCont) should be equal to one. Once, the light fixture is restocked by
the journeyman in the lift, all three go up in 1 minute and wait in the AsmbdLtFxLft, LftUpWt
and JrnymnWt queues.

148

2
>0 , 1

LitFix

Assemble

300

1

AsmbdLitFixA
0

9
>0 , 1

1

AprntcWt
1

Figure 5.4: Simulation model (Assembling Cycle) for the “Base” crew case of installation of
light fixtures

AsmbdLitFixA
0

>0 , 1

AsmbdLtFxLf

LftGoinUpWt

t

>0 , 1

LftGoinUp

0

0
1

1
1

1
1

1

JrnymnWt

Restocking
>0 , 1

1

==1 , 1

LftUpWt

1

0

>0 , 1

RestockCont

LftDwnWt

1

1

LftGoinDwn
1

==1 , 1

LftGoinDwnWt
0

1

Figure 5.5: Simulation model (Restocking Cycle) for the “Base” crew case of installation of light
fixture
149

AsmbdLtFxLf
t
0
>0 , 1
LftUpWt
2

>0 , 1

0

Install
7
>0 , 1

1

JrnymnWt

WirnClip

1

8.5
1
1
LampnFnsh

LftGoinDwnWt
0

7.5
1
1

1

RestockCont
1

LitFixInPlc

Figure 5.6: Simulation model (Installation Cycle) for the “Base” crew case of installation of light
fixtures

150

The installation cycle starts with the activity install taking 7 minutes when journeyman, lift and
light fixture are all available (AsmbdLtFxLft, LftUpWt and JrnymnWt >0), followed by wire &
clipping (WirnClip) taking 8.5 minutes and lamp & finish (LampnFnsh) taking 7.5 minutes.
After, lamp & finish activity is finished one light fixture is released to the LitFixInPlc queue and
both journeyman and lift goes down in 1 min to restock the next assembled light fixture. All
three cycles of assembling, restocking and installation repeat until 300 light fixtures are finished
and in place.
Results from the simulation of the “Base” case of installation of light fixtures have been
summarized in Table below.
Table 5.2: Simulation results of the “Base” crew of installation of light fixtures
PERFORMANCE OUTCOMES

BUDGETED

BASE CASE

125.60 hours

130.15 hours

8.32 USD

9.06 USD

Apprentice Utilization

-

34.65%

Journeyman Utilization

-

89.88%

Lift Utilization

-

99.70%

Average Crew Utilization

-

62.26%

Average Apprentice Idle Time

-

16.97 min

Average Journeyman Idle Time

-

1.01 min

Average Work-In-Progress (WIP)

-

97.64 Fixtures

Operation Production Rate (OPR)

-

2.31 Fixtures/Hr

Maximum Production Rate (MPR)

-

6.67 Fixtures/Hr

OPR-MPR Ratio

-

34.63%

Time of Operation
Unit Cost of Installation

The “Base” crew takes 130.15 hours to complete the operation and install 300 light fixtures;
that is 4.55 hours more than that was actually budgeted (125.6 hours) by the construction staff.
151

The unit cost of operation also came out to be $0.72 more than the budgeted unit cost of $8.32.
The resource utilization originally not budgeted by the construction staff came out to be 34.65%
for apprentice, 89.88% for journeyman, and 99.70% for lift; that shows although the journeyman
and the lift were utilized for most of the operation, apprentice was only utilized for around onethird of the operation. This is supplemented by the fact that average idle time for the apprentice
is very high at 16.97 minutes whereas average idle time for the journeyman is only 1.01 min. The
average work-in-progress (WIP) is also high at 97.64 fixtures that means around 97.64 light
fixtures have been assembled but are waiting to be restocked and installed indicating a very low
operation production (2.31 fixtures/hr) to maximum production (6.67 fixture/hr) ratio at only
34.63% (Figure 5.7).
At Normal Capacity

8

6.67

7
Light Fixtures / Hour

At Maximum Capacity
6.67

6
5
4
3
2

2.31

2.31

2.21

2.31

1
0
Operation
Maximum
Assembling Cycle Installation Cycle
Production Rate Production Rate Production Rate Production Rate
(OPR)
(MPR)
(ACPR)
(ICPR)
Figure 5.7: Production rates for the "Base" crew case
The simulation results for the “Base” crew of installation of light fixtures are analyzed to
evaluate the performance of the crew design in terms of the performance outcomes proposed in

152

the previous chapter 4: (1) duration of work performed; (2) cost of work performed; (3)
productivity; (4) quality of work performed; (5) worker satisfaction; (6) worker learning; (7)
worker motivation; (8) pride in work performed.
1. Duration of Work Performed:
Duration of work performed (time of operation) is directly related to the processing time,
non-processing time, and variability in the work performed. The over budget duration
(130.15 hours i.e. 7809 minutes vs. 125.60 hours i.e. 7536 minutes) in this case, is due to:
a. High non-processing time present in the form of high wait/idle time for the apprentice
who is waiting an average of 16.97 minutes. The apprentice assembles 300 fixtures
@9minutes in total of 2700 minutes and then is just idle for the rest of the operation for
approx. 5109 minutes. Also, the non-processing time is increased when the journeyman
has to come down to restock for every light fixture and then go up to install leading to
waste in terms of high transportation/operating/motion time.
b. High over-production present in the form of large intermediate inventory/work-inprogress (AsmbdLitFixA), as production rate for assembling cycle (6.67 fixtures/hour) is
very high compared to the restocking and installation cycle, that means more number of
light fixtures are assembled than that can actually be installed by the journeyman. High
work-in-progress in front of journeyman may also lead to overburdening him, potentially
leading to lowering his task productivity.
c. Non-uniform work load between the assembling and installing cycle, leading to gap in the
cycle times (assembling cycle time – 9 minutes vs. restocking and installation cycle time
– 26 minutes) results in non-continuous flow, where each light fixture waits in the queue

153

before being restocked and installed. It also implies that a push system instead of pull
system is used.
2. Cost of Work Performed:
Cost of work performed is directly related to the crew size, crew flexibility, processing
and non-processing cost and defective products. The high unit cost of operation ($9.06)
in this case is due to:
a. High non-processing cost resulting from the high non-processing time as discussed
earlier.
b. Low crew flexibility is present in this case. As seen from the model results, although
apprentice is idle for two-third (65.35%) of the time of operation, he is not being
utilized to help the journeyman in the installation cycle or involved in multi-tasking
when idle that would lead to shortening the duration of this operation and
correspondingly saving the labor costs.
3. Productivity:
Productivity is related to resource utilization, crew flexibility, overburden, task
experience, task complexity etc. In this case, the crew productivity is low as the crew is
not very flexible; apprentice being utilized very less. The gap between the operation
production rate and maximum production rate (2.31 fixtures/hour vs. 6.67 fixtures/hr) is
very high and any reduction in that gap would definitely lead to a higher overall crew
productivity.
4. Quality of Work Performed:
Quality of work is directly related to the defective products, task experience, task
complexity, overburden, task interdependence. But in this case, these soft or qualitative

154

crew design characteristics cannot be empirically quantified using discrete event
simulation.
5. Worker Learning:
Worker learning is related to crew design indicators such as coordination, crew
flexibility, task variety etc. In this case, these soft or qualitative crew design
characteristics cannot be empirically quantified using discrete event simulation. But from
the model and simulation results, a qualitative inference can be made that there is very
less/no team work/crew flexibility/multi-tasking/coordination/task variety present in this
case. Hence, learning among crew is minimal to non-existent for the apprentice, except in
assembling fixtures.
6. Pride in Work performed:
Pride in the work performed is related to the task identity and self-management. In this
case, these soft or qualitative crew design characteristics cannot be empirically quantified
using discrete event simulation. But from the model and simulation results, a qualitative
inference can be made that only pride that may be present could be in Journeyman as
installation that is the major part of the operation is solely executed by the journeyman
and he can identify with the work performed by him and take pride in it.
7. Worker Satisfaction:
Worker satisfaction directly relates to crew design indicators such as coordination,
support, task significance, reasonable autonomy etc. But in this case, these soft or
qualitative crew design characteristics cannot be empirically quantified using discrete
event simulation.

155

8. Worker Motivation:
Worker motivation directly relates to crew design indicators such as coordination, task
identity, task significance, reasonable autonomy etc. But in this case, these soft or
qualitative crew design characteristics cannot be empirically quantified using discrete
event simulation.
5.3.2

DES Improved Crew Model

In the “Improved” crew case, the same amount of work was performed i.e. the installation of 300
light fixtures. The earlier “Base” crew case was improved and following changes (Howell et al
1993) to the “Base” crew design were introduced:
1. Two journeymen with dedicated lifts were added;
2. A temporary rack developed by the crew held pieces for 16-20 fixtures- minimizing
loading/restocking frequency;
3. One apprentice and one journeyman prepared the initial buffers of assembled fixtures
while the other two journeymen modified the scissor-lifts.
The size of the crew is 4 workers with crew composition of 1 apprentice, 3 journeymen, and
3 lifts. Activities required to perform the operation of installation of light fixtures are same in the
“Improved” case i.e. assemble, restocking, lift going up, install, wire & clip, lamp & finish, and
lift going down; but in this crew design, structuring of work among the crew was different from
the “base” crew design case. Here, the apprentice was again assigned the responsibility for
assembling the fixtures in 9 minutes but, each journeyman executed only one activity in the
restocking and installation cycle (one executed install, the second executed wire & clip, and the
third executed lamp & finish). Also, each of the 3 journeymen would go down using their own
lift once they have completed 15 light fixtures to restock and again execute the same activity for
156

another 15 fixtures. For the purpose of modelling and simulation, it was assumed that restocking
required 3 minutes with lifts taking 1 minute each in going up and coming down. The sequence
of activities for this operation is shown in the Figure 5.8.

LAMP&FINISH WIRE&CLIP INSTALLATIO RESTOCKIN ASSEMBLIN
CYCLE
CYCLE
N CYCLE
G CYCLE
G CYCLE

IMPROVED CASE – ACTIVTY SEQUENCE
ASSEMBLE
9 MIN

RESTOCKING
3 MIN

LIFT GOING
UP
1 MIN

INSTALL
4 MIN

LIFT GOING
DOWN
1 MIN

RESTOCKI
NG
1 MIN

LIFT GOING
UP
1 MIN

WIRE&CLIP
4 MIN

LIFT GOING
DOWN
1 MIN

RESTOCKI
NG
1 MIN

LIFT GOING
UP
1 MIN

LAMP&FINIS
H
4 MIN

LIFT GOING
DOWN
1 MIN

RESTOCKI
NG
1 MIN

LIFT GOING
UP
1 MIN

Figure 5.8: Sequence of activities in the “Improved” crew case of installation of light fixtures
For example, in this model for activity install (combi) to start; the conditions set by the draw
link should be met for the light fixture, journeyman, and lift. Similarly, for activity lift A going
down, the table shows that only the journeyman A and lift A are needed to execute that activity,
as the light fixture is already installed and is not needed in the going back cycle.
Table 5.3 shows the activities and their types involved in this process along with the entity,
resource types, time and size data that are needed to execute each type of activity. The two
157

activity types used to model this process are combi (i.e. conditional activity) and queues where
entity and resource types accumulate.
Table 5.3: DES "Improved" crew model data
DES “IMPROVED” MODEL DATA

S.NO

ACTIVITY

TYPE

TIME
(min)

ENTITY

RESOURCE

Light Fixture

Apprentice

Journeyman

Lift

1

Assemble

Combi

9

√

√

X

X

2

Restocking A

Combi

C

√

X

√

√

3

Lift A Going
Up

Combi

1

√

X

√

√

4

Install

Combi

4

√

X

√

√

6

Lift A Going
Down

Combi

1

X

X

√

√

7

Restocking B

Combi

3

X

X

√

√

8

Lift B Going
Up

Combi

1

X

X

√

√

9

Wire & Clip

Combi

8

√

X

√

√

10

Lift B Going
Down

Combi

1

X

X

√

√

11

Restocking C

Combi

3

X

X

√

√

Combi

1

X

X

√

√

Combi

4

√

X

√

√

Combi

1

X

X

√

√

300

1

3

3

12
13
14

Lift C Going
Up
Lamp &
Finish
Lift C Going
Down

SIZE/NUMBER

158

The assembling cycle for the improved case simulation model is same as the base case. However, the installation cycle for the three
journeymen is different for the “Improved” crew case of installation of light fixtures and is shown in the Figures 5.9, 5.10, 5.11.

>0 , 1

AsmbLtFxLftA

LftAGoinUpWt

0

2

>0 , 1

0
1

15

1
AsmbLitFixA

JrnymnARestck

15

>0 , 1

1

1
1

>0 , 1

1
JrnymnAWt
1
1

>0 , 1

1

4

1

>0 , 1

LftADwnWt

Install

1

3

==15 , 15

1

LftAGoinUp

LftAUpWt

InstldLitFix

>0 , 1

0

0
1

LftAGoinDwn

==15 , 15

1

LftAGoinDwn
Wt
0

Figure 5.9: Simulation model for the installation activity cycle in the “Improved” crew case of installation of light fixtures
159

>0 , 1

AsmbLtFxLftB

2

LftBGoinUpWt >0 , 1
0

0
15

1

>0 , 1
1
LftBDwnWt

JrnymnBRestck

>0 , 1

1

1
1

>0 , 1

1
JrnymnBWt

>0 , 1

1
1
LftBUpWt

WirnClipLitfix

>0 , 1

0

0

1
1

8

1

>0 , 1

LftBGoinDwn

WirenClipLitFix

1

3

1

1

LftBGoinUp

==15 , 15

LftBGoinDwn
Wt
0

InstldLitFix
0

Figure 5.10: Simulation model for the wire and clip activity cycle in the “Improved” crew case of installation of light fixtures
160

>0 , 1

AsmbLtFxLftC

2

LftCGoinUpWt >0 , 1
0

0
15

1

>0 , 1
1
LftCDwnWt

JrnymnCRestck

>0 , 1

1

1
1

>0 , 1

1
JrnymnCWt

>0 , 1

1
1
LftCUpWt

LitFixInPlc

>0 , 1

0

0

1
1

4

1

>0 , 1

LftCGoinDwn

LmpnFns

1

3

1

1

LftCGoinUp

==15 , 15

LftCGoinDwn
Wt
0

WirnClipLitfix
0

Figure 5.11: Simulation model for the lamp & finish activity cycle in the “Improved” crew of installation of light fixtures
161

In this model, initially 15 light fixtures have already been pre-assembled and are waiting in
the assembled light fixture queue (AsmbdLitFixA) and 285 light fixtures are waiting in the light
fixture waiting queue (LitFix) at the start time of simulation. In the install activity cycle,
restocking of light fixture (JrnymnARestck) takes place when all journeyman, lift waiting down
and 15 assembled light fixtures are waiting (i.e. JrnymnARestck==15, JrnymnAWt &
LftADwnWt >0). Once the restocking of light fixtures takes place in 3 minutes, 15 light fixtures
go up with the journeyman and lift in 1 minute and wait in the AsmbLtFxLftA, LftAUpWt, and
JrnymnAWt queues to be installed. Then, the journeyman installs each light fixture in 4 mins
(when

AsmbLtFxLftA,

LftAUpWt,

and

JrnymnAWt

>0)

and

goes

down

(when

LftAGoinDwnWt ==15) to restock again only after 15 light fixtures have been installed.
The wire & clip (WirnClip) and lamp & finish (LmpnFns) activity also have same cycles as the
install cycle except the fact that journeyman A goes down to restock the 15 assembled light
fixtures whereas, journeyman B & C go down the lift to restock the material and tools required to
execute their respective activities. Activity WirnClip starts when installed light fixture
(InstldLitFix >0), material & tools restocked for 15 light fixtures (AsmbLtFxLftB>0), lift B up
waiting (LftBUpWt>0) and journeyman B waiting (JrnymnBWt>0). And activity LmpnFns starts
when wire & clipped light fixture (WirnClipLitfix>0), material & tools restocked for 15 light
fixtures (AsmbLtFxLftC>0), lift C up waiting (LftCUpWt>0) and journeyman C waiting
(JrnymnCWt>0). Note that although Journeyman A can start and continue installing 15 fixtures
continuously, but journeyman B can only wire & clip if a complete light fixture has been
installed by the journeyman A and is ready to be wire & clipped (although he has the material &
tools for 15 light fixtures, there could be wait times if a installed fixture is not ready, same is true
for journeyman C for lamp & finishing activity).

162

Results from the simulation of the “Improved” case of installation of light fixtures have been
summarized in Table below.
Table 5.4: Simulation results for the “Improved” crew of installation of light fixtures
PERFORMANCE OUTCOMES

BUDGETED

IMPROVED CASE

125.60 hours

45.03 hours

8.32 USD

7.61 USD

Apprentice Utilization

-

94.94%

Journeyman A Utilization

-

46.63%

Lift A Utilization

-

48.11%

Journeyman B Utilization

-

91.15%

Lift B Utilization

-

92.67%

Journeyman C Utilization

-

46.74%

Lift C Utilization

-

48.26%

Average labor Utilization

-

66.43%

Average Apprentice Idle Time

-

0.48

Average Journeyman A Idle Time

-

4.49 min

Time of Operation
Unit Cost of Installation

Average Journeyman B Idle Time

0.00 min

Average Journeyman C Idle Time

4.47 min

Average Work-In-Progress (WIP)

-

28.29 Fixtures

Operation Production Rate (OPR)

-

6.66 Fixtures/Hr

Maximum Production Rate (MPR)

-

6.67 Fixtures/Hr

OPR-MPR Ratio

-

99.85%

The “Improved” crew takes 45.03 hours to complete the operation and install 300 light
fixtures. The time of operation here is 85.12 hours less than the “Base” crew case (130.15), and
80.57 hours less than that was actually budgeted (125.6 hours) by the construction staff. The unit
cost of operation also came out to be $0.71 less than the budgeted unit cost of $8.32 and $1.45
163

less than the “Base” case. There is no significant savings in the cost as compared to the time of
operation that can be attributed to the large crew size of 4 workers and 3 lifts. The resource
utilization originally not budgeted by the construction staff came out to be 94.94% for
apprentice, 46.63%, 91.15%, and 46.74% for journeyman A, B & C and 48.11%, 92.67%, and
48.26% for lift A, B, & C. Also, the average labor utilization in this was slightly more at 66.43%
than the “Base” case (62.26%). These above utilizations clearly show that though the apprentice
and journeyman B were almost more than 90% utilized, journeyman A & C were only less than
50% utilized and were waiting for more than 50% of the operation time. This is supplemented by
the fact that Average Idle Time for both apprentice and journeyman B is almost zero or less than
half a minute whereas the Average Idle Time for both the journeyman A & C is around 4.49 and
4.47 minutes respectively.
The average work-in-progress (WIP) was reduced from 97.64 fixtures to 28.29 fixtures,
although it was not a single piece flow and a batch and single piece mixed flow. The reduced
WIP also supplements the better operation production rates for all the activities and for the
operation at 6.66 light fixtures/hour compared to the 2.31 light fixtures/hour in the “Base” case.
The individual production rates for all the activities at their normal and maximum capacity for
this “improved” crew case has been shown in the figure 5.12.

164

At Normal Capacity

16

At Maximum Capacity

15

15

14

12.79

12.19

Light Fixtures / Hour

12
10
8
6

8.31
6.66

6.67

6.33

6.67

7.067

7.5
6.41
4.67

4
2
0
Operation Maximum Assembling
Install
Install
Wire & Clip Wire & Clip
Lamp &
Lamp &
Production Production Production Production Restocking Production Restocking
Finish
Finish
Rate (OPR) Rate (MPR)
Rate
Rate
Production
Rate
Production Production Restocking
Rate
Rate
Rate
Production
Rate
Figure 5.12: Production rates for the "Improved" crew case

165

The simulation results for the “Improved” crew of installation of light fixtures are analyzed
to evaluate the performance of the crew design in terms of the performance outcomes proposed
in the previous chapter 4: (1) duration of work performed; (2) cost of work performed; (3)
productivity; (4) quality of work performed; (5) worker satisfaction; (6) worker learning; (7)
worker motivation; (8) pride in work performed.
(1) Duration of Work Performed:
The “Improved” case shows significant savings in the hours of operation. The reduced
duration of work (45.03 hours i.e. 2701.8 minutes in the “Improved” case vs. 130.15 hours
i.e. 7809 minutes in the “base” case) is due to:
a. Reduction in the processing timeIn the “Base” crew case, the install, wire & clip and lamp & finish activities were
talking 7, 8.5 and 7.5 minutes respectively whereas in this case after the crew was
redesigned, the install, wire & clip and lamp & finish activities took 4, 8, and 4
minutes respectively to execute the installation cycle. This reduction in the activity
durations considerably reduced the installation cycle time and consequently the
duration of operation.
b. Reduction in non-processing timei.

Reduction in motion and transportation of light fixturesIn the “Base” case, each light fixture took 1 minute to restock, 1 minute to go
up and 1 minute for the lift to come down to restock the next fixture; making
it a total of 3 minutes each for 300 light fixtures requiring 300 moves to
restock all the fixtures. Hence a total of 900 minutes (300 moves * 3 minutes
each) were spent in motion and transportation of 300 light fixtures.

166

Whereas, in this “Improved” case, restocking was done by 3 journeymen for a
batch size of 15 light fixtures; 3 minutes to restock, 1 minute to go up and 1
minute for the lift to come down to restock the next fixture; making it a total
of 5 minutes each for 15 light fixtures requiring 20 moves to restock all the
300 fixtures. Hence a total of 300 minutes ((3*300*({3+1+1}/15)) were spent
in motion and transportation of 300 light fixtures.
ii.

Reduction in the wait/idle timeThe wait time for both apprentice and the journeyman has been considerably
reduced from the “base” case. The apprentice average wait time has been
reduced from 16.97 minutes to only 0.48 minutes. And, the journeyman
average wait time in the installation cycle earlier which was 1.01 minutes has
been increased to 4.49 and 4.47 minutes for journeyman A and Journeyman C
and reduced to 0.00 minutes for journeyman B (if considered for 15 light
fixtures together).

c. Over-production is still present in this case present in the form of intermediate
inventory/work-in-progress (InstldLitFix). This evident from the fact that install
activity production rate (7.067 fixtures/hour) is high compared to the wire & clip
activity production rate (6.41 fixtures/hour).

That means more number of light

fixtures are installed by the journeyman A than that can actually be wire & Clipped
by the journeyman B.
However, compared to the “Base” case, the work-in-progress in the “Improved” case
has been reduced considerably from 97.64 fixtures to 0.53 fixtures. This low WIP
leads to less variability resulting in reducing the duration of the work performed.

167

However, the WIP in front of journeyman B may lead to overburdening him,
potentially leading to lowering his task productivity.
d. There is non-uniform work load present between the assembling, install, wire & Clip,
and lamp & finish activity cycle, leading to gap in the cycle times that results in noncontinuous flow.
For example, considering the batch size of 15 fixtures: assembling cycle time is 135
minutes vs. install activity cycle time is 65 minutes), where journeyman A after
installing 15 light fixtures waits in the queue for around 70 minutes every time to start
restocking the next 15 light fixtures that take 135 minutes to be assembled.
(2) Cost of Work Performed:
Cost of work performed is directly related to the crew size, crew flexibility, processing and
non-processing cost and defective products. The reduced unit cost of operation ($7.61) in this
case is due to:
a. Reduced non-processing cost resulting from reduced non-processing time in the
operation as discussed earlier.
b. Reduced processing cost resulting from reduced processing time in the operation as
discussed earlier.
c. There is increase in the size of the crew from 2 workers and one lift to 4 workers and
3 lifts. Due to this, labor and equipment cost per hour definitely increased from
$20.9/hour to $50.7/hour but in this “Improved” crew case due to the significant
reduction in the duration of the operation, the total cost of operations is also reduced.

168

(3) Productivity:
Productivity is related to resource utilization, crew flexibility, overburden, task experience,
task complexity etc. In this “Improved” case, the crew productivity is better than the “Base”
case, as is evident from the increase in the operation production rate from 2.31 light
fixtures/hour (base case) to 6.67 light fixtures/hour (improved case). Also as discussed
earlier, an increase in the resource utilizations of overall crew (average system increase rather
than single point increase) would increase the overall crew productivity; this is evident from
the fact that apprentice utilization has increased from 34.65% to 94.94% along with
journeyman B utilization has increased to above 90%.
(4) Quality of Work Performed:
Quality of work is directly related to the defective products, task experience, task complexity,
overburden, task interdependence. But in this case, these soft or qualitative crew design
characteristics cannot be empirically quantified using discrete event simulation.
(5) Worker Learning:
Worker learning is related to crew design indicators such as coordination, crew flexibility,
task variety etc. In this case, from the model and simulation results, a qualitative inference
can

be

made

that

there

is

very

less/no

team

work/crew

flexibility/multi-

tasking/coordination/task variety present in this case similar to the “Base” crew case. Hence,
learning among crew is minimal to non-existent for the apprentice, except in assembling
fixtures.
(6) Pride in Work performed:
Pride in the work performed is related to the task identity and self-management. In this case,
from the model and simulation results, a qualitative inference can be made that pride is now

169

lost to everyone involved as the work performed has been divided into sub-activity cycles
that are performed by different workers and no crew member can now identify himself with
the whole operation and take pride in it, because they only execute a part of the operation.
(7) Worker Satisfaction:
Worker satisfaction directly relates to crew design indicators such as coordination, support,
task significance, reasonable autonomy etc. But in this case, these soft or qualitative crew
design characteristics cannot be empirically quantified using discrete event simulation.
(8) Worker Motivation:
Worker motivation directly relates to crew design indicators such as coordination, task
identity, task significance, reasonable autonomy etc. But in this case, these soft or qualitative
crew design characteristics cannot be empirically quantified using discrete event simulation.
5.3.3

DES Lean Crew Model

After evaluating the previous 2 crew cases (base and improved), the “Improved” crew design is
enhanced & optimized based on the lean based crew design guidelines proposed in the chapter 4.
For the “Lean” crew case, the size of the crew selected is 3 workers with crew composition of 2
apprentice, 1 journeyman, and 2 lifts. Activities required to perform the operation of installation
of light fixtures were kept the same in the “Improved” case i.e. assemble, restocking, lift A going
up, install, wire & clip, lamp & finish, and lift A going down; but in this “Lean” crew design,
structuring of work among the crew was different from the “Base” and “Improved” crew design
case. Here, the apprentice A was again assigned the responsibility for assembling the fixtures in
9 minutes, but he was also assigned restocking of the assembled light fixtures and taking them up
in the lift A and coming down in the lift A after handing over the assembled light fixtures to the
journeyman A and apprentice B who are on the other lift waiting above to carry out the
170

installation cycle consisting of install, wire & clip and lamp & finish activities taking 3.5, 4.25,
and 3.75 minutes respectively. The journeyman A and apprentice B are working together on the
same lift B and stay up until they complete 300 light fixtures. In this “Lean” crew case, single
piece flow was adopted and only light fixture was assembled, restocked and installed at a time.
Also, for the purpose of modelling and simulation, it was assumed that restocking required 1
minute with lift A taking 1 minute each in going up and coming down. The sequence of activities
for this operation is shown in the Figure 5.13.

ASSEMBLING
CYCLE

RESTOCKING
1 MIN

INSTALLATION
CYCLE

ASSEMBLE
9 MIN

RESTOCKING
CYCLE

LEAN CASE – ACTIVTY SEQUENCE

LIFT GOING UP
1 MIN

LIFT GOING DOWN
1 MIN

WIRE&CLIP
4.25 MIN

INSTALL
3.5 MIN

LAMP&FINISH
3.75 MIN

Figure 5.13: Sequence of activities in the “Lean” crew case of installation of light fixtures

Table 5.5 shows the activities and their types involved in this process along with the entity,
resource types, time and size data that are needed to execute each type of activity. The two
activity types used to model this process are combi (conditional) and normal along with queues
where entity and resource types accumulate For example, in this model for activity restocking
(combi) to start; the conditions set by the draw link should be met for the light fixture, apprentice
171

A and lift A. Similarly, for activity install, the table shows that both the journeyman A,
apprentice B, and lift B are needed to execute that activity.
Table 5.5: DES "Lean" crew model data
DES “LEAN” MODEL DATA
ENTITY
S.NO

ACTIVITY

TYPE

TIME
(min)

Light
Fixture

RESOURCE
Apprentice

Journeyman

Lift

A

B

A

A

B

1

Assemble

Combi

9

√

√

X

X

√

X

2

Restocking

Combi

1

√

√

X

X

√

X

Combi

1

√

√

X

X

√

X

Combi

1

X

√

X

X

√

X

3
4

Lift A Going
Up
Lift A Going
Down

6

Install

Combi

3.5

X

X

√

√

X

√

7

Wire & Clip

Combi

4.25

X

X

√

√

X

√

8

Lamp &
Finish

Combi

3.75

X

X

√

√

X

√

300

1

1

1

1

1

SIZE/NUMBER

The Figures 5.14 & 5.15 shows the simulation model developed for the “Lean” crew case of
installation of light fixtures. In this model, initially 300 light fixtures are waiting in the LitFix
queue to be assembled. For the conditional activity assemble to start, both apprentice and light
fixtures should be available (i.e. AprntcAWt and LitFix >0); so once both these are available
assembling takes place for 9 minutes and releases one each of apprentice and assembled light
fixture that waits in the apprentice wait queue (AprntcAWt) and assembled light fixture queue
(RestockCount) respectively. For conditional activity restocking to start, there should be at least
one light fixture assembled, one apprentice, one lift A waiting in the RestockCount, AprntcAWt,
LftADwnWt queue should be equal to one.
172

1

1

LftAGoinUpWt

>0 , 1

LftAGoinUp
1

0

Restocking
1

1

1

>0 , 1

LftADwnWt

==1 , 1
>0 , 1

>0 , 1

0

1

RestockCount

AprntcAWt

0

1
>0 , 1

1

AsmbdLtFxLftB
LftAGoinDwn

0

1

1

1

2
1

1
LftAUpWt

Assemble
>0 , 1

LitFix

9

LftADwnCount

300

1
>0 , 1

Figure 5.14: Simulation model for the apprentice, assembling & restocking activity cycle in the “Lean” crew of installation of light
fixtures
173

Install
>0 , 1

AsmbdLtFx
LftB

>0 , 1

3.5
>0 , 1

>0 , 1

1

JrnymnAWt

AprntcBWt

1

1

LftBWt

WirenClip

1

4.25

0

1
1

LampnFinish

LitFixInPlc

1

3.75
1

Figure 5.15: Simulation model for the journeyman, installation cycle in the “Lean” crew of installation of light fixtures

174

Once, the light fixture is restocked by the apprentice A, he goes up and comes down in the
lift A in 1 minute each (LftAGoinUp & LftAGoinDwn), releasing one assembled light fixture in
AsmbldLtFxLftB queue. The installation cycle starts with the activity install taking 3.5 minutes
when journeyman A, lift B, apprentice B and assembled light fixture are all available
(JrnymnAWt, LftBWt, AprntcBWt, AsmbldLtFxLftB >0), followed by wire & clipping
(WirenClip) taking 4.25 minutes and lamp & finish (LampnFnsh) taking 3.75 minutes. After,
lamp & finish activity is finished one light fixture is released to the LitFixInPlc queue.
Results from the simulation of the “Lean” crew case of installation of light fixtures have been
summarized in Table 5.6. The “Lean” crew takes 60.17 hours to complete the operation and
install 300 light fixtures. The time of operation here is 15.14 hours more than the “Improved”
crew case (45.03 hours), but is 69.98 hours less than the “Base” crew case (130.15 hours), and
80.57 hours less than the budgeted (125.6 hours) showing significant saving in the duration of
the operation. The unit cost of operation also came out to be $2.63 less than the budgeted and
$3.37 less than the “Base” case and &1.92 less than the improved case.
The resource utilizations in this case are very high compared to all the previous cases at around
95% apart from lift A utilization that came out to be around 25%. Also, the average labor
utilization in this crew case was considerably more at 94.43% than the “Base” case (62.26%).
The high resource utilizations are supplemented by the fact that Average Idle Times for both
apprentice A and journeyman A (&apprentice B) are very low at 1.02 and 0.53 minute
respectively.

175

Table 5.6: Simulation results of the “Lean” crew of installation of light fixtures
PERFORMANCE OUTCOMES

BUDGETED

LEAN CASE

125.60 hours

60.17 hours

Unit Cost of Installation

8.32 USD

5.69 USD

Apprentice A Utilization

-

92.19%

Apprentice B Utilization

-

95.55%

Journeyman Utilization

-

95.55%

Lift A Utilization

-

24.92%

Lift B Utilization

-

95.56%

Average Apprentice A Idle Time

-

1.02 min

Average Journeyman A Idle Time

-

0.53 min

Time of Operation

Average Apprentice B Idle Time

0.53 min

Average Work-In-Progress (WIP)

-

0.0 Fixtures

Operation Production Rate (OPR)

-

4.98 Fixtures/Hr

Maximum Production Rate (MPR)

-

5.00 Fixtures/Hr

OPR-MPR Ratio

-

99.60%

The average work-in-progress (WIP) was reduced completely from 97.64 fixtures (base case) to
0.0 fixtures (lean case) following a single piece flow by reducing the batch size to light fixture.
The reduced WIP also supplements the better operation production rates for all the activities and
for the operation at 4.98 light fixtures/hour compared to the 2.31 light fixtures/hour in the “Base”
case. The operation production rate is very close to the maximum production rate (5 light
fixtures/hour) that can be possible for this operation at maximum capacity. Also, the individual
production rates for all the activities at their normal and maximum capacity for this “Lean” crew
case has been shown in the figure 5.16.

176

At Normal Capacity

At Maximum Capacity

6

Light Fixtures / Hour

5

4.98

5

5.21

5
4.59

4.58

4
3
2
1
0
Operation Production Maximum Production Assembling Cycle
Rate (OPR)
Rate (MPR)
Production Rate

Installation Cycle
Production Rate

Figure 5.16: Production rates for the "Lean" crew case
The simulation results for the “Lean” crew of installation of light fixtures are analyzed to
evaluate the performance of the crew design in terms of the performance outcomes proposed in
the previous chapter 4: (1) duration of work performed; (2) cost of work performed; (3)
productivity; (4) quality of work performed; (5) worker satisfaction; (6) worker learning; (7)
worker motivation; (8) pride in work performed.
1. Duration of Work Performed:
The “Lean” crew case shows significant savings in the hours of operation. The reduced duration
of work (60.17 hours i.e. 3610.20 minutes) is due to:
a. Reduction in the processing timeIn this case, after the crew was redesigned, the install, wire & clip and lamp & finish
activities took 3.5, 4.25, 3.75 minutes respectively to execute the installation cycle that is

177

lower than compared to both the previous crew cases This reduction in the activity durations
considerably reduced the installation cycle time and consequently the duration of operation.
b. Reduction in non-processing timei.

Reduction in motion and transportation of light fixturesIn this case, restocking was done for each light fixture by the apprentice A rather than
the journeyman. Hence, motion and transportation was reduced for the journeyman
that earlier had to go up and down in the lift for restocking. The restocking time that
is added on to the apprentice A further helps in balancing the cycle times and work
flow among the assembling and installation cycles.

ii.

Reduction in the wait/idle timeThe wait time for both apprentice and the journeyman has been considerably reduced
from the previous cases. The apprentice A average wait time has been reduced to only
1.02 minutes. And, the journeyman A (apprentice B) average wait time in the
installation cycle 0.53 minutes. This significant reduction in the wait time results in
reduced duration of the work.

c. The problem of Over-production is controlled in this case that was earlier present in the in
the form of intermediate inventory/work-in-progress. This evident from the fact that in this
case, there is almost no WIP and also, the assembling production rate (4.59 light
fixtures/hour) is almost equal to the installation activity production rate (4.58 light
fixtures/hour), meaning the number of light fixtures assembled by the apprentice A can now
be installed at the almost the same rate without any wait times in between.
d. Uniform work load is finally achieved in this case between the assembling and installation
cycle leading to continuous flow. This is evident from the fact the assembling and installation

178

cycle times are almost equal (12 minutes vs. 11.5 minutes). Hence now the light fixture don’t
have to wait in the queue after being assembled and can straight away start getting installed.
2. Cost of Work Performed:
Cost of work performed is directly related to the crew size, crew flexibility, processing and nonprocessing cost and defective products. The reduced unit cost of operation ($5.69) in this case is
due to:
a. Reduced non-processing cost resulting from reduced non-processing time in the operation as
discussed earlier.
b. Reduced processing cost resulting from reduced processing time in the operation as
discussed earlier.
c. Reduction in the size of the crew from the improved crew case. There is decrease in the size
of the crew from 4 workers and 3 lifts to 3 workers and 2 lifts. Due to this, labor and
equipment cost per hour reduces from $50.7/hour in the “Improved” crew case to $26.9/hour
in the lean crew case.
3. Productivity:
Productivity is related to resource utilization, crew flexibility, overburden, task experience, task
complexity etc. In this “Lean” case, the crew productivity is better than the “Base” case, as is
evident from the increase in the operation production rate from 2.31 light fixtures/hour (Base
Case) to 4.98 light fixtures/hour. Also as discussed earlier, an increase in the resource utilizations
of overall crew (average system increase rather than single point increase) would increase the
overall crew productivity; this is evident from the fact that apprentice A utilization has increased
from 34.65% to 92.19% and with journeyman B (apprentice B) utilization has increased to above
95.55%.

179

4. Quality of Work Performed:
Quality of work is directly related to the defective products, task experience, task complexity,
overburden, task interdependence. But in this case, these soft or qualitative crew design
characteristics cannot be empirically quantified using discrete event simulation.
5. Worker Learning:
Worker learning is related to crew design indicators such as coordination, crew flexibility, task
variety etc. In this case, from the model and simulation results, a qualitative inference can be
made that there is team work and learning in the crew because the apprentice is working along
with the journeyman helping him to install the light fixtures and also learn on the way.
6. Pride in Work performed:
Pride in the work performed is related to the task identity and self-management. In this case,
from the model and simulation results, a qualitative inference can be made that pride is taken by
the crew as both the journeyman and the apprentice now work on the major part i.e. installation
of light fixtures and can identify themselves with contributing towards the operation and taking
pride in the finished product.
7. Worker Satisfaction:
Worker satisfaction directly relates to crew design indicators such as coordination, support, task
significance, reasonable autonomy etc. But in this case, these soft or qualitative crew design
characteristics cannot be empirically quantified using discrete event simulation.
8. Worker Motivation:
Worker motivation directly relates to crew design indicators such as coordination, task identity,
task significance, reasonable autonomy etc. But in this case, these soft or qualitative crew design
characteristics cannot be empirically quantified using discrete event simulation.

180

5.4

System Dynamics Model

As, shown in the previous sections, DES is very effective in modeling and evaluating the
construction processes at an operational level. The use of DES allowed us to evaluate three
different crew designs (base, improved and lean crew design) for the construction operation of
installation of light fixtures and determine information about expected cost, duration, and other
process level details such as resource utilizations, production rates, idleness of resources,
waiting/delays, work-in-progress etc.
However in DES, it is difficult to model the qualitative or softer aspects such pride, learning,
satisfaction etc. and to evaluate the various nonlinear cause-and-effect relationships and feedback
that exist in a construction process. These cause-and-effect relationships and feedback among the
variables/aspects can be analyzed more effectively with the help of system dynamics (SD).
In this section, SD models (qualitative and quantitative) are developed to represent the 3
crew designs for the operation of installation of light fixtures. The SD models are developed to
demonstrate how both the quantitative (hard factors/aspects) and qualitative (soft factors/aspects)
can be modeled using SD. The proposed qualitative SD model demonstrates the various causeand-effect relationships and feedback among the crew design characteristics and the crew
performance outcomes proposed in the chapter 4. And, the quantitative SD models demonstrate
how the operational/process level details can be modeled using stock and flows in the form of
mathematical equations representing the rate of change of the crew design variables over the
time. The quantitative SD models help in determining and evaluating the pattern of behavior of
various crew design characteristics and crew performance outcomes that is not possible with the
use of DES. For modeling the SD qualitative and quantitative models Vensim PLE, a free
shareware is used.

181

5.4.1

SD Qualitative Model

In the chapter 4, crew design characteristics and crew performance outcomes are proposed to
guide the design of a crew. The crew design characteristics are divided into a number of primary
and intermediate characteristics that have various causal and feedback relationships among them
affecting the performance of the crew in terms of cost, duration, productivity, learning, pride etc.
The causal relationships among the crew design characteristics, intermediate crew design
characteristics were shown in the previous chapter 4 (Figure 4.4 and 4.5); however the feedback
among these was not shown.
Hence, a SD qualitative model is developed and proposed to demonstrate the causal and
feedback relationships among the primary and intermediate crew design characteristics affecting
the performance of the crew. First, qualitative models are developed for each of the crew
performance outcomes such as cost, duration, productivity, learning, pride etc. proposed in the
chapter 4 and then these individual qualitative models are combined together to provide a crew
performance model that shows all the causal and feedback relationships present among the
primary crew design characteristics, intermediate crew design characteristics & outcomes and
final crew performance outcomes.
For example in a SD qualitative duration model, an increase in the processing time will
increase the cycle time that in result will again increase the duration of work performed. And in
the same way, any increase in the movement speed will decrease the processing time resulting in
a decrease in the cycle time and consequently reducing the duration of work performed.
Similarly, other models such as cost, productivity, quality, satisfaction, learning, motivation,
pride are developed and combined together with the feedback relationships in order to develop

182

the qualitative crew performance model that can be accessed at the following website:
https://docs.google.com/file/d/0BxrMYHbElFuNam5keHp1RWVnQTQ/edit
5.4.2

SD Quantitative Models

In this section, SD quantitative models are developed to represent the 3 crew designs for the
operation of installation of light fixtures. The SD quantitative models are developed to
demonstrate how both the causal/feedback relationships and operational level detail of the
construction process of installation of light fixtures can be modeled using continuous simulation
such as system dynamics. The accuracy and reliability of SD models are tested by comparing the
SD simulation results with the DES results.
In system dynamics quantitative modeling, the causal loop/feedback diagrams developed in
the previous section are converted into stock and flow diagrams to represent the dynamic
behavior of a system with the help of mathematical equations.
To model a process in SD, the DES modeling elements such as activity, entity and resources
can be represented in the form of stock and flows. The two types of activities (normal & combi)
and queues are represented by flows and stocks respectively. One of the main differences in
modeling a process in SD verses DES is the representation of the activities, entity and resources.
In DES (STROBOSCOPE), a resource or entity can be distinguished from the other with the help
of different arrows and therefore different type and amount of resources/entities can either enter
or released from the same activity. In SD (Vensim), same activity from DES has to be
represented by different flows for each of the different stocks (represented by entities/resources
in DES). Hence, each cycle in DES would be represented by a number of cycles in SD (equal to
the number of entity/resources); this is discussed in detail in the following section of SD
quantitative models for the 3 crew design cases.

183

As, discussed in the previous section, in the “Base” crew design for the construction
operation of installation of light fixtures, 300 light fixtures are assembled and installed by 1
apprentice, 1 journeyman, and 1 lift. The entity in this case is the light fixture and the resources
are apprentice, journeyman and lift. Now, since there are four different entity and resources, four
different cycles of stock and flow diagram is needed to represent the life cycle of each one of
them from. Also, the sequence of activities is assembling, restocking, lift going up, install, wire
& clip, lamp & finish, and lift going down taking the same time as mentioned in the DES “Base”
crew model. Representation of assembling cycle in SD (Vensim) with conversion of activities
and queues (in DES) into stock and flows is shown in the Figure 5.17.

Apprentice
Returning

Apprentice Waiting
Apprentice
Assembling

Lit Fix Waiting

Apprentice Finished
Assembling

Lit Fix Assembled
Lit Fix
Assembling Rate

Figure 5.17: Representation of assembling cycle for DES VS. SD “Base” case
In the above Figure 5.17, the activity assembling (DES) is represented by two flows (Lit Fix
Assembling Rate and Apprentice Assembling) one for the entity (light fixture) and one for the
184

resource (apprentice) respectively. The accumulation of the entity in DES in shown by the
Queues LitFix where light fixtures are waiting; and after each light fixture is assembled in 9
minutes in activity Assemble, it waits in the AsmbdLitFixA queue. This is represented in SD by
the rate Lit Fix Assembling Rate that draws out 1/9 light fixture each minute from the stock Lit
Fix Waiting into the Lit Fix Assembled stock (hence in 9 minutes one light fixture is assembled).
Similarly Apprentice is drawn out of the Apprentice Waiting stock at 1/9 Apprentice each minute
into the Apprentice Finished Assembling (hence 1 Apprentice finishes assembling in 9 minutes)
and Apprentice is finished assembling in 9 minutes, he returns to waiting again through
Apprentice Returning flow into the Apprentice Waiting stock. The Figure 5.18 shows the pattern
of behavior of the apprentice in the assembling cycle for the first 50 minutes of assembling
cycle.
Apprentice Finished Assembling
Apprentice Finished Assembling

Apprentice Waiting
Apprentice Waiting
11

11

.75
0.75

.75
0.75

0.5
.50

0.5
.50

0.25
.25

0.25
.25

00

00
0

5

10

15

20 25 30
20 25 30
Time (Minute)

35

40

45

50

0 5 10 15
35 40 45 50
Time (Minutes)
Apprentice Waiting : Current
Apprentice Waiting: Current

0
5
10 15 20 25 30 35 40 45 50
0 5 10 15 20 25 30 35 40 45 50
Time (Minute)
Time (Minutes)
Apprentice Finished Assembling : Current
Apprentice Finished Assembling: Current

Figure 5.18: Pattern of behavior for the apprentice in assembling cycle for “base” crew case in
SD (Part A).

185

Apprentice Waiting
Apprentice Assembling

Apprentice Returning
Apprentice Returning

0.2
0.2

11

0.15
0.15

.75
0.75

0.1
0.1

.500.5

0.05
0.05

0.25
.25

00

00
0

5

10

15

20
25
30
20Time (Minute)
25 30

35

40

45

0

50

0 5

5

10

15

20

25

30

35

40

45

50

10 15 20 25 30 35 40 45 50
Time (Minute)
Time (Minutes)
Apprentice Returning : Current
Apprentice Finished Assembling: Current

0 5 10 15
35 40 45 50
Time (Minutes)
Apprentice Assembling : Current
Apprentice Waiting: Current

Figure 5.19: Pattern of behavior for the apprentice in assembling cycle for “base” crew case in
SD (Part B)
The above Figure 5.19 shows behavior of the apprentice changing over time. The apprentice
assembling remains constant at 0.1111 (1/9 apprentice) for every 9 minutes showing that
apprentice is busy assembling for 9 minutes; also apprentice waiting value changes between 1
and 0 respectively every 9 minutes showing the apprentice idle and busy states (apprentice
waiting==1 shows the idle state whereas apprentice waiting<1 shows the assembling state).
Similarly, the light fixture also behaves in the same pattern (Figure 5.20).
Light Fixture Assembling Rate
Light Fixture Assembling Rate

Light Fixture Waiting
Light Fixture Waiting
340
340

0.2
.02

335
335

.15
0.15

330
330

.10
0.1

325
325

0.05
.05

00

320
325
0

5

10

15

20
25
30
20 25 30
Time (Minute)

35

40

45

50

0 5 10 15
35 40 45 50
Lit Fix Waiting : Current
Time (Minutes)
LiTFix Waiting: Current

0

5

0 5

10

15

20

25

30

35

40

45

50

10 15 20 25 30 35 40 45 50
Time (Minute)
Time (Minutes)
Lit Fix Assembling Rate : Current
LiTFix Assembling Rate: Current

Figure 5.20: Pattern of behavior for the light fixture in assembling cycle for “base” crew case in
SD (Part A)
186

The above discussion proves that assembling cycle in SD has been built right and is a correct
representation of the assembling process developed in DES (STROBOSCOPE).
Light Fixture Restocked
Light Fixture Restocked

Light FixtureAssembled
Light Fixture Assembled
4
4

11

3
3

0.75
.75

2
2

0.5
.50

1
1

0.25
.25

00

00
0

5

10

15

20
25
20Time (Minute)30
25 30

35

40

45

50

0 5 10 15
35 40 45 50
Time (Minutes)
Lit Fix Assembled : Current
LiTFix Assembled: Current

0

5
0 5

10
15
25
10 15 20Time (Minute)30 35 40 45 50
20 25 30 35 40 45 50
Lit Fix Restocked : Current Time (Minutes)
LiTFix Restocked: Current

Figure 5.21: Pattern of behavior for the light fixture in assembling cycle for “base” crew case in
SD (Part B)
The restocking and installation cycle for the “Base” crew case in DES (Figure 5.4) is converted
into the SD using stock and flows. Three different flow cycles (light fixture cycle, journeyman
cycle and lift cycle) were developed to represent the construction operation of installation of
light fixtures in the quantitative SD model. The complete quantitative SD model for the “Base”
crew case is shown in the Figure 5.22. The mathematical equations used to develop the model
and graphs for pattern of behavior for the quantitative SD model for “Base” crew case are
provided in the appendices. After modeling the quantitative SD model for the “Base” crew case
of installation of light fixtures; the “Improved” and “Lean” crew cases are also converted into
quantitative SD models. The complete quantitative SD model for the “Base” crew case is shown
in the Figure 5.23. The mathematical equations used to develop the model and graphs for pattern
of behavior for the quantitative SD model for “Base” crew case are provided in the appendices.

187

Apprentice
Returning

Apprentice Waiting
Apprentice
Assembling
Lit Fix Waiting
Lit Fix
Assembling Rate

Lift Waiting

Lit Fix Assembled

Lift Restocking

<Lit Fix
Assembled>

Lift Finished Restocking

Lit Fix
Restocking Rate
Lit Fix Restocked
Lit Fix Going Up
Rate

Lift Going Up

Lift Up

Lit Fix Up

Lift Starting
Installation

Partial Lit Fix
Installation Rate

<Lift Finished Full
Lamp&Finish>
Lift Finished Partial
Installation

Lift Going Down

Partial Lit Fix Installed

Lift Completing
Installation

Lift Finished Full
Installation

Complete Lit Fix
Installation Rate
Full Lit Fix Installed

Apprentice Finished
Assembling

Lit Fix Assembled

Journeyman Waiting
Journeyman
Restocking
Journeyman Finished
Restocking

<Lit Fix
Assembled>

Journeyman
Going Up
Journeyman Up
Journeyman Starting
Installation
<Journeyman Finished
Full Lamp&Finish>
Journeyman Finished
Partial Installation
Journeyman Going
Down
Journeyman
Completing Installation
<Lit Fix In Place>
Journeyman Finished
Full Installation

Figure 5.22: Representation of the quantitative SD model for “Base” crew case (Part A)
188

p
Lift Finished Partial
Installation

Lift Going Down

Journeyman Finished
Partial Installation
Journeyman Going
Down

Partial Lit Fix Installed

Lift Completing
Installation

Journeyman
Completing Installation

Complete Lit Fix
Installation Rate

<Lit Fix In Place>
Lift Finished Full
Installation

<Lit Fix In Place>

Lift Starting
Wire&Clip

Lift Finished Partial
Wire&Clip
Lift Completing
Wire&Clip

Lift Finished Full
Wire&Clip
Lift Starting
Lamp&Finish

Lift Finished Partial
Lamp&Finish
Lift Completing
Lamp&Finish

Lift Finished Full
Lamp&Finish

Journeyman Finished
Full Installation

Full Lit Fix Installed

Journeyman Starting
Wire&Clip

Partial Lit Fix
Wire&Clip Rate

Journeyman Finished
Partial Wire&Clip

Partial Lit Fix
Wire&Clipped

Journeyman
Completing Wire&Clip

Complete Lit Fix
Wire&Clip Rate

Full Lit Fix
Wire&Clipped

Journeyman Finished
Full Wire&Clip

Partial Lit Fix
Lamp&Finish Rate

Partial Lit Fix
Lamp&Finished

Journeyman Finished
Partial Lamp&Finish

Complete Lit Fix
Lamp&Finish Rate

Lit Fix In Place

Journeyman Starting
Lamp&Finish

Journeyman Completing
Lamp&Finish

Journeyman Finished
Full Lamp&Finish

Figure 5.23: Representation of the quantitative SD model for “Base” crew case (Part B)
189

Time of
Operation

<Time>
Unit Apprentice
Cost
Total Apprentice
Cost per hour

Total
Cost of
Operation

Total Journeyman
Cost per hour

Total Lift Cost
per hour
Unit Cost
of
Operation

Lift
Utilization

Lift Idle
Percentage

Number of
Apprentice
Unit Journeyman
Cost
Number of
Journeyman
Number of Lift

Unit Lift Cost
<Lit Fix In Place>

<Lift Waiting>

Apprentice
Idle
Percentage

Journeyman
Idle
Percentage

<Apprentice
Waiting>

<Journeyman
Waiting>

Apprentice
Utilization

Journeyman
Utilization

Operation
Production
Rate

<Lit Fix In Place>

Assembling
Production
Rate

<Lit Fix
Assembled>

Figure 5.24: Representation of the quantitative SD model for “Base” crew case Part (C)

190

<Time of
Operation>

After modeling the three crew designs in SD, crew design characteristics and performance
outcomes such as duration of operation, total cost of operation, unit cost of operation, resource
utilizations, wait/idle times and production rates etc. are generated for the three quantitative SD
models and compared with DES simulation results in the following Table 5.7.
Table 5.7: Comparison of DES and SD simulation results
DES (Stroboscope)
Crew
Case

SD (Vensim)
Time
Ratio

Cost
Ratio

Time
(Hr)

Unit
Cost ($)

OPR
Fix/Hr

Time
(hr)

Unit
Cost ($)

OPR
Fix/Hr

Base

130.15

9.06

2.31

130.13

9.062

2.30

0.99984

1.0002

Improved

45.03

7.61

6.66

50.23

8.48

5.97

1.1154

1.1143

Lean

60.17

5.69

4.98

60.16

5.69

4.98

0.99983

1.0000

In order to validate the three crew design models in SD, simulation results from SD are
compared with the DES results. The SD simulation results for time of operation, unit cost and
operation production rates are very similar to DES simulation results.
The SD results for the “Base” and “Lean” crew cases are almost the same with less than
0.0003% variation; however some variation is present in the results for the “Improved” crew
case. This could be attributed due to the complexity present in the process of the “Improved”
crew case in DES. When the “Improved” crew case DES model is converted into SD certain
assumptions are made in order to represent the same process and flow of entity (light fixtures)
and resources (1 apprentice, 3 journeymen, and 3 lifts) that results in variation in the results.

191

5.5

Crew Design Analysis and Comparison

In the previous section, a well-documented construction operation focusing on the installation of
light fixtures (Howell et al 1993) was modeled and simulated using DES and SD for two
different crew designs (Base and Improved) having different crew size and composition. The
“Base” and “Improved” crew designs are enhanced & optimized based on the lean crew design
guidelines proposed in the chapter 4 to develop a “Lean” crew design. In this section, the “Lean”
crew design is compared with the “Base” and “Improved” crew designs to finalize the impact of
lean crew design guidelines. For the scope of this research, it was not possible to analyze and
compare the crew designs based on all the proposed lean crew design characteristics. Hence, few
characteristics from both the flow and work characteristics were selected to analyze and compare
the three crew designs (Base, Improved and Lean).
The flow characteristics selected were: (1) motion, (2) transport, (3) waiting, (4) overproduction, (5) inventory, and (6) defects. And, the work characteristics selected were: (7) crew
size, (8) crew flexibility, (9) crew heterogeneity, (10) task variety, (11) task identity, and (12)
support.
1) MOTION:
Motion is an important flow characteristic affecting crew performance in terms of nonprocessing time and consequently the duration and cost of work performed by a crew. Motion
(number of moves) in the 3 crew designs is shown in the Table 5.8.
Table 5.8: Motion in 3 crew designs
Motion/Number of
Moves
(to restock 300 light
fixtures)

Base

Improved

Lean

300

20

300

192

In the “Improved” crew design, all three Journeymen have to go down after working on the 15
light fixtures taking a total of 20 moves to restock 300 light fixtures at the rate of 15 light
fixtures at a time signifying excessive motion among all three Journeymen. The “Base” and
“Lean” crew designs both take the same number of moves.
However, in the “lean” crew design, motion is restricted to the Apprentice A who is
responsible for restocking unlike the “Base” crew design where Journeyman is coming down for
restocking while the Apprentice is waiting towards the end. Hence, motion is best designed in
the “Lean” crew leading to least duration and unit cost of work performed among the three crew
designs.
2) TRANSPORT:
Transport is an important flow characteristic affecting crew performance in terms of nonprocessing time and consequently the duration and cost of work performed by a crew. Nonprocessing time in the 3 crew designs is shown in the Table 5.9.
Table 5.9: Non-processing time in 3 crew designs
Non-processing Time
(to restock 300 light
fixtures)

Base

Improved

Lean

900 mins

300 mins

900 mins

In the “Improved” case, restocking was done for a batch size of 15 light fixtures; 3 minutes to
restock, 1 minute to go up and 1 minute for the lift to come down to restock the next fixture;
making it a total of 5 minutes each for 15 light fixtures requiring 20 moves to restock all the 300
fixtures. Hence a total of 300 minutes ((3*300*({3+1+1}/15)) were spent in motion and
transportation of 300 light fixtures.

193

Whereas, a total of 900 minutes (300 moves * 3 minutes each) were spent in motion and
transportation of 300 light fixtures in both the “Base” and “Lean” crew designs. However, in the
“Lean” crew design, the operating time to restock was included in the assembling cycle utilizing
the journeyman that was earlier utilized only one-third of the time. Hence, transportation (nonprocessing time) is best designed in the “Lean” crew leading to least duration and unit cost of
work performed among the three crew designs.
3) WAITING:
Waiting is an important flow characteristic affecting crew performance in terms of resource
utilization, non-processing time and consequently the overall crew productivity, duration and
cost of work performed by a crew. Resource wait/Idle time in the 3 crew designs is shown in the
Figure 5.25.
Average Apprentice Idle Time
Average Labor Idle Time
18
Average Idle/Wait Time

16

Average Journeyman Idle Time
Percent Improvement

16.97

91.38

90
80

14

70

12

60

10

8.99

8

50

8.96

40

6

30

4.48

4
2

100

1.01

0
Base Crew

20
0.48

1.02

Improved Crew

0.53
Lean Crew

10

0.775

0
Percent
Improvement

Figure 5.25: Resource Idle/Wait times in 3 crew designs

194

The above Figure 5.25 shows that “Lean” crew has the lowest resource idle/wait times among
the three crew designs leading to decrease in the non-processing time and consequently reducing
the duration and cost of work performed.
Also, the reduced idle/wait times leads to higher resource utilizations in the “Lean” crew
(Figure 5.26) that in turn leads to increasing the overall crew productivity.

Apprentice Utlization

Journeyman Utlization

Average Labor Utlization

Percent Improvement
40

Resource Utilizations (%)

120
94.94

100

89.88

40

35
30
25

61.51 69.86

62.26

60

94.43

92.19

80

34.07

95.55

20

34.65

15
10

20

5
0

0
Base Crew

Improved Crew

Lean Crew

Percent
Improvement

Figure 5.26: Resource Utilizations in 3 crew designs
4) OVER-PRODUCTION:
Over-production is an important flow characteristic affecting crew performance in terms of
work-in-progress (WIP), non-uniform work load (production rates), non-continuous flow and
duration and cost of work performed.

195

The Figure 5.27 shows that high work-in-progress (WIP) in the “Base” and “Improved” crew
designs has been completely eliminated in the “Lean” crew.
Work-in-progress (WIP)

Percent Improvement
120

120

Work-in-progress (WIP)

100

100

97.64

100

80

80

60

60

40

40

28.29

20

20
0

0
Base Crew

Improved Crew

0

Lean Crew

Percent
Improvement

Figure 5.27: Work-in-progress (WIP) in the three crew designs
This is also evident from the assembling and installation cycle times in the three crew
designs shown in the Table 5.10:
Table 5.10: Cycle times of assembling and installation activities in three crew designs
Activity

Task

Base Crew

Improved Crew

Assembling

Assemble

9 min

9 min

3 min

5 min

Restocking
Install
Installation

Wire & Clip
Lamp & Finish

7 min
8.5
min
7.5
min

4 min
26 min

8 min
4 min

196

Lean Crew
9 min
3 min

12 min

3.5
21 min

4.25
3.75

11.5
min

The cycle times for assembling and installation activities in the “Lean” crew design are almost
balanced and hence all the light fixtures assembled are processed and installed without having
any work in progress leading to a continuous flow.
5) INVENTORY
High Inventory also leads work-in-progress. In the “Lean” crew, the batch size was reduced to
only one fixture at a time from batch size of 15 fixtures in the “Improved” crew design to control
the issue of over production.
The light fixtures in the “Lean” crew were only restocked by the apprentice as needed for
installation unlike the “Improved” crew design where 15 light fixtures were restocked at a time.
Also, reduction in the batch size along with leveling production rates of different cycles and
applying pull system led to generate high operation throughput (Figure 5.28).

Operation Production Rate
(Fixture/Hour)

6
4.98

5

53.61

50

4
3

60

40
30

2.31

2

20

1

10

0

0
Base Crew

Lean Crew

Operation Production Rate

Percent Improvement
Percent Improvement

Figure 5.28: Operation production rates between Base and Lean crew design
6) DEFECTS
Defect is an important flow characteristic affecting crew performance in terms of quality,
duration and cost of work performed. From the evaluation of the crew size and composition in
197

the three crew design, it can be qualitatively inferred that due to journeyman and apprentice
working together in the installation cycle, there is team work present that may reduce the number
of defects that are overlooked in the earlier crew designs and improve the quality of the work
performed.
7) CREW SIZE:
Crew Size is an important work characteristic affecting crew performance in terms of production
capacity, satisfaction, learning, duration and cost of work performed. A reasonable increase in
the size of the crew in the “Lean” crew increased the production rate of the installation cycle
from 2.21 fixtures/hour (“Base” crew design) to 4.58 fixtures/hour (“Lean crew design) without

Production Rates (Fixture/Hour)

increasing the cost of work performed (Figure 5.29).
10

$9.06

8

6.67

$5.69

6

4.58

4
2

2

5

3

2.21

0
Base Crew

Lean Crew

Crew Size

Installation Production Rate

Maximum production Rate

Unit Cost

Figure 5.29: Impact of crew size on the production rates between the Base and Lean crew design
8) CREW FLEXIBILITY
Crew flexibility is an important work characteristic affecting crew performance in terms of
satisfaction, learning, duration and cost of work performed. It can be qualitatively inferred that in
the “Lean” crew design, the crew flexibility is higher than the other two “Base” and “Improved”
198

crew designs with the addition of one more apprentice in the installation cycle and assigning
restocking to the first apprentice.
Higher crew flexibility increases the overall crew productivity (operation production rate) by
increasing the utilization and reducing the wait time for the first apprentice. Also, higher crew
flexibility may lead to satisfaction and learning among the crew members arising from the
feeling of helping each other in the installation cycle (journeyman and apprentice working
together) and learning from the each other as a result of working together in a team.
9) CREW HETEROGENIETY
Crew heterogeneity is an important work characteristic affecting crew performance in terms of
support, motivation and learning. It can be qualitatively inferred that in the “Lean” crew design,
the crew heterogeneity is higher than the other two “Base” and “Improved” crew designs.
The “Lean” crew design of having the workers with different skill, knowledge and
experience such as journeyman working together with the apprentice in the installation cycle,
may lead to motivation in the journeyman to share his expertise and knowledge with the
apprentice and consequently may lead to learning in the apprentice working together with him.
10) TASK VARIETY
Task variety is an important work characteristic affecting crew performance in terms of
satisfaction and learning and pertains to the concept of job-enlargement from the STS theory. It
can be qualitatively inferred that in the “Lean” crew design, the task variety is higher than the
other two “Base” and “Improved” crew designs resulting from the crew design where apprentice
is given the opportunity to be part of more than one task (in Lean crew design apprentice takes
part in both the assembling and installation cycle).

199

Hence, being part of more than one task, may lead to more satisfaction in the apprentice as they
would not be executing the same task/work over and over again. Also, some learning may occur
in the apprentice arising from opportunity to work on a variety of tasks now.
11) TASK IDENTITY
Task identity is an important work characteristic affecting crew performance in terms of pride,
satisfaction and motivation. It can be qualitatively inferred that in the “Lean” crew design, the
task identity is higher than the other two “Base” and “Improved” crew designs resulting from the
fact that now the apprentice is involved in executing both the assembling and installation cycles
and consequently can now relate themselves with a both assembling and the installation
(considered to be major part in the operation) taking ownership of the finished product. Also,
increased task identity for the apprentice may also lead to higher satisfaction and motivation
levels in the “Lean” crew design.
12) SUPPORT
Support is an important work characteristic affecting crew performance in terms of learning,
satisfaction and motivation. It can be qualitatively inferred that in the “Lean” crew design, the
support is higher than the other two “Base” and “Improved” crew designs resulting from the fact
that now the apprentice and journeyman are working together providing help to each other if
needed.
5.6

Chapter Summary

This chapter provided an overview of DES and SD computer simulation models with
experimentation about crew interaction/design attributes in order to demonstrate how the crew
design characteristics affect the performance of the crew. The simulation models are developed
first using Discrete Event Simulation (DES) and then moving to System Dynamics (SD)
200

providing the qualitative and quantitative crew design models. And, finally analysis of the
simulation results took place to finalize the impact of the different crew design guidelines.

201

6 CONCLUSION
This research proposed lean crew design guidelines to guide the design of crews for construction
operations. Discrete Event Simulation and System Dynamics models were developed to test the
impact of guidelines on the crew design. The guidelines optimize both the crew and the work
performed, resulting in better overall crew performance. This chapter provides an overview,
contributions and conclusions of this research.
6.1

Research Overview

The overall goal of this research is to develop crew design guidelines for construction operations.
The purpose of this research is achieved by conducting an extensive literature review on the
existing construction models, production system theories, and crew & work design models.
Research studies on “work design/team design” that have well developed theoretical and
empirical foundations in Lean Production, Lean Construction, Socio-Technical System theory,
Social and Organizational Psychology were investigated to identify important soft and hard
factors and synthesized into lean crew design characteristics.
After, synthesizing the existing crew design and work design research into lean crew design
characteristics, lean-based crew design guidelines are proposed to guide the crew design to
achieve lean construction ideals of minimizing waste and maximizing value for construction
operations. A Lean-based Crew Design (LbCD) Model was developed and proposed to guide the
crew design process in a production setting involving initial formation of the crew, structuring of
the work, execution of work/implementation (actual work/simulation model execution) and
refining the crew design to maximize the crew performance based on the lean crew design
guidelines.
202

A combination of DES and SD computer simulation models were developed in order to
effectively demonstrate how the crew design characteristics affect the performance of a crew. A
case study involving a well-documented construction operation from a prior research work
conducted by Howell et al (1993) was selected for the purpose of modeling and simulation. The
case study consists of a real time construction operation focusing on the installation of light
fixtures by a construction crew. Three crew designs (Base, Improved, and Lean) with different
crew size and composition were analyzed and compared to determine the impact of lean crew
design guidelines on the performance of the crew.
6.2

Research Contributions

Performance of crews is very crucial to the work being executed and consequently for the overall
project performance. There are numerous factors which affect the crew executing the various
interconnected tasks involved in any operation. How to form or design a crew to carry out an
operation in a dynamically changing construction environment is a complex task. However,
despite the clear significance of crew design, only a modest scholarly interest in this area has
been seen in the last two decades.
While research on the design of construction operations has come a long way, this research
adds a first step in expanding it to the design of crews conducting construction operations, which
will get us closer in achieving Lean Construction ideals. It benefits the construction industry by
providing basis for designing crew interaction and roles during performing a construction
operation.
This research aimed at contributing to the gap in the crew design research in literature by
developing and proposing lean crew design characteristics and guidelines to maximize the
performance of construction crews to achieve lean construction ideals. It contributes in
203

identifying the important soft and hard factors from existing production system, crew and work
design models in the construction industry affecting the performance of crews and by
synthesizing the existing crew and work design approaches from lean production, lean
construction, socio-technical system theory, and social and organizational psychology into lean
crew design characteristics.
The lean crew design characteristics are comprised of “primary crew design characteristics”,
“intermediate crew design characteristics and outcomes” and “final crew performance
outcomes”. The primary work and flow characteristics that are again divided into waste, task,
social and crew characteristics. The lean-based crew design guidelines are developed for
maximizing joint optimization of both the crew and the work performed. They lead to better
overall crew performance by optimizing performance indicators such as (1) duration of work
performed; (2) cost of work performed; (3) productivity; (4) quality of work performed; (5)
worker satisfaction; (6) worker learning; (7) worker motivation; (8) pride in work performed. In
addition, simulation modeling has traditionally been carried out using discrete event simulation;
however, this research ventured into the area of construction simulation by using a combination
of simulation techniques such as Discrete Event Simulation and System Dynamics. In this
research, three crew design model were developed and simulated using both DES and SD in
order to demonstrate and analyze both the hard and soft factors affecting crew design.
DES was used to demonstrate the process and operational details of the crew design model
and SD was used to develop qualitative model demonstrating the various cause-and-effect
relationships and feedback among the crew design characteristics and the crew performance
outcomes. And, the quantitative SD models helped in demonstrating and evaluating the pattern

204

of behavior of various crew design characteristics and crew performance outcomes that is not
possible with the use of DES.
6.3

Conclusion

6.3.1

Quantitative Findings from Case Study

In this research, DES and SD computer simulation models are used to demonstrate the use of
lean-based crew design guidelines. The simulation models involved experimentation about crew
interaction/design attributes in order to demonstrate how the crew design characteristics affect
the performance of the crew. A well-documented construction operation focusing on the
installation of light fixtures (Howell et al 1993) was modeled and simulated using DES and SD.
Three crew design models were developed in order to demonstrate and analyze both the hard and
soft factors affecting crew design. DES was used to demonstrate the process and operational
details of the crew design model and SD was used to develop qualitative model demonstrating
the various cause-and-effect relationships and feedback among the crew design characteristics
and the crew performance outcomes. And, the quantitative SD models helped in demonstrating
and evaluating the pattern of behavior of various crew design characteristics and crew
performance outcomes that is not possible with the use of DES.
For the scope of this research, few characteristics from both the flow and work characteristics
were selected to analyze and compare the three crew designs (Base, Improved and Lean)
modeled and simulated in both DES and SD. Simulation analysis and comparison of the three
crew design (Base, Improved and Lean) showed that the “Lean” crew design based on the lean
crew design guidelines resulted in maximum performance of the crew:
1) Duration of work performed in the “Lean” crew was reduced to only 60.17 hours from
130.15 hours with 53% time savings (Figure 6.1).
205

2) Unit cost of work performed in the “Lean” crew was reduced to only $5.69 from $9.06

12
Unit Cost $

10
8

125.60

130.15

$8.32

$7.61
$9.06

6

$5.69

4

60.17

45.03

2
0
Budgeted

Base Crew

Improved Crew

140
120
100
80
60
40
20
0

Time of Operation Hrs

with 37.20% cost savings (Figure 6.1)

Lean Crew

Figure 6.1: Unit Cost and Time of Operation Savings in Lean crew
3) Overall crew productivity (operation production rate) in “Lean” crew was increased to
4.98 light fixtures/hour from 2.31 light fixtures/hour with 53.61% operation production
improvement (Figure 6.2).

Production Rates
(Fixture/Hour)

6
4.98

5

53.61

4
3
2

60
50
40
30

2.31

20

1

10

0

0
Base Crew

Lean Crew

Operation Production Rate

Percent Imrovement
Percent Imrovement

Figure 6.2: Overall Crew productivity (production rate) improvement in Lean crew

206

6.3.2

Qualitative Findings from Literature Review

In this research, lean crew design characteristics were developed by synthesizing the existing
crew and work design approaches from lean production, lean construction, socio-technical
system theory, and social and organizational psychology. These characteristics formed the basis
for developing the proposed lean-based crew design guidelines.
The lean crew design guidelines were used to refine the “Base” and “Improved” crew designs
from the light fixture case study and crew performance was qualitatively evaluated in terms of:
quality of work performed, worker satisfaction, worker learning, worker motivation, pride in
work performed. Based on the extensive literature review and the lean crew design guidelines,
qualitative comparisons of the three crew designs (Base, Improved and Lean) indicate that the
crew performance (quality of work performed, worker satisfaction, worker learning, worker
motivation, pride in work performed) is potentially improved in lean crew design compared to
the base and improved crew designs (discussed in detail in the crew design analysis and
comparison section in chapter 5).
In the opinion of the author, application of lean crew design guidelines would result in
maximizing the performance of the Lean crew in the following way:
1) Quality of work performed in the lean crew design would be improved due to reduction
in overburden and task interdependence compared to the base and improved crew
designs, where there is high overburden on the journeyman installing the fixtures alone.
2) Worker satisfaction in the lean crew design would be improved due to higher crew
flexibility, crew heterogeneity and increased support with the addition of an apprentice to
install the light fixtures along with the journeyman. Also with the addition of an
apprentice for installation activity, task variety and task identity are increased in the lean

207

crew design compared to the base and improved crew design where apprentice is only
assembling the fixtures with limited task variety, crew flexibility and crew heterogeneity.
3) Worker learning in the lean crew design would be improved due to higher crew
flexibility, crew heterogeneity, coordination and task variety where worker learning is
encouraged between the apprentice and the journeyman working together compared to
the base and improved crew designs.
4) Worker motivation in the lean crew design would be improved due to higher crew
heterogeneity, coordination and task identity among the crew compared to the base and
improved crew designs.
5) Pride in the work performed in the lean crew design would be improved due to increased
task identity with the apprentice now taking pride in executing a significant part of the
operation such as installation of light fixtures compared to the base and improved crew
designs where the apprentice is only assembling the light fixtures having very limited
pride.

208

APPENDICES

209

Appendix A: DES “Base” Crew Model Formulation and Simulation Results.

The input and output parameters used to formulate the discrete event simulation “Base” model
are shown below:
1. AprntcCost = Cost of 1 Apprentice per Hour = $6
2. JrnymnCost = Cost of 1 Apprentice per Hour = $13.4
3. LftCost = Cost of 1 Lift per Hour = $1.5
4. AprntcNumb = Number of Apprentice in Crew = AprntcWt.CurCount
5. JrnymnNumb = Number of Journeyman in Crew = JrnymnWt.CurCount
6. TOO = Time of Operation in Hours = SimTime/60
7. AprntcUtlizR

=

Apprentice

Utilization

in

%

(Resource)

=

(1-

(AprntcWtQ/AprntcNumb))*100
8. AprntcUtlizT = Apprentice Utilization in % (Time) = (1-(JrnymnWtQ/JrnymnNumb))*100
9. JrnymnUtlizR = Journeyman Utilization in % (Resource)
=((Restocking.AveDur+LftGoinUp.AveDur+Install.AveDur)/((LftGoinUpWt.AveWait+Jrny
mnWt.AveWait) +(Restocking.AveDur+LftGoinUp.AveDur+Install.AveDur)))*100
10. JrnymnUtlizT = Journeyman Utilization in % (Time)
=((Restocking.AveDur+LftGoinUp.AveDur+Install.AveDur)/((LftGoinUpWt.AveWait+Jrny
mnWt.AveWait) +(Restocking.AveDur+LftGoinUp.AveDur+Install.AveDur)))*100
11. LftUtlizR = Lift Utilization in in % (Resource)
= (1- ((LftDwnWt.AveCount+LftGoinUpWt.AveCount)/LftNumb))*100
12. LiftUtlizT = Lift Utilization in % (Time)

210

=(Restocking.AveDur+LftGoinUp.AveDur+Install.AveDur+LftGoinDwn.AveDur)/((LftGoi
nUpWt.AveWait+LftDwnWt.AveWait)
+(Restocking.AveDur+LftGoinUp.AveDur+Install.AveDur+LftGoinDwn.AveDur)))*100
13. TotalCost = Total Cost of the Operation ($)
=TOO*((AprntcCost*AprntcNumb) +(JrnymnCost*JrnymnNumb)+(LftCost*LftNumb))
14. UnitCost = Unit Cost of the Operation ($)
= TotalCost/LitFixInPlcQ
Please note:
The discrete event simulation results for the “Base” model are uploaded online and can be
accessed at the following website:
https://docs.google.com/file/d/0BxrMYHbElFuNNkhCdi1JdmRQdEk/edit

Appendix B: DES “Improved” Crew Model Formulation and Simulation Results.
The input and output parameters used to formulate the discrete event simulation “Improved”
model are shown below:
1. AprntcCost = Cost of 1 Apprentice per Hour = $6
2. JrnymnCost = Cost of 1 Apprentice per Hour = $13.4
3. LftACost = Cost of Lift A per Hour= $1.5
4. LftBCost = Cost of Lift B per Hour= $1.5
5. LftCCost = Cost of Lift C per Hour= $1.5
6. JrnymnANumb = Number of Journeyman A in Crew = JrnymnAWt.CurCount
7. JrnymnBNumb = Number of Journeyman B in Crew = JrnymnBWt.CurCount
8. JrnymnCNumb = Number of Journeyman C in Crew = JrnymnCWt.CurCount
9. AprntcNumb = Number of Apprentice in Crew = AprntcWt.CurCount
10. LitFixInPlcQ = Total Number of Light Fixtures in Place = LitFixInPlc.CurCount
211

11. AprntcWtQ = Average Content of Apprentice in Queue = AprntcWt.AveCount
12. JrnymnAWtQ = Average Content of Journeyman A in Queue = JrnymnAWt.AveCount
13. JrnymnBWtQ = Average Content of Journeyman B in Queue = JrnymnBWt.AveCount
14. JrnymnCWtQ = Average Content of Journeyman C in Queue = JrnymnCWt.AveCount
15. LftAUpWtQ = Average Content of Lift A Up in Queue = LftAUpWt.AveCount
16. LftADwnWtQ = Average Content of Lift A Down in Queue = LftADwnWt.AveCount
17. LftAGoinUpWtQ

=

Average

Content

of

Lift

A

GoingUp

in

Queue

=

LftAGoinUpWt.AveCount
18. LftAGoinDwnWtQ
= Average Content of Lift A Going Down in Queue = LftAGoinDwnWt.AveCount
19. LftBUpWtQ = Average Content of Lift B Up in Queue = LftBUpWt.AveCount
20. LftBDwnWtQ = Average Content of Lift B Down in Queue = LftBDwnWt.AveCount
21. LftBGoinUpWtQ

=

Average

Content

of

Lift

B

GoingUp

in

Queue

=

LftBGoinUpWt.AveCount
22. LftBGoinDwnWtQ = Average Content of Lift B Going Down in Queue
= LftBGoinDwnWt.AveCount
23. LftCUpWtQ = Average Content of Lift C Up in Queue = LftCUpWt.AveCount
24. LftCDwnWtQ = Average Content of Lift C Down in Queue = LftCDwnWt.AveCount
25. LftCGoinUpWtQ

=

Average

Content

of

Lift

C

GoingUp

LftCGoinUpWt.AveCount
26. LftCGoinDwnWtQ = Average Content of Lift C Going Down in Queue
= LftCGoinDwnWt.AveCount
27. AssmblD = Average Duration of Assemble = Assemble.AveDur

212

in

Queue

=

28. JrnymnARestckD = Average Duration of Restocking by Journeyman A
= JrnymnARestck.AveDur
29. JrnymnBRestckD = Average Duration of Restocking by Journeyman B
= JrnymnARestck.AveDur
30. JrnymnCRestckD = Average Duration of Restocking by Journeyman C
= JrnymnCRestck.AveDur
31. InstallD = Average Duration of Installation = Install.AveDur
32. WirnClipD = Average Duration of Wire and Clipping = WirnClip.AveDur
33. LmpnFnsD = Average Duration of Lamp and Finish = LmpnFns.AveDur
34. LftAGoinDwnD = Average Duration of of Lift A Going Down = LftAGoinDwn.AveDur
35. LftAGoinUpD = Average Duration of of Lift A Going Up = LftAGoinUp.AveDur
36. LftBGoinDwnD = Average Duration of of Lift B Going Down = LftBGoinDwn.AveDur
37. LftBGoinUpD = Average Duration of Lift B Going Up = LftBGoinUp.AveDur
38. LftCGoinDwnD = Average Duration of of Lift C Going Down
39. LftCGoinUpD = Average Duration of Lift C Going Up = LftCGoinUp.AveDur
40. JrnymnAWtD = Average Duration of Journeman A in Queue = JrnymnAWt.AveWait
41. JrnymnBWtD = Average Duration of Journeman B in Queue = JrnymnBWt.AveWait
42. JrnymnCWtD = Average Duration of Journeman C in Queue = JrnymnCWt.AveWait
43. AprntcWtD = Average Duration of Apprentice in Queue = AprntcWt.AveWait
44. LftAUpWtD = Average Duration of Lift A Up in Queue = LftAUpWt.AveWait
45. LftADwnWtD = Average Duration of Lift A Down in Queue = LftADwnWt.AveWait
46. LftAGoinUpWtD

=

Average

Duration

LftAGoinUpWt.AveWait

213

of

Lift

A

Going

Up

in

Queue

=

47. LftAGoinDwnWtD = Average Duration of Lift A Going Down in Queue
= LftAGoinDwnWt.AveWait
48. LftBUpWtD = Average Duration of Lift B Up in Queue =LftBUpWt.AveWait
49. LftBDwnWtD = Average Duration of Lift B Down in Queue = LftBDwnWt.AveWait
50. LftBGoinUpWtD

=

Average

Duration

of

Lift

B

Going

Up

in

Queue

=

LftBGoinUpWt.AveWait
51. LftBGoinDwnWtD = Average Duration of Lift B Going Down in Queue
= LftBGoinDwnWt.AveWait
52. LftCUpWtD = Average Duration of Lift C Up in Queue = LftCUpWt.AveWait
53. LftCDwnWtD = Average Duration of Lift C Down in Queue = LftCDwnWt.AveWait
54. LftCGoinUpWt

=

DAverage

Duration

of

Lift

C

Going

Up

in

Queue

=

LftCGoinUpWt.AveWait
55. LftCGoinDwnWtD = Average Duration of Lift C Going Down in Queue
= LftCGoinDwnWt.AveWait
56. AprntcUtlizR

=

Apprentice

Utilization

in

%

(Resource)

=

(1-

(Time)

=

(AprntcWtQ/AprntcNumb))*100
57. AprntcUtlizT

=

Apprentice

Utilization

in

%

((AssmblD)/(AprntcWtD+AssmblD))*100
58. JrnymnAUtlizR = Journeyman A Utilization in % (Resource)
= (1-(JrnymnAWtQ/JrnymnANumb))*100
59. JrnymnAUtlizT = Journeyman A Utilization in % (Time)
=(JrnymnARestckD+LftAGoinUpD+InstallD)/((LftAGoinUpWtD+LftAUpWtD+JrnymnA
WtD)+(JrnymnARestckD+LftAGoinUpWtD+InstallD)))*100

214

60. JrnymnBUtlizR = Journeyman B Utilization in % (Resource)
= (1-(JrnymnBWtQ/JrnymnBNumb))*100
61. JrnymnBUtlizT = Journeyman B Utilization in % (Time)
=((JrnymnBRestckD+LftBGoinUpD+WirnClipD)/((LftBGoinUpWtD+LftBUpWtD+Jrnymn
BWtD)+(JrnymnBRestckD+LftBGoinUpD+WirnClipD)))*100
62. JrnymnCUtlizR = Journeyman C Utilization in % (Resource)
= (1(1-(JrnymnCWtQ/JrnymnCNumb))*100
63. JrnymnCUtlizT = Journeyman C Utilization in % (Time)
=((JrnymnCRestckD+LftCGoinUpD+WirnClipD)/((LftCGoinUpWtD+LftCUpWtD+Jrnymn
CWtD)+(JrnymnCRestckD+LftCGoinUpD+LmpnFnsD)))*100
64. LftAUtlizR = Lift A Utilization in %(Resource)
= (1-((LftADwnWtQ+LftAGoinUpWtQ+LftAUpWtQ)/1))*100
65. LftAUtlizT = Lift A Utilization in %(Time)
=((JrnymnARestckD+LftAGoinUpD+InstallD+LftAGoinDwnD)/((LftADwnWtD+LftAGoin
UpWtD+LftAUpWtD)+(JrnymnARestckD+LftAGoinUpD+InstallD+LftAGoinDwnD)))*10
0
66. LftBUtlizR = Lift B Utilizationin in %(Resource)
= (1-((LftBDwnWtQ+LftBGoinUpWtQ+LftBUpWtQ)/1))*100
67. LftBUtlizT = Lift B Utilizationin in %(Time)
=((JrnymnBRestckD+LftBGoinUpD+WirnClipD+LftBGoinDwnD)/((LftBDwnWtD+LftBG
oinUpWtD+LftBUpWtD)+(JrnymnBRestckD+LftBGoinUpD+WirnClipD+LftBGoinDwnD)
))*100
68. LftCUtlizR = Lift C Utilizationin in %(Resource)

215

= (1-((LftCDwnWtQ+LftCGoinUpWtQ+LftCUpWtQ)/1))*100
69. TotalCostTotal = Cost of the Operation ($)
=TOO*((AprntcCost*AprntcNumb)+(JrnymnACost*1)+(JrnymnBCost*1)+(JrnymnCCost*1
)+(LftACost*1)+(LftBCost*1)+(LftCCost*1))
70. UnitCost = Unit Cost of the Operation ($) = TotalCost/LitFixInPlcQ
Please note:
The discrete event simulation results for the “Improved” model are uploaded online and can be
accessed at the following website:
https://docs.google.com/open?id=0BxrMYHbElFuNRnhtVklSeTE4UVE
Appendix C: DES “Lean” Crew Model Formulation and Simulation Results

The input and output parameters used to formulate the discrete event simulation “Lean” model
are shown below:
1. AprntcACost = Cost of 1 Apprentice per Hour = $6
2. JrnymnACost = Cost of Journeyman A per Hour = $13.4
3. AprntcBCost = Cost of Apprentice B per Hour = $6
4. LftACost = Cost of Lift A per Hour= $1.5
5. LftBCost = Cost of Lift B per Hour= $1.5
6. TOO = Time of Operation in Hours = SimTime/60
7. AprntcANumb = Number of Apprentice in Crew = AprntcAWt.CurCount
8. AprntcBNumb = Number of Apprentice B in Crew = AprntcBWt.CurCount
9. JrnymnANumb = Number of Journeyman A in Crew = JrnymnAWt.CurCount
10. LftANumb = Number of Lift A = LftADwnWt.CurCount
11. LftBNumb = Number of Lift B = LftBWt.CurCount

216

12. LitFixInPlcQ = Total Number of Light Fixtures in Place = LitFixInPlc.CurCount
13. AprntcBWtQ = Average Content of Apprentice B in Queue = AprntcBWt.AveCount
14. AprntcAWtQ = Average Content of Apprentice A in Queue = AprntcAWt.AveCount
15. JrnymnAWtQ = Average Content of Journeyman A in Queue = JrnymnAWt.AveCount
16. LftAUpWtQ = Average Content of Lift A Up in Queue = LftAUpWt.AveCount
17. LftADwnWtQ = Average Content of Lift A Down in Queue = LftADwnWt.AveCount
18. LftAGoinUpWtQ

=

Average

Content

of

Lift

A

GoingUp

in

Queue

=

LftAGoinUpWt.AveCount
19. LftBWtQ = Average Content of Lift B in Queue = LftBWt.AveCount
20. LftAGoinUpDur = Average Duration of of Lift A Going Up = LftAGoinUp.AveDur
21. LftAGoinDwnDur = Average Duration of of Lift A going Down = LftAGoinDwn.AveDur
22. RestockDur = Average Duration of Restocking = Restocking.AveDur
23. InstallDur = Average Duration of Installation = Install.AveDur
24. WirenClipDur = Average Duration of Wire and Clipping = WirenClip.AveDur
25. LampnFinishDur = Average Duration of Lamp and Finish = LampnFinish.AveDur
26. AprntcAWtDur = Average Duration of Aprntc A in Queue = AprntcAWt.AveWait
27. JrnymnAWtDur = Average Duration of Journeman A in Queue = JrnymnAWt.AveWait
28. AprntcBWtDur = Average Duration of Apprentice B in Queue = AprntcBWt.AveWait
29. LftAUpWtDur = Average Duration of Lift A Up in Queue = LftAUpWt.AveWait
30. LftADwnWtDur = Average Duration of Lift A Down in Queue = LftADwnWt.AveWait
31. LftAGoinUpWtDur =

Average Duration of Lift A Going Up in Queue =

LftAGoinUpWt.AveWait
32. LftBWtDur = Average Duration of Lift B in Queue = LftBWt.AveWait

217

33. AprntcBUtlizR

=

Apprentice

B

Utilization

in

%

(Resource)

=

(1-

(AprntcBWtQ/AprntcBNumb))*100
34. AprntcBUtlizT = Apprentice B Utilization in % (Time)
=((InstallDur+WirenClipDur+LampnFinishDur)/((AprntcBWtDur)+(InstallDur+WirenClipD
ur+LampnFinishDur)))*100
35. AprntcAUtlizR

=

Apprentice

A

Utilization

in

%

(Resource)

(1-

(AprntcAWtQ/AprntcANumb))*100
36. AprntcAUtlizT = Apprentice A Utilization in % (Time)
=((AssmblDur+RestockDur+LftAGoinUpDur+LftAGoinDwnDur)/((AprntcAWtDur+LftAU
pWtDur+LftAGoinUpWtDur)+(AssmblDur+RestockDur+LftAGoinUpDur+LftAGoinDwnD
ur)))*100
37. JrnymnAUtlizR = Journeyman A Utilization in % (Resource)
=(1-(JrnymnAWtQ/JrnymnANumb))*100
38. JrnymnAUtlizT = Journeyman A Utilization in % (Time)
=((InstallDur+WirenClipDur+LampnFinishDur)/((JrnymnAWtDur)+(InstallDur+WirenClip
Dur+LampnFinishDur)))*100
39. JrnymnAUtlizT = Journeyman A Utilization in % (Time)
=((InstallDur+WirenClipDur+LampnFinishDur)/((JrnymnAWtDur)+(InstallDur+WirenClip
Dur+LampnFinishDur)))*100
40. LftAUtlizRES = Lift A Utilizationin in %(Resource)
=(1-((LftAUpWtQ+LftADwnWtQ+LftAGoinUpWtQ)/LftANumb))*100
41. LftAUtlizTIM = Lift A Utilizationin in %(Time)

218

=((RestockDur+LftAGoinUpDur+LftAGoinDwnDur)/((LftAUpWtDur+LftADwnWtDur+Lf
tAGoinDwnDur)+(RestockDur+LftAGoinUpDur+LftAGoinDwnDur)))*100
42. LftBUtlizRES = Lift B Utilizationin in %(Resource) = (1-(LftBWtQ/LftBNumb))*100
43. LftBUtlizTIM = Lift B Utilizationin in %(Time)
=((InstallDur+WirenClipDur+LampnFinishDur)/((LftBWtDur)+(InstallDur+WirenClipDur+
LampnFinishDur)))*100
44. TotalCost = Total Cost of the Operation ($)
=TOO*((AprntcACost*AprntcANumb)+(AprntcBCost*AprntcBNumb)+(JrnymnACost*Jrn
ymnANumb)+(LftACost*LftANumb)+(LftBCost*LftBNumb))
45. UnitCost = Unit Cost of the Operation ($) = TotalCost/LitFixInPlcQ
Please note:
The discrete event simulation results for the “Lean” model are uploaded online and can be
accessed at the following website:
https://docs.google.com/file/d/0BxrMYHbElFuNRFhEWW9tOVRaRTQ/edit
Appendix D: SD “Base” Crew Model Formulation
The mathematical equations used to formulate the system dynamics “Base” model are shown
below:
(01)

Apprentice Assembling=
IF THEN ELSE( Apprentice Waiting>=1/9:AND:Lit Fix Waiting>=1/9 , 1/9 , 0
)
Units: (labor/Minute)

(02)

Apprentice Finished Assembling= INTEG (
+Apprentice Assembling-Apprentice Returning,
0)
Units: labor

(03)

Apprentice Idle Percentage=
IF THEN ELSE(Apprentice Waiting=1:OR:Apprentice Waiting=1/9, Apprentice
Waiting
219

*100 , 0 )
Units: **undefined**
(04)

Apprentice Returning=
IF THEN ELSE(Apprentice Finished Assembling=1, 1 , 0 )
Units: labor/Minute

(05)

Apprentice Utilization=
100-Apprentice Idle Percentage
Units: **undefined**

(06)

Apprentice Waiting= INTEG (
+Apprentice Returning-Apprentice Assembling ,
1)
Units: labor

(07)

Assembling Production Rate=
ZIDZ(Lit Fix Assembled, Time of Operation )
Units: **undefined**

(08)

Complete Lit Fix Installation Rate=
IF THEN ELSE(Partial Lit Fix Installed>=1, 1 , 0 )
Units: light fixture/Minute

(09)

"Complete Lit Fix Lamp&Finish Rate"=
IF THEN ELSE("Partial Lit Fix Lamp&Finished">=1, 1 , 0 )
Units: light fixture/Minute

(10)

"Complete Lit Fix Wire&Clip Rate"=
IF THEN ELSE("Partial Lit Fix Wire&Clipped">=1, 1 , 0 )
Units: light fixture/Minute

(11)

FINAL TIME = 7900
Units: Minute
The final time for the simulation.

(12)

Full Lit Fix Installed= INTEG (
+Complete Lit Fix Installation Rate-"Partial Lit Fix Wire&Clip Rate",
0)
Units: light fixture

(13)

"Full Lit Fix Wire&Clipped"= INTEG (
+"Complete Lit Fix Wire&Clip Rate"-"Partial Lit Fix Lamp&Finish Rate",
0)
Units: light fixture

220

(14)

INITIAL TIME = 0
Units: Minute
The initial time for the simulation.

(15)

Journeyman Completing Installation=
IF THEN ELSE(Journeyman Finished Partial Installation>=1, 1 , 0 )
Units: labor/Minute

(16)

"Journeyman Completing Lamp&Finish"=
IF THEN ELSE("Journeyman Finished Partial Lamp&Finish">=1, 1 , 0 )
Units: labor/Minute

(17)

"Journeyman Completing Wire&Clip"=
IF THEN ELSE("Journeyman Finished Partial Wire&Clip">=1, 1 , 0 )
Units: labor/Minute

(18)

Journeyman Finished Full Installation= INTEG (
+Journeyman Completing Installation-"Journeyman Starting Wire&Clip",
0)
Units: labor

(19)

"Journeyman Finished Full Lamp&Finish"= INTEG (
+"Journeyman Completing Lamp&Finish"-Journeyman Going Down,
0)
Units: labor

(20)

"Journeyman Finished Full Wire&Clip"= INTEG (
+"Journeyman Completing Wire&Clip"-"Journeyman Starting Lamp&Finish",
0)
Units: labor

(21)

Journeyman Finished Partial Installation= INTEG (
+Journeyman Starting Installation-Journeyman Completing Installation,
0)
Units: labor

(22)

"Journeyman Finished Partial Lamp&Finish"= INTEG (
+"Journeyman Starting Lamp&Finish"-"Journeyman Completing Lamp&Finish",
0)
Units: labor

(23)

"Journeyman Finished Partial Wire&Clip"= INTEG (
+"Journeyman Starting Wire&Clip"-"Journeyman Completing Wire&Clip",
0)
Units: labor

221

(24)

Journeyman Finished Restocking= INTEG (
+Journeyman Restocking-Journeyman Going Up,
0)
Units: labor

(25)

Journeyman Going Down=
IF THEN ELSE("Journeyman Finished Full Lamp&Finish">=1:AND:Lit Fix In

Place
>=1, 1 , 0 )
Units: labor/Minute
(26)

Journeyman Going Up=
IF THEN ELSE(Journeyman Finished Restocking>0:AND:Journeyman Finished
Restocking
<=1, 1 , 0 )
Units: labor/Minute
(27)

Journeyman Idle Percentage=
IF THEN ELSE(Journeyman Waiting=1, Journeyman Waiting*100 , 0 )
Units: **undefined**

(28)

Journeyman Restocking=
IF THEN ELSE(Journeyman Waiting>=1:AND:Lit Fix Assembled>=1, 1 , 0 )
Units: labor/Minute

(29)

Journeyman Starting Installation=
IF THEN ELSE(Journeyman Up>0:AND:Journeyman Up<=1, 1/6 , 0 )
Units: labor/Minute

(30)

"Journeyman Starting Lamp&Finish"=
IF THEN ELSE("Journeyman Finished Full Wire&Clip">0:AND:"Journeyman
Finished Full Wire&Clip"
<=1, 1/6 , 0 )
Units: labor/Minute
(31)

"Journeyman Starting Wire&Clip"=
IF THEN ELSE(Journeyman Finished Full Installation>0:AND:Journeyman
Finished Full Installation
<=1, 1/8 , 0 )
Units: labor/Minute
(32)

Journeyman Up= INTEG (
+Journeyman Going Up-Journeyman Starting Installation,
0)
Units: labor

222

(33)

Journeyman Utilization=
100-Journeyman Idle Percentage
Units: **undefined**

(34)

Journeyman Waiting= INTEG (
+Journeyman Going Down-Journeyman Restocking,
1)
Units: labor

(35)

Lift Completing Installation=
IF THEN ELSE(Lift Finished Partial Installation>=1, 1 , 0 )
Units: labor/Minute

(36)

"Lift Completing Lamp&Finish"=
IF THEN ELSE("Lift Finished Partial Lamp&Finish">=1, 1 , 0 )
Units: labor/Minute

(37)

"Lift Completing Wire&Clip"=
IF THEN ELSE("Lift Finished Partial Wire&Clip">=1, 1 , 0 )
Units: labor/Minute

(38)

Lift Finished Full Installation= INTEG (
+Lift Completing Installation-"Lift Starting Wire&Clip",
0)
Units: labor

(39)

"Lift Finished Full Lamp&Finish"= INTEG (
+"Lift Completing Lamp&Finish"-Lift Going Down,
0)
Units: labor

(40)

"Lift Finished Full Wire&Clip"= INTEG (
+"Lift Completing Wire&Clip"-"Lift Starting Lamp&Finish",
0)
Units: labor

(41)

Lift Finished Partial Installation= INTEG (
+Lift Starting Installation-Lift Completing Installation,
0)
Units: labor

(42)

"Lift Finished Partial Lamp&Finish"= INTEG (
+"Lift Starting Lamp&Finish"-"Lift Completing Lamp&Finish",
0)
Units: labor

223

(43)

"Lift Finished Partial Wire&Clip"= INTEG (
+"Lift Starting Wire&Clip"-"Lift Completing Wire&Clip",
0)
Units: labor

(44)

Lift Finished Restocking= INTEG (
+Lift Restocking-Lift Going Up,
0)
Units: labor

(45)

Lift Going Down=
IF THEN ELSE("Lift Finished Full Lamp&Finish">=1:AND:Lit Fix In Place>=1,
1,0)
Units: labor/Minute

(46)

Lift Going Up=
IF THEN ELSE(Lift Finished Restocking>0:AND:Lift Finished Restocking<=1,
1,0)
Units: labor/Minute

(47)

Lift Idle Percentage=
IF THEN ELSE(Lift Waiting<=1/9:OR:Lift Waiting>=8/9, Lift Waiting*100 , 0
)
Units: **undefined**

(48)

Lift Restocking=
IF THEN ELSE(Lift Waiting>=1:AND:Lit Fix Assembled>=1, 1 , 0 )
Units: labor/Minute

(49)

Lift Starting Installation=
IF THEN ELSE(Lift Up>0:AND:Lift Up<=1, 1/6 , 0 )
Units: labor/Minute

(50)

"Lift Starting Lamp&Finish"=
IF THEN ELSE("Lift Finished Full Wire&Clip">0:AND:"Lift Finished Full
Wire&Clip"
<=1, 1/6 , 0 )
Units: labor/Minute
(51)

"Lift Starting Wire&Clip"=
IF THEN ELSE(Lift Finished Full Installation>0:AND:Lift Finished Full
Installation
<=1, 1/8 , 0 )
Units: labor/Minute
(52)

Lift Up= INTEG (
224

+Lift Going Up-Lift Starting Installation,
0)
Units: labor
(53)

Lift Utilization=
100-Lift Idle Percentage
Units: **undefined**

(54)

Lift Waiting= INTEG (
+Lift Going Down-Lift Restocking,
1)
Units: labor

(55)

Lit Fix Assembled= INTEG (
+Lit Fix Assembling Rate-Lit Fix Restocking Rate,
0.0001)
Units: light fixture

(56)

Lit Fix Assembling Rate=
IF THEN ELSE( Lit Fix Waiting>=1/9:AND:Apprentice Waiting>=1/9, 1/9 , 0 )
Units: light fixture/Minute

(57)

Lit Fix Going Up Rate=
IF THEN ELSE(Lit Fix Restocked>0:AND:Lit Fix Restocked<=1, 1 , 0 )
Units: light fixture/Minute

(58)

Lit Fix In Place= INTEG (
+"Complete Lit Fix Lamp&Finish Rate",
0.1111)
Units: light fixture

(59)

Lit Fix Restocked= INTEG (
+Lit Fix Restocking Rate-Lit Fix Going Up Rate,
0)
Units: light fixture

(60)

Lit Fix Restocking Rate=
IF THEN ELSE(Lit Fix Assembled>=1:AND:Lift Waiting>=1:AND:Journeyman
Waiting
>=1, 1 , 0 )
Units: light fixture/Minute
(61)

Lit Fix Up= INTEG (
+Lit Fix Going Up Rate-Partial Lit Fix Installation Rate,
0)
Units: light fixture
225

(62)

Lit Fix Waiting= INTEG (
-Lit Fix Assembling Rate,
330)
Units: light fixture

(63)

Number of Apprentice=
1
Units: labor

(64)

Number of Journeyman=
1
Units: labor

(65)

Number of Lift=
1
Units: labor

(66)

Operation Production Rate=
ZIDZ(Lit Fix In Place, Time of Operation )
Units: **undefined**

(67)

"OPR-APR Ratio"=
ZIDZ(Operation Production Rate, Assembling Production Rate )
Units: **undefined**

(68)

Partial Lit Fix Installation Rate=
IF THEN ELSE(Lit Fix Up>0:AND:Lit Fix Up<=1, 1/6 , 0 )
Units: light fixture/Minute

(69)

Partial Lit Fix Installed= INTEG (
+Partial Lit Fix Installation Rate-Complete Lit Fix Installation Rate,
0)
Units: light fixture

(70)

"Partial Lit Fix Lamp&Finish Rate"=
IF THEN ELSE("Full Lit
Wire&Clipped"
<=1, 1/6 , 0 )
Units: light fixture/Minute
(71)

Fix

Wire&Clipped">0:AND:"Full

Lit

"Partial Lit Fix Lamp&Finished"= INTEG (
+"Partial Lit Fix Lamp&Finish Rate"-"Complete Lit Fix Lamp&Finish Rate",
0)
Units: light fixture

226

Fix

(72)

"Partial Lit Fix Wire&Clip Rate"=
IF THEN ELSE(Full Lit Fix Installed>0:AND:Full Lit Fix Installed<=1, 1/8
,0)
Units: light fixture/Minute

(73)

"Partial Lit Fix Wire&Clipped"= INTEG (
+"Partial Lit Fix Wire&Clip Rate"-"Complete Lit Fix Wire&Clip Rate",
0)
Units: light fixture

(74)

SAVEPER =
TIME STEP
Units: Minute [0,?]
The frequency with which output is stored.

(75)

Time of Operation=
Time/60
Units: hours

(76)

TIME STEP = 1
Units: Minute [0,?]
The time step for the simulation.

(77)

Total Apprentice Cost per hour=
Number of Apprentice*Unit Apprentice Cost
Units: dollars/hour

(78)

Total Cost of Operation=
Time of Operation*(Total Apprentice Cost per hour+Total Journeyman Cost per

hour
+Total Lift Cost per hour)
Units: dollars
(79)

Total Journeyman Cost per hour=
Number of Journeyman*Unit Journeyman Cost
Units: dollars/hour

(80)

Total Lift Cost per hour=
Number of Lift*Unit Lift Cost
Units: dollars/hour

(81)

Unit Apprentice Cost=
6
Units: dollars/(labor*hour)

(82)

Unit Cost of Operation=
227

Total Cost of Operation/Lit Fix In Place
Units: dollars/light fixture
(83)

Unit Journeyman Cost=
13.4
Units: dollars/(labor*hour)

(84)

Unit Lift Cost=
1.5
Units: dollars/(labor*hour)

The complete system dynamics model for the “Base” case is uploaded online and can be
accessed at the following website:
https://docs.google.com/open?id=0BxrMYHbElFuNUUdUX1F6WGE3Y3M
Appendix E: SD “Improved” Crew Model Formulation
The mathematical equations used to formulate the system dynamics “Improved” model are
shown below:
Model mathematical equations:
(001) "1st Half Lit Fix Restocked"= INTEG (
+"1st Half Lit Fix Restocking Rate"-"2nd Half Lit Fix Restocking Rate",
0)
Units: light fixture
(002) "1st Half Lit Fix Restocking Rate"=
IF THEN ELSE(Lit Fix Assembled>=15:AND:Lift A
Waiting>=1:AND:Journeyman A Waiting
>=1, 15 , 0 )
Units: light fixture/Minute
(003) "2nd Half Light Fix Restocked"= INTEG (
+"2nd Half Lit Fix Restocking Rate"-Complete Lit Fix Restocking Rate,
0)
Units: light fixture
(004) "2nd Half Lit Fix Restocking Rate"=
IF THEN ELSE("1st Half Lit Fix Restocked">=15, 15 , 0 )
Units: light fixture/Minute
(005) Apprentice Assembling=
IF THEN ELSE( Apprentice Waiting>=1/9:AND:Lit Fix Waiting>=1/9 , 1/9 , 0
228

)
Units: labor/Minute
(006) Apprentice Finished Assembling= INTEG (
+Apprentice Assembling-Apprentice Returning,
0)
Units: labor
(007) Apprentice Idle Percentage=
IF THEN ELSE(Apprentice Waiting<=1/9:OR:Apprentice Waiting>=8/9,
Apprentice Waiting
*100 , 0 )
Units: **undefined**
(008) Apprentice Returning=
IF THEN ELSE(Apprentice Finished Assembling=1, 1 , 0 )
Units: labor/Minute
(009) Apprentice Utilization=
100-Apprentice Idle Percentage
Units: **undefined**
(010) Apprentice Waiting= INTEG (
+Apprentice Returning-Apprentice Assembling ,
1)
Units: labor
(011) Assembling Cycle Production Rate=
ZIDZ( Lit Fix Assembled , Time of Operation )
Units: **undefined**
(012) Complete Lit Fix Installation Rate=
IF THEN ELSE(Partial Lit Fix Installed>=1, 1 , 0 )
Units: light fixture/Minute
(013) "Complete Lit Fix Lamp&Finish Rate"=
IF THEN ELSE("Partial Lit Fix Lamp&Finished">=1, 1 , 0 )
Units: light fixture/Minute
(014) Complete Lit Fix Restocking Rate=
IF THEN ELSE("2nd Half Light Fix Restocked">=15, 15 , 0 )
Units: light fixture/Minute
(015) "Complete Lit Fix Wire&Clip Rate"=
IF THEN ELSE("Partial Lit Fix Wire&Clipped">=1, 1 , 0 )
Units: light fixture/Minute
229

(016) FINAL TIME = 4000
Units: Minute
The final time for the simulation.
(017) Full Lit Fix In Place= INTEG (
+"Complete Lit Fix Lamp&Finish Rate",
0)
Units: light fixture
(018) Full Lit Fix Installed= INTEG (
+Complete Lit Fix Installation Rate-"Partial Lit Fix Wire&Clip Rate",
0)
Units: light fixture
(019) Full Lit Fix Restocked= INTEG (
+Complete Lit Fix Restocking Rate-Lit Fix Going Up Rate,
0)
Units: light fixture
(020) "Full Lit Fix Wire&Clipped"= INTEG (
+"Complete Lit Fix Wire&Clip Rate"-"Partial Lit Fix Lamp&Finish Rate",
0)
Units: light fixture
(021) INITIAL TIME = 0
Units: Minute
The initial time for the simulation.
(022) Journeyman A Completing Installation=
IF THEN ELSE(Journeyman A Finished Partial Installation>=1, 1 , 0 )
Units: labor/Minute
(023) Journeyman A Completing Restock=
IF THEN ELSE(Journeyman A Finished 2nd Half Restock>=1, 1 , 0 )
Units: labor/Minute
(024) Journeyman A Finished 2nd Half Restock= INTEG (
+Journeyman A Starting 2nd Half Restock-Journeyman A Completing Restock,
0)
Units: labor
(025) Journeyman A Finished Full Installation= INTEG (
+Journeyman A Completing Installation-Journeyman Going Down,
0)
Units: labor
230

(026) Journeyman A Finished Full Restock= INTEG (
+Journeyman A Completing Restock-Journeyman A Going Up,
0)
Units: labor
(027) Journeyman A Finished Partial Installation= INTEG (
+Journeyman A Starting Installation-Journeyman A Completing Installation,
0)
Units: labor
(028) Journeyman A Finished Partial Restocking= INTEG (
+Journeyman A Starting 1st Half Restock-Journeyman A Starting 2nd Half
Restock
,
0)
Units: labor
(029) Journeyman A Going Up=
IF THEN ELSE(Journeyman A Finished Full Restock>=1, 1 , 0 )
Units: labor/Minute
(030) Journeyman A Idle Percentage=
IF THEN ELSE(Journeyman A Waiting=1, Journeyman A Waiting*100 , 0 )
Units: **undefined**
(031) Journeyman A Starting 1st Half Restock=
IF THEN ELSE(Journeyman A Waiting>=1:AND:Lit Fix Assembled>=15, 1 , 0 )
Units: labor/Minute
(032) Journeyman A Starting 2nd Half Restock=
IF THEN ELSE(Journeyman A Finished Partial Restocking>=1, 1 , 0 )
Units: labor/Minute
(033) Journeyman A Starting Installation=
IF THEN ELSE(Journeyman A Up>0:AND:Journeyman A Up<=1, 1/59 , 0 )
Units: labor/Minute
(034) Journeyman A Up= INTEG (
+Journeyman A Going Up-Journeyman A Starting Installation,
0)
Units: labor
(035) Journeyman A Utilization=
100-Journeyman A Idle Percentage
Units: **undefined**
231

(036) Journeyman A Waiting= INTEG (
+Journeyman Going Down-Journeyman A Starting 1st Half Restock,
1)
Units: labor
(037) Journeyman B Completing Restock=
IF THEN ELSE(Journeyman B Finished 2nd Half Restock>=1, 1 , 0 )
Units: labor/Minute
(038) "Journeyman B Completing Wire&Clip"=
IF THEN ELSE("Journeyman B Finished Partial Wire&Clip">=1, 1 , 0 )
Units: labor/Minute
(039) Journeyman B Finished 2nd Half Restock= INTEG (
+Journeyman B Starting 2nd Half Restock-Journeyman B Completing Restock,
0)
Units: labor
(040) Journeyman B Finished Full Restock= INTEG (
+Journeyman B Completing Restock-Journeyman B Going Up,
0)
Units: labor
(041) "Journeyman B Finished Full Wire&Clip"= INTEG (
+"Journeyman B Completing Wire&Clip"-Journeyman B Going Down,
0)
Units: labor
(042) Journeyman B Finished Partial Restocking= INTEG (
+Journeyman B Starting 1st Half Restock-Journeyman B Starting 2nd Half
Restock
,
0)
Units: labor
(043) "Journeyman B Finished Partial Wire&Clip"= INTEG (
+"Journeyman B Starting Wire&Clip"-"Journeyman B Completing Wire&Clip",
0)
Units: labor
(044) Journeyman B Going Down=
IF THEN ELSE("Journeyman B Finished Full Wire&Clip">=1, 1 , 0 )
Units: labor/Minute
(045) Journeyman B Going Up=
232

IF THEN ELSE(Journeyman B Finished Full Restock>=1, 1 , 0 )
Units: labor/Minute
(046) Journeyman B Idle Percentage=
IF THEN ELSE(Journeyman B Waiting=1, Journeyman B Waiting*100 , 0 )
Units: **undefined**
(047) Journeyman B Starting 1st Half Restock=
IF THEN ELSE(Journeyman B Waiting>=1, 1 , 0 )
Units: labor/Minute
(048) Journeyman B Starting 2nd Half Restock=
IF THEN ELSE(Journeyman B Finished Partial Restocking>=1, 1 , 0 )
Units: labor/Minute
(049) "Journeyman B Starting Wire&Clip"=
IF THEN ELSE(Journeyman B Up>0:AND:Journeyman B Up<=1, 1/59 , 0 )
Units: labor/Minute
(050) Journeyman B Up= INTEG (
+Journeyman B Going Up-"Journeyman B Starting Wire&Clip",
0)
Units: labor
(051) Journeyman B Utilization=
100-Journeyman B Idle Percentage
Units: **undefined**
(052) Journeyman B Waiting= INTEG (
+Journeyman B Going Down-Journeyman B Starting 1st Half Restock,
1)
Units: labor
(053) "Journeyman C Completing Lamp&Finish"=
IF THEN ELSE("Journeyman C Finished Partial Lamp&Finish">=1, 1 , 0 )
Units: labor/Minute
(054) Journeyman C Completing Restock=
IF THEN ELSE(Journeyman C Finished 2nd Half Restock>=1, 1 , 0 )
Units: labor/Minute
(055) Journeyman C Finished 2nd Half Restock= INTEG (
+Journeyman C Starting 2nd Half Restock-Journeyman C Completing Restock,
0)
Units: labor

233

(056) "Journeyman C Finished Full Lamp&Finish"= INTEG (
+"Journeyman C Completing Lamp&Finish"-Journeyman C Going Down,
0)
Units: labor
(057) Journeyman C Finished Full Restock= INTEG (
+Journeyman C Completing Restock-Journeyman C Going Up,
0)
Units: labor
(058) "Journeyman C Finished Partial Lamp&Finish"= INTEG (
+"Journeyman C Starting Lamp&Finish"-"Journeyman C Completing
Lamp&Finish"
,
0)
Units: labor
(059) Journeyman C Finished Partial Restocking= INTEG (
+Journeyman C Starting 1st Half Restock-Journeyman C Starting 2nd Half
Restock
,
0)
Units: labor
(060) Journeyman C Going Down=
IF THEN ELSE("Journeyman C Finished Full Lamp&Finish">=1, 1 , 0 )
Units: labor/Minute
(061) Journeyman C Going Up=
IF THEN ELSE(Journeyman C Finished Full Restock>=1, 1 , 0 )
Units: labor/Minute
(062) Journeyman C idle Percentage=
IF THEN ELSE(Journeyman C Waiting=1, Journeyman C Waiting*100 , 0 )
Units: **undefined**
(063) Journeyman C Starting 1st Half Restock=
IF THEN ELSE(Journeyman C Waiting>=1, 1 , 0 )
Units: labor/Minute
(064) Journeyman C Starting 2nd Half Restock=
IF THEN ELSE(Journeyman C Finished Partial Restocking>=1, 1 , 0 )
Units: labor/Minute
(065) "Journeyman C Starting Lamp&Finish"=
IF THEN ELSE(Journeyman C Up>0:AND:Journeyman C Up<=1, 1/59 , 0 )
234

Units: labor/Minute
(066) Journeyman C Up= INTEG (
+Journeyman C Going Up-"Journeyman C Starting Lamp&Finish",
0)
Units: labor
(067) Journeyman C Utilization=
100-Journeyman C idle Percentage
Units: **undefined**
(068) Journeyman C Waiting= INTEG (
+Journeyman C Going Down-Journeyman C Starting 1st Half Restock,
1)
Units: labor
(069) Journeyman Going Down=
IF THEN ELSE(Journeyman A Finished Full Installation>=1:AND:Full Lit Fix
Installed
>=1, 1 , 0 )
Units: labor/Minute
(070) Lift A Completing Installation=
IF THEN ELSE(Lift A Finished Partial Installation>=1, 1 , 0 )
Units: labor/Minute
(071) Lift A Completing Restock=
IF THEN ELSE(Lift A Finished 2nd Half Restock>=1, 1 , 0 )
Units: labor/Minute
(072) Lift A Finished 1st Half Restock= INTEG (
+Lift A Starting 1st Half Restock-Lift A Starting 2nd Half Restock,
0)
Units: labor
(073) Lift A Finished 2nd Half Restock= INTEG (
+Lift A Starting 2nd Half Restock-Lift A Completing Restock,
0)
Units: labor
(074) Lift A Finished Full Installation= INTEG (
+Lift A Completing Installation-Lift A Going Down,
0)
Units: labor
(075) Lift A Finished Full Restock= INTEG (
235

+Lift A Completing Restock-Lift A Going Up,
0)
Units: labor
(076) Lift A Finished Partial Installation= INTEG (
+Lift A Starting Installation-Lift A Completing Installation,
0)
Units: labor
(077) Lift A Going Down=
IF THEN ELSE(Lift A Finished Full Installation>=1:AND:Full Lit Fix Installed
>=1, 1 , 0 )
Units: labor/Minute
(078) Lift A Going Up=
IF THEN ELSE(Lift A Finished Full Restock>=1, 1 , 0 )
Units: labor/Minute
(079) Lift A Idle Percentage=
IF THEN ELSE(Lift A Waiting=1, Lift A Waiting*100 , 0 )
Units: **undefined**
(080) Lift A Starting 1st Half Restock=
IF THEN ELSE(Lift A Waiting>=1:AND:Lit Fix Assembled>=15, 1 , 0 )
Units: labor/Minute
(081) Lift A Starting 2nd Half Restock=
IF THEN ELSE(Lift A Finished 1st Half Restock>=1, 1 , 0 )
Units: labor/Minute
(082) Lift A Starting Installation=
IF THEN ELSE(Lift A Up>0:AND:Lift A Up<=1, 1/59 , 0 )
Units: labor/Minute
(083) Lift A Up= INTEG (
+Lift A Going Up-Lift A Starting Installation,
0)
Units: labor
(084) Lift A Utilization=
100-Lift A Idle Percentage
Units: **undefined**
(085) Lift A Waiting= INTEG (
+Lift A Going Down-Lift A Starting 1st Half Restock,
1)
236

Units: labor
(086) Lift B Completing Restock=
IF THEN ELSE(Lift B Finished 2nd Half Restock>=1, 1 , 0 )
Units: labor/Minute
(087) "Lift B Completing Wire&Clip"=
IF THEN ELSE("Lift B Finished Partial Wire&Clip">=1, 1 , 0 )
Units: labor/Minute
(088) Lift B Finished 1st Half Restock= INTEG (
+Lift B Starting 1st Half Restock-Lift B Starting 2nd Half Restock,
0)
Units: labor
(089) Lift B Finished 2nd Half Restock= INTEG (
+Lift B Starting 2nd Half Restock-Lift B Completing Restock,
0)
Units: labor
(090) Lift B Finished Full Restock= INTEG (
+Lift B Completing Restock-Lift B Going Up,
0)
Units: labor
(091) "Lift B Finished Full Wire&Clip"= INTEG (
+"Lift B Completing Wire&Clip"-Lift B Going Down,
0)
Units: labor
(092) "Lift B Finished Partial Wire&Clip"= INTEG (
+"Lift B Starting Wire&Clip"-"Lift B Completing Wire&Clip",
0)
Units: labor
(093) Lift B Going Down=
IF THEN ELSE("Lift B Finished Full Wire&Clip">=1, 1 , 0 )
Units: labor/Minute
(094) Lift B Going Up=
IF THEN ELSE(Lift B Finished Full Restock>=1, 1 , 0 )
Units: labor/Minute
(095) Lift B Idle Percentage=
IF THEN ELSE(Lift B Waiting=1, Lift B Waiting*100 , 0 )
Units: **undefined**
237

(096) Lift B Starting 1st Half Restock=
IF THEN ELSE(Lift B Waiting>=1, 1 , 0 )
Units: labor/Minute
(097) Lift B Starting 2nd Half Restock=
IF THEN ELSE(Lift B Finished 1st Half Restock>=1, 1 , 0 )
Units: labor/Minute
(098) "Lift B Starting Wire&Clip"=
IF THEN ELSE(Lift B Up>0:AND:Lift B Up<=1, 1/59 , 0 )
Units: labor/Minute
(099) Lift B Up= INTEG (
+Lift B Going Up-"Lift B Starting Wire&Clip",
0)
Units: labor
(100) Lift B Utilization=
100-Lift B Idle Percentage
Units: **undefined**
(101) Lift B Waiting= INTEG (
+Lift B Going Down-Lift B Starting 1st Half Restock,
1)
Units: labor
(102) "Lift C Completing Lamp&Finish"=
IF THEN ELSE("Lift C Finished Partial Lamp&Finish">=1, 1 , 0 )
Units: labor/Minute
(103) Lift C Completing Restock=
IF THEN ELSE(Lift C Finished 2nd Half Restock>=1, 1 , 0 )
Units: labor/Minute
(104) Lift C Finished 1st Half Restock= INTEG (
+Lift C Starting 1st Half Restock-Lift C Starting 2nd Half Restock,
0)
Units: labor
(105) Lift C Finished 2nd Half Restock= INTEG (
+Lift C Starting 2nd Half Restock-Lift C Completing Restock,
0)
Units: labor
(106) "Lift C Finished Full Lamp&Finish"= INTEG (
238

+"Lift C Completing Lamp&Finish"-Lift C Going Down,
0)
Units: labor
(107) Lift C Finished Full Restock= INTEG (
+Lift C Completing Restock-Lift C Going Up,
0)
Units: labor
(108) "Lift C Finished Partial Lamp&Finish"= INTEG (
+"Lift C Starting Lamp&Finish"-"Lift C Completing Lamp&Finish",
0)
Units: labor
(109) Lift C Going Down=
IF THEN ELSE("Lift C Finished Full Lamp&Finish">=1, 1 , 0 )
Units: labor/Minute
(110) Lift C Going Up=
IF THEN ELSE(Lift C Finished Full Restock>=1, 1 , 0 )
Units: labor/Minute
(111) Lift C Idle Percentage=
IF THEN ELSE(Lift C Waiting=1, Lift C Waiting*100 , 0 )
Units: **undefined**
(112) Lift C Starting 1st Half Restock=
IF THEN ELSE(Lift C Waiting>=1, 1 , 0 )
Units: labor/Minute
(113) Lift C Starting 2nd Half Restock=
IF THEN ELSE(Lift C Finished 1st Half Restock>=1, 1 , 0 )
Units: labor/Minute
(114) "Lift C Starting Lamp&Finish"=
IF THEN ELSE(Lift C Up>0:AND:Lift C Up<=1, 1/59 , 0 )
Units: labor/Minute
(115) Lift C Up= INTEG (
+Lift C Going Up-"Lift C Starting Lamp&Finish",
0)
Units: labor
(116) Lift C Utilization=
100-Lift C Idle Percentage
Units: **undefined**
239

(117) Lift C Waiting= INTEG (
+Lift C Going Down-Lift C Starting 1st Half Restock,
1)
Units: labor
(118) Lit Fix Assembled= INTEG (
+Lit Fix Assembling Rate-"1st Half Lit Fix Restocking Rate",
15)
Units: light fixture
(119) Lit Fix Assembling Rate=
IF THEN ELSE( Lit Fix Waiting>=1/9:AND:Apprentice Waiting>=1/9, 1/9 , 0 )
Units: light fixture/Minute
(120) Lit Fix Going Up Rate=
IF THEN ELSE(Full Lit Fix Restocked>=15, 15 , 0 )
Units: light fixture/Minute
(121) Lit Fix Up= INTEG (
+Lit Fix Going Up Rate-Partial Lit Fix Installation Rate,
0)
Units: light fixture
(122) Lit Fix Waiting= INTEG (
-Lit Fix Assembling Rate,
330)
Units: light fixture
(123) Number of Apprentice=
1
Units: labor
(124) Number of Journeyman=
3
Units: labor
(125) Number of Lift=
3
Units: labor
(126) Operation Production Rate=
ZIDZ(Full Lit Fix In Place, Time of Operation )
Units: **undefined**
(127) Partial Lit Fix Installation Rate=
240

IF THEN ELSE(Lit Fix Up>0:AND:Lit Fix Up<=15, 1/3 , 0 )
Units: light fixture/Minute
(128) Partial Lit Fix Installed= INTEG (
+Partial Lit Fix Installation Rate-Complete Lit Fix Installation Rate,
0)
Units: light fixture
(129) "Partial Lit Fix Lamp&Finish Rate"=
IF THEN ELSE("Full Lit Fix Wire&Clipped">0:AND:Journeyman C
Up>0:AND:Lift C Up
>0, 1/3 , 0 )
Units: light fixture/Minute
(130) "Partial Lit Fix Lamp&Finished"= INTEG (
+"Partial Lit Fix Lamp&Finish Rate"-"Complete Lit Fix Lamp&Finish Rate",
0)
Units: light fixture
(131) "Partial Lit Fix Wire&Clip Rate"=
IF THEN ELSE(Full Lit Fix Installed>0:AND:Journeyman B Up>0:AND:Lift B
Up
>0, 1/7 , 0 )
Units: light fixture/Minute
(132) "Partial Lit Fix Wire&Clipped"= INTEG (
+"Partial Lit Fix Wire&Clip Rate"-"Complete Lit Fix Wire&Clip Rate",
0)
Units: light fixture
(133) SAVEPER =
TIME STEP
Units: Minute [0,?]
The frequency with which output is stored.
(134) Time of Operation=
Time/60
Units: hours
(135) TIME STEP = 1
Units: Minute [0,?]
The time step for the simulation.
(136) Total Apprentice Cost per hour=
Number of Apprentice*Unit Apprentice Cost
Units: dollars/hour
241

(137) Total Cost of Operation=
Time of Operation*(Total Apprentice Cost per hour+Total Journeyman Cost per
hour
+Total Lift Cost per hour)
Units: dollars
(138) Total Journeyman Cost per hour=
Number of Journeyman*Unit Journeyman Cost
Units: dollars/hour
(139) Total Lift Cost per hour=
Number of Lift*Unit Lift Cost
Units: dollars/hour
(140) Unit Apprentice Cost=
6
Units: dollars/(labor*hour)
(141) Unit Cost of Operation=
ZIDZ(Total Cost of Operation, Full Lit Fix In Place )
Units: dollars/light fixture
(142) Unit Journeyman Cost=
13.4
Units: dollars/(labor*hour)
(143) Unit Lift Cost=
1.5
Units: dollars/(labor*hour)
The complete system dynamics model for the “Improved” case is uploaded online and can be
accessed at the following website:
https://docs.google.com/open?id=0BxrMYHbElFuNSHJiZXozTDU2eUk

Appendix F: SD “Lean” Crew Model Formulation
The mathematical equations used to formulate the system dynamics “Lean” model are shown
below:
Model mathematical equations:
(01)

Apprentice Assembling=
242

IF THEN ELSE( Apprentice Waiting>=1/9:AND:Lit Fix Waiting>=1/9 , 1/9 , 0
)
Units: (labor/Minute)
(02)

Apprentice Finished Assembling= INTEG (
+Apprentice Assembling-Apprentice Returning,
0)
Units: labor

(03)

Apprentice Idle Percentage=
IF THEN ELSE(Apprentice Waiting=1:OR:Apprentice Waiting=1/9, Apprentice
Waiting
*100 , 0 )
Units: **undefined**
(04)

Apprentice Returning=
IF THEN ELSE(Apprentice Finished Assembling=1, 1 , 0 )
Units: labor/Minute

(05)

Apprentice Utilization=
100-Apprentice Idle Percentage
Units: **undefined**

(06)

Apprentice Waiting= INTEG (
+Apprentice Returning-Apprentice Assembling ,
1)
Units: labor

(07)

Assembling Production Rate=
ZIDZ(Lit Fix Assembled, Time of Operation )
Units: **undefined**

(08)

Complete Lit Fix Installation Rate=
IF THEN ELSE(Partial Lit Fix Installed>=1, 1 , 0 )
Units: light fixture/Minute

(09)

"Complete Lit Fix Lamp&Finish Rate"=
IF THEN ELSE("Partial Lit Fix Lamp&Finished">=1, 1 , 0 )
Units: light fixture/Minute

(10)

"Complete Lit Fix Wire&Clip Rate"=
IF THEN ELSE("Partial Lit Fix Wire&Clipped">=1, 1 , 0 )
Units: light fixture/Minute

(11)

FINAL TIME = 7900
Units: Minute
243

The final time for the simulation.
(12)

Full Lit Fix Installed= INTEG (
+Complete Lit Fix Installation Rate-"Partial Lit Fix Wire&Clip Rate",
0)
Units: light fixture

(13)

"Full Lit Fix Wire&Clipped"= INTEG (
+"Complete Lit Fix Wire&Clip Rate"-"Partial Lit Fix Lamp&Finish Rate",
0)
Units: light fixture

(14)

INITIAL TIME = 0
Units: Minute
The initial time for the simulation.

(15)

Journeyman Completing Installation=
IF THEN ELSE(Journeyman Finished Partial Installation>=1, 1 , 0 )
Units: labor/Minute

(16)

"Journeyman Completing Lamp&Finish"=
IF THEN ELSE("Journeyman Finished Partial Lamp&Finish">=1, 1 , 0 )
Units: labor/Minute

(17)

"Journeyman Completing Wire&Clip"=
IF THEN ELSE("Journeyman Finished Partial Wire&Clip">=1, 1 , 0 )
Units: labor/Minute

(18)

Journeyman Finished Full Installation= INTEG (
+Journeyman Completing Installation-"Journeyman Starting Wire&Clip",
0)
Units: labor

(19)

"Journeyman Finished Full Lamp&Finish"= INTEG (
+"Journeyman Completing Lamp&Finish"-Journeyman Going Down,
0)
Units: labor

(20)

"Journeyman Finished Full Wire&Clip"= INTEG (
+"Journeyman Completing Wire&Clip"-"Journeyman Starting Lamp&Finish",
0)
Units: labor

(21)

Journeyman Finished Partial Installation= INTEG (
+Journeyman Starting Installation-Journeyman Completing Installation,
0)
244

Units: labor
(22)

"Journeyman Finished Partial Lamp&Finish"= INTEG (
+"Journeyman Starting Lamp&Finish"-"Journeyman Completing Lamp&Finish",
0)
Units: labor

(23)

"Journeyman Finished Partial Wire&Clip"= INTEG (
+"Journeyman Starting Wire&Clip"-"Journeyman Completing Wire&Clip",
0)
Units: labor

(24)

Journeyman Finished Restocking= INTEG (
+Journeyman Restocking-Journeyman Going Up,
0)
Units: labor

(25)

Journeyman Going Down=
IF THEN ELSE("Journeyman Finished Full Lamp&Finish">=1:AND:Lit Fix In

Place
>=1, 1 , 0 )
Units: labor/Minute
(26)

Journeyman Going Up=
IF THEN ELSE(Journeyman Finished Restocking>0:AND:Journeyman Finished
Restocking
<=1, 1 , 0 )
Units: labor/Minute
(27)

Journeyman Idle Percentage=
IF THEN ELSE(Journeyman Waiting=1, Journeyman Waiting*100 , 0 )
Units: **undefined**

(28)

Journeyman Restocking=
IF THEN ELSE(Journeyman Waiting>=1:AND:Lit Fix Assembled>=1, 1 , 0 )
Units: labor/Minute

(29)

Journeyman Starting Installation=
IF THEN ELSE(Journeyman Up>0:AND:Journeyman Up<=1, 1/6 , 0 )
Units: labor/Minute

(30)

"Journeyman Starting Lamp&Finish"=
IF THEN ELSE("Journeyman Finished Full Wire&Clip">0:AND:"Journeyman
Finished Full Wire&Clip"
<=1, 1/6 , 0 )
Units: labor/Minute
245

(31)

"Journeyman Starting Wire&Clip"=
IF THEN ELSE(Journeyman Finished Full Installation>0:AND:Journeyman
Finished Full Installation
<=1, 1/8 , 0 )
Units: labor/Minute
(32)

Journeyman Up= INTEG (
+Journeyman Going Up-Journeyman Starting Installation,
0)
Units: labor

(33)

Journeyman Utilization=
100-Journeyman Idle Percentage
Units: **undefined**

(34)

Journeyman Waiting= INTEG (
+Journeyman Going Down-Journeyman Restocking,
1)
Units: labor

(35)

Lift Completing Installation=
IF THEN ELSE(Lift Finished Partial Installation>=1, 1 , 0 )
Units: labor/Minute

(36)

"Lift Completing Lamp&Finish"=
IF THEN ELSE("Lift Finished Partial Lamp&Finish">=1, 1 , 0 )
Units: labor/Minute

(37)

"Lift Completing Wire&Clip"=
IF THEN ELSE("Lift Finished Partial Wire&Clip">=1, 1 , 0 )
Units: labor/Minute

(38)

Lift Finished Full Installation= INTEG (
+Lift Completing Installation-"Lift Starting Wire&Clip",
0)
Units: labor

(39)

"Lift Finished Full Lamp&Finish"= INTEG (
+"Lift Completing Lamp&Finish"-Lift Going Down,
0)
Units: labor

(40)

"Lift Finished Full Wire&Clip"= INTEG (
+"Lift Completing Wire&Clip"-"Lift Starting Lamp&Finish",
0)
246

Units: labor
(41)

Lift Finished Partial Installation= INTEG (
+Lift Starting Installation-Lift Completing Installation,
0)
Units: labor

(42)

"Lift Finished Partial Lamp&Finish"= INTEG (
+"Lift Starting Lamp&Finish"-"Lift Completing Lamp&Finish",
0)
Units: labor

(43)

"Lift Finished Partial Wire&Clip"= INTEG (
+"Lift Starting Wire&Clip"-"Lift Completing Wire&Clip",
0)
Units: labor

(44)

Lift Finished Restocking= INTEG (
+Lift Restocking-Lift Going Up,
0)
Units: labor

(45)

Lift Going Down=
IF THEN ELSE("Lift Finished Full Lamp&Finish">=1:AND:Lit Fix In Place>=1,
1,0)
Units: labor/Minute

(46)

Lift Going Up=
IF THEN ELSE(Lift Finished Restocking>0:AND:Lift Finished Restocking<=1,
1,0)
Units: labor/Minute

(47)

Lift Idle Percentage=
IF THEN ELSE(Lift Waiting<=1/9:OR:Lift Waiting>=8/9, Lift Waiting*100 , 0
)
Units: **undefined**

(48)

Lift Restocking=
IF THEN ELSE(Lift Waiting>=1:AND:Lit Fix Assembled>=1, 1 , 0 )
Units: labor/Minute

(49)

Lift Starting Installation=
IF THEN ELSE(Lift Up>0:AND:Lift Up<=1, 1/6 , 0 )
Units: labor/Minute

247

(50)

"Lift Starting Lamp&Finish"=
IF THEN ELSE("Lift Finished Full Wire&Clip">0:AND:"Lift Finished Full
Wire&Clip"
<=1, 1/6 , 0 )
Units: labor/Minute
(51)

"Lift Starting Wire&Clip"=
IF THEN ELSE(Lift Finished Full Installation>0:AND:Lift Finished Full
Installation
<=1, 1/8 , 0 )
Units: labor/Minute
(52)

Lift Up= INTEG (
+Lift Going Up-Lift Starting Installation,
0)
Units: labor

(53)

Lift Utilization=
100-Lift Idle Percentage
Units: **undefined**

(54)

Lift Waiting= INTEG (
+Lift Going Down-Lift Restocking,
1)
Units: labor

(55)

Lit Fix Assembled= INTEG (
+Lit Fix Assembling Rate-Lit Fix Restocking Rate,
0.0001)
Units: light fixture

(56)

Lit Fix Assembling Rate=
IF THEN ELSE( Lit Fix Waiting>=1/9:AND:Apprentice Waiting>=1/9, 1/9 , 0 )
Units: light fixture/Minute

(57)

Lit Fix Going Up Rate=
IF THEN ELSE(Lit Fix Restocked>0:AND:Lit Fix Restocked<=1, 1 , 0 )
Units: light fixture/Minute

(58)

Lit Fix In Place= INTEG (
+"Complete Lit Fix Lamp&Finish Rate",
0.1111)
Units: light fixture

(59)

Lit Fix Restocked= INTEG (
+Lit Fix Restocking Rate-Lit Fix Going Up Rate,
248

0)
Units: light fixture
(60)

Lit Fix Restocking Rate=
IF THEN ELSE(Lit Fix Assembled>=1:AND:Lift Waiting>=1:AND:Journeyman
Waiting
>=1, 1 , 0 )
Units: light fixture/Minute
(61)

Lit Fix Up= INTEG (
+Lit Fix Going Up Rate-Partial Lit Fix Installation Rate,
0)
Units: light fixture

(62)

Lit Fix Waiting= INTEG (
-Lit Fix Assembling Rate,
330)
Units: light fixture

(63)

Number of Apprentice=
1
Units: labor

(64)

Number of Journeyman=
1
Units: labor

(65)

Number of Lift=
1
Units: labor

(66)

Operation Production Rate=
ZIDZ(Lit Fix In Place, Time of Operation )
Units: **undefined**

(67)

Partial Lit Fix Installation Rate=
IF THEN ELSE(Lit Fix Up>0:AND:Lit Fix Up<=1, 1/6 , 0 )
Units: light fixture/Minute

(68)

Partial Lit Fix Installed= INTEG (
+Partial Lit Fix Installation Rate-Complete Lit Fix Installation Rate,
0)
Units: light fixture

(69)

"Partial Lit Fix Lamp&Finish Rate"=

249

IF THEN ELSE("Full
Wire&Clipped"
<=1, 1/6 , 0 )
Units: light fixture/Minute

Lit

Fix

Wire&Clipped">0:AND:"Full

Lit

Fix

(70)

"Partial Lit Fix Lamp&Finished"= INTEG (
+"Partial Lit Fix Lamp&Finish Rate"-"Complete Lit Fix Lamp&Finish Rate",
0)
Units: light fixture

(71)

"Partial Lit Fix Wire&Clip Rate"=
IF THEN ELSE(Full Lit Fix Installed>0:AND:Full Lit Fix Installed<=1, 1/8
,0)
Units: light fixture/Minute

(72)

"Partial Lit Fix Wire&Clipped"= INTEG (
+"Partial Lit Fix Wire&Clip Rate"-"Complete Lit Fix Wire&Clip Rate",
0)
Units: light fixture

(73)

SAVEPER =
TIME STEP
Units: Minute [0,?]
The frequency with which output is stored.

(74)

Time of Operation=
Time/60
Units: hours

(75)

TIME STEP = 1
Units: Minute [0,?]
The time step for the simulation.

(76)

Total Apprentice Cost per hour=
Number of Apprentice*Unit Apprentice Cost
Units: dollars/hour

(77)

Total Cost of Operation=
Time of Operation*(Total Apprentice Cost per hour+Total Journeyman Cost per

hour
+Total Lift Cost per hour)
Units: dollars
(78)

Total Journeyman Cost per hour=
Number of Journeyman*Unit Journeyman Cost
Units: dollars/hour
250

(79)

Total Lift Cost per hour=
Number of Lift*Unit Lift Cost
Units: dollars/hour

(80)

Unit Apprentice Cost=
6
Units: dollars/(labor*hour)

(81)

Unit Cost of Operation=
Total Cost of Operation/Lit Fix In Place
Units: dollars/light fixture

(82)

Unit Journeyman Cost=
13.4
Units: dollars/(labor*hour)

(83)

Unit Lift Cost=
1.5
Units: dollars/(labor*hour)

The complete system dynamics model for the “Lean” case is uploaded online and can be
accessed at the following website:
https://docs.google.com/file/d/0BxrMYHbElFuNbzgtU3g2cjJ2dFE/edit

251

BIBLIOGRAPHY

252

BIBLIOGRAPHY

Abdelhamid, T. (2004, August). The self-destruction and renewal of lean construction theory: a
prediction from Boyd’s Theory’. In Proceedings of the 12th Annual Conference of the
International Group for Lean Construction (IGLC12), Helsingore, Denmark.
Abdelhamid, TS, & Everett, J. (2002). Physical demands of construction work: A source of
workflow unreliability (pp. 6-8). Presented at the Proceedings of the 10 th Conference of
the International Group for Lean Construction.
Abdelhamid, T.S., El-Gafy, M., and Salem, O. (2008). “Lean Construction: Fundamentals And
Principles.” American Professional Constructor Journal.
AbouRizk, S. M., & Hajjar, D. (1998). A framework for applying simulation in construction.
Canadian Journal of Civil Engineering, 25, 604-617.
Abu-Taieh, E., & Rahman El Sheikh, A. (2008). Methodologies and Approaches in Discrete
Event Simulation.
Adrian, J. J., & Boyer, L. R. T. (1976). Modeling method productivity. Journal of the
Construction Division, 102(1), 157-168.
Ahuja, H. N., & Nandakumar, V. (1985). “Simulation model to forecast project completion
time.” Journal of construction engineering and management, 111, 325.
Alexander, M. J. (1974). Information systems analysis: theory and applications, Science
Research Associates.
Alderfer, C. P. (1977). Organization development. Annual Review of Psychology, 28(1), 197223.
Antoniadis, D.N., Edum-Fotwe, F.T., Thorpe, A., and McCaffer, R. (2008). “Exploring
Complexity in Construction Projects.” 4th SCPM & 1st IPMA/MedNet Conference
Chios Island, Greece.
Argote, L., & McGrath, J. E. (1993). Group processes in organizations: Continuity and change.
International review of industrial and organizational psychology, 8(1993), 333-389.
Baccarini, D. (1996). “The concept of project complexity – a review.” International Journal of
Project Management, 14(4), 201-204.
Balbontin-Bravo, E. 1998. Simulation of large precast operations. Proceedings of 1998 Winter
Simulation Conference, ed. D. J. Medeiros and E. F. Watson, Washington, D.C., pp.
1311-1317.

253

Balci, O. (1988). The implementation of four conceptual frameworks for simulation modeling in
high-level languages. [Systems Research Center], Department of Computer Science,
Virginia Polytechnic Institute and State University.
Ballard, G., & Howell, G. (1994). Implementing lean construction: stabilizing work flow. Lean
construction, 101-110.Ballard, Glenn. (1999). Work Structuring. LCI White Paper, 5.
Retrieved from http://www.leanconstruction.org/pdf/WP4-OperationsDesign.pdf
Ballard, G., and Howell, G. (1994a). “Implementing Lean Construction: Stabilizing Work Flow.”
Proceedings of the 2nd Annual Meeting of the International Group for Lean
Construction, Santiago, Chile.
Ballard, G., and Howell, G. (1994b). “Implementing Lean Construction: Improving Performance
Behind the Shield.” Proceedings of the 2nd Annual Meeting of the International Group
for Lean Construction, Santiago, Chile.
Ballard, G. and Howell, G.A. (1997). “Shielding Production: An Essential Step in Production
Control.” J. of Constr. Engineering and Mgmt., ASCE, New York, NY, 124(1), 11-17.
Ballard, G. (1999). Work Structuring. LCI White Paper,
http://www.leanconstruction.org/pdf/WP4-OperationsDesign.pdf

5.

Retrieved

from

Ballard, Glenn, & Tommelein, I. (1999). Aiming for Continuous Flow. LCI White Paper, 3.
Retrieved from http://www.leanconstruction.org/pdf/WP3-ContinuousFlow.pdf
Ballard, Glenn. (2000a). Phase Scheduling. LCI White Paper, 7. Retrieved from
http://www.leanconstruction.org/pdf/WP7-PhaseScheduling.PDF
Ballard, Glenn. (2000b). Lean Project Delivery System. LCI White Paper, 8. Retrieved from
http://www.leanconstruction.org/pdf/WP_8_LPDS.pdf
Ballard, Glenn, Koskela, L., Howell, G., & Zabelle, T. R. (2001). Production System Design:
Work Structuring Revisited. LCI White Paper, 11. Retrieved from
http://www.leanconstruction.org/pdf/WP_11_Work_Structuring.pdf
Ballard, G., and Howell, G. A. (2003). “Competing Construction Management Paradigms.”
Proceedings of the 2003 ASCE Construction Research Congress, 19-21 March 2003,
Honolulu, Hawaii.
Banker, R., Field, J., Schroeder, R., & Sinha, K. (1996). Impact of work teams on manufacturing
performance: A longitudinal field study. Academy of Management Journal, 39(4), 867890.
Banks, J., & Carson, J. S. (1984). Discrete event system simulation.
Banks, J. (1999). Introduction to simulation. In Simulation Conference Proceedings, 1999 Winter
(Vol. 1, pp. 7-13). IEEE.

254

Banks, J., J. S. Carson, B. L. Nelson, and D. M. Nicol. (2000). Discrete-Event System
Simulation. 3rd ed. Upper Saddle River, New Jersey: Prentice-Hall, Inc.
Barnes, R. M. (1980). Motion and time study: Design and measurement of work. Seventh Ed.,
John Wiley-Interscience, New York, N.Y.
Bayer, S., & Gann, D. (2006). Balancing work: bidding strategies and workload dynamics in a
project based professional service organisation. System Dynamics Review, 22(3), 185211.
Bellinger, Gene. “Systems Thinking: An Operational Perspective of the Universe.” ©2004. 26
Sep. 2005. <http://www.systems-thinking.org/>.
Bennett, J. (1991). “International Construction Project Management: General Theory and
Practice.” Butterworth-Heinemann, Oxford.
Bérard, C. (2010). Group model building using system dynamics: an analysis of methodological
frameworks. Electronic Journal of Business Research Methods, 8(1), 35.
Bertelsen, S. (2003a). “Complexity – Construction in a new Perspective”. Accepted for the
proceedings of IGLC 11.
Bertelsen, S. (2003b). “Construction as a Complex System”. Accepted for the proceedings of
IGLC 11.
Bertelsen, S. (2003). Complexity–Construction in a new Perspective. IGLC-11, Blacksburg,
Virginia.
Borshchev, A., & Filippov, A. (2004). From system dynamics and discrete event to practical
agent based modeling: reasons, techniques, tools. In Proceedings of the 22nd
International Conference of the System Dynamics Society (No. 22).
Bostrom, R. P., & Heinen, J. S. (1977a). MIS problems and failures: a socio-technical
perspective. Part I: The causes. MIS quarterly, 17-32.
Bostrom, R. P., & Heinen, J. S. (1977). MIS problems and failures: a socio-technical perspective,
part II: the application of socio-technical theory. MIS quarterly, 11-28.
Brailsford, S. C., and N. A. Hilton. "A comparison of discrete event simulation and system
dynamics for modelling health care systems." (2001).
Callahan, K., & Brooks, L. (2004). Essentials of strategic project management. Hoboken, NJ:
John Wiley & Sons.
Campion, M. A. (1988). Interdisciplinary approaches to job design: A constructive replication
with extensions. Journal of Applied Psychology, 73(3), 467.

255

Campion, M. A., & McClelland, C. L. (1991). Interdisciplinary examination of the costs and
benefits of enlarged jobs: A job design quasiexperiment. Journal of Applied Psychology,
76, 186–198.
Campion, M. A., Medsker, G. J., & Higgs, A. C. (1993). Relations between work group
characteristics and effectiveness: Implications for designing effective work groups.
Personnel psychology, 46(4), 823-847.
Campion, M. A., Mumford, T. V., Morgeson, F. P., & Nahrgang, J. D. (2005). Work redesign:
Eight obstacles and opportunities. Human Resource Management-New York, 44(4), 367390.
Campion, M. A., & Thayer, P W. (1985). Development and field evaluation of an
interdisciplinary measure of job design. Journal of Applied Psychology, 70, 29–43.
Carson, J.S., 1993. "Modeling and Simulation World Views," in Proceedings of the 1993 Winter
Simulation Conference, eds., G.W. Evans, M. Mollaghasemi, E.C. Russell, and W.E.
Biles, Institute of Electrical and Electronics Engineers, pp. 18-23. 1-4, Piscataway, NJ.
Cherns, A. (1976). The Principles of Sociotechnical Design1. Human relations, 29(8), 783-792.
Cherns, A. (1987). Principles of sociotechnical design revisted. Human Relations, 40(3), 153161.
Clegg, C. (2000). Sociotechnical principles for system design. Applied ergonomics, 31(5), 463477.
Cohen, S., & Bailey, D. (1997). What makes teams work: Group effectiveness research from the
shop floor to the executive suite. Journal of management, 23(3), 239.
Cooper, K., & Lee, G. (2009). Managing the Dynamics of Projects and Changes at Fluor.
Cor, H. (1998). Using simulation to quantify the impacts of changes in construction work.
Virginia Polytechnic Institute and State University.
Cummings, T. G. (1978). Self-regulating work groups: A socio-technical synthesis. The
Academy of Management Review, 3(3), 625-634.
Cutcher Gershenfeld, J. (n.d.). Associates.(1994). Japanese team based work systems in North
America: Explaining the diversity. California Management Review, 37(1), 42-64.
Dai, Jiukun, Goodrum, P. M., & Maloney, W. F. (2009). Construction craft workers’ perceptions
of the factors affecting their productivity. Journal of Construction Engineering and
Management, 135, 217.
Dai, Jiukun, Goodrum, P. M., Maloney, W. F., & Srinivasan, C. (2009). Latent Structures of the
Factors Affecting Construction Labor Productivity. Journal of Construction Engineering
and Management, 135, 397.
256

Damron, A. J., de Las Casas, R., & Kuennen, S. T. (2004). Managing constructibility reviews to
reduce highway project durations. Journal of construction engineering and management,
130, 33.
Davis, L. E., & Wacker, G. L. (1982). Job design. In G. Salvendy (Ed.), Handbook of industrial
engineering (pp. 2.5.1–2.5.31). New York: Wiley
Davis, L. E., & Wacker, G. J. (1987). Job design.
Department of Defense Modeling and Simulation (M&S) Glossary", DoD 5000.59-M,
Department of Defense, 1998 [1]
DICTIONARY (2006) Collins English dictionary and thesaurus. Glasgow: Harper Collins.
Diekmann, J., and Heinz, J. _2001_. “Determinants of jobsite productivity.” Construction
Industry Institute Research Rep. No. 143–11, Univ. of Texas at Austin.
Dukkipati, R. V. (2007). Solving Engineering System Dynamics Problems with MATLAB. New
Age International.
Dyer, L. (1985). Strategic human resources management and planning. Research in personnel
and human resource management, 3(1), 30.
El-Gafy, M. A., & Abdelhamid, T. (2008). Using system Dynamics Modeling as a Lean Construction Work
Structuring Tool. Proceedings IGLC-16.

El-Gafy M. and Abdelhamid T. (2010) “Lean Work Structuring Using System Dynamic
Modeling”, NICMAR Journal of Construction Management, Vol. XXV, No. III, pp.2227. (ISSN: 0970-3675).
Emery, F. E. (1959). Characteristics of socio-technical systems. In Emery, F. (ed.), The
Emergence of a New Paradigm of Work. (Centre for Continuing Education, Australian
National University, Canberra, 1978, pp. 38–86.)
Emery, F. E. (1978). Analytical model for sociotechnical systems.
Emery, F. E., & Trist, E. L. (1972). Towards a social ecology: Contextual appreciations of the
future in the present. London: Plenum Press. Engel, G. L.
Fathi, H., Yomo, H., Kuijpers, G., & Christiansen, H. (2003). CNTK Deliverable: Simulator
requirements-a comparative evaluation of tools. CNTK Project.
Ford, F. A. (1999). Modeling the environment: An introduction to system dynamics models of
environmental systems. Island Pr.
Forrester, Jay W. (1961): Industrial Dynamics, Cambridge.
Forrester, J. (1965). “Industrial Dynamics.” MIT Press, 4th Ed., MA, USA

257

Forrester, J. W. (1971). World dynamics (Vol. 59). Wright-Allen Press Cambridge, Mass.;
Forrester, J. W. (1995). The beginning of system dynamics. McKinsey Quarterly, 4-17.
Forrester, J. W., & Collins, J. F. (1969). Urban dynamics. MIT Press.
Gidado, K. (1996). “Project complexity: The focal point of construction production planning.”
Construction Management and Economics. 14, 213-225.
Gidado, K. (2004). “Enhancing the prime contractors pre construction planning.” Journal of
Construction Research. 5, 87-106.
Gladstein, D. (1984). Groups in context: A model of task group effectiveness. Administrative
Science Quarterly, 29(4), 499-517.
Gleeson, F. and Townend J. (2007). "Lean construction in the corporate world of the U.K.
construction industry", University of Manchester, School of Mechanical, Aerospace,
Civil and Construction Engineering.
Goldratt, E.M., 1997. Critical Chain. North River Press, Great Barrington, MA
Guzzo, R. A., & Shea, G. P. (1992). Group performance and intergroup relations in
organizations. Handbook of industrial and organizational psychology, 3, 269-313.
Guzzo, R., & Dickson, M. (1996). Teams in organizations: Recent research on performance and
effectiveness. Annual review of psychology, 47.
Hackman, J. (1980). Work redesign and motivation. Professional Psychology, 11(3), 445-455.
Hackman, J. (1987). The design of work teams. Handbook of organizational behavior, 315, 342.
Hackman, J., & Oldham, G. (1980). Work redesign. Prentice Hall.
Hackman, J. (1990). Groups that work (and those that don’t): Creating conditions for effective
teamwork. Jossey-Bass San Francisco, CA.
Hackman, J. R. (1992). Group influences on individuals in organizations.
Hackman, J. R. (2002). Leading teams: Setting the stage for great performances. Boston:
Harvard Business School Press.
Hackman, J. R., & Lawler, E. E. (1971). Employee reactions to job characteristics. Journal of
applied psychology, 55(3), 259.
Hackman. J. R.. & Morris. C. G. (1975). Group tasks, group interaction process, and group
performance effectiveness: A review and proposed integration. In L. Berkowitz (Ed.),
Advances in experimental social psychology. New York: Academic Press. Vol. 8.

258

Hackman, J. R., & Oldham, G. R. (1976). Motivation through the design of work: Test of a
theory. Organizational behavior and human performance, 16(2), 250-279.
Hajjar, D., AbouRizk, S., (1996) Building a Special Purpose Simulation Tool for Earth Moving
Operations, Proceedings of the 1996 Winter Simulation Conference, pp. 1313-1320
Hajjar, D., and AbouRizk, S. (1998). ‘‘Modeling and analysis of aggregate production
operations.’’ J. Constr. Eng. Manage., 124(5), 390–401.
Hajjar, D., & AbouRizk, S. (1999). Simphony: an environment for building special purpose
construction simulation tools (pp. 998-1006). Presented at the wsc, IEEE.
Halpin, D. W. (1973). “An investigation of the use of simulation networks for modeling
construction operations.” Ph.D. Dissertation of University of Illinois at UrbanaChampaign, IL, U.S.A.
Halpin, DW, & Riggs, L. (1992). Planning and analysis of construction operations. WileyInterscience.
Halpin, D. W., and Woodhead, R. (1976). Design of construction and process operations. Wiley,
New York.
Halpin, D. W., Jen, H., and Kim, J. (2003). "A construction process simulation web service."
Proceedings of the 2003 Winter Simulation Conference, New Orleans, LA, 1503-1509.
Hanna, A. S., Russell, J. S., Gotzion, T. W., and Nordheim, E. V. (1999). “Impact of change
orders on labor efficiency for mechanical construction.” J. Constr. Eng. Manage.,
125_3_, 176–184.
Hassanein, A., & Melin, J. (1997). Construction crew design processes. International Journal of
Project Management, 15(2), 85-91.
Hassanein, A., and Melin, J. (1997). “Crew Design Methodology for Construction Contractors”
J. Constr. Engg. and Mgmt. Volume 123, Issue 3, pp. 203-207
Herbsman, Z., and Ellis, R. (1990). “Research of factors influencing construction productivity.”
Constr. Manage. Econom., 8_1_, 49–61.
Herbst, P. G. (1974). Socio-technical design. Tavistock Publications.
Horman, H. R., Thomas, M. J., Souza, U. E. L. de, & Zavrˇski, I. (2002). Reducing variability to
improve performance as a lean construction principle. Journal of Construction
Engineering and Management, 128, 144.
Homer, J. B., & Hirsch, G. B. (2006). System dynamics modeling for public health: background
and opportunities. American Journal of Public Health, AJPH-2005.
Howell, G. A. (1999). What is lean construction-1999. In Proceedings IGLC (Vol. 7, p. 1).
259

Howell, G.A. (1998). “The Dice Game.” Game played at 6th Ann. Conf. Intl. Group for Lean
Constr., IGLC-6, Guaruja, Brazil, July.
Howell, G., Ballard, G., & Alarcon, L. (1994). Implementing lean construction: reducing inflow
variation. Lean Construction.
Howell, G., Alexander, L., Ballard, G. (1993). “Interaction between subcycles –One key to
improved methods.” Journal of Construction Engineering and Management, ASCE, 119
(4) 714-728
Howell, G. and Ballard, G. (1999). “Design of Construction Operations” White Paper-4
(unpublished), Lean Constr. Inst., Ketchum, ID, http://www.leanconstruction.org/
Howell, G. A., & Koskela, L. J. (2000, July). Reforming project management: the role of lean
construction. In Proceedings of the 8th Annual Conference of the International Group for
Lean Construction.
Huckman, R. S., Staats, B. R., & Upton, D. M. (2009). Team familiarity, role experience, and
performance: Evidence from Indian Software Services. Management Sciences, 55, 85100.
Humphrey, S. E., Nahrgang, J. D., & Morgeson, F. P. (2007). Integrating motivational, social,
and contextual work design features: A meta-analytic summary and theoretical extension
of the work design literature. Journal of Applied Psychology, 92(5), 1332.
Ioannou, P., & Martinez, J. (1996a). Simulation of complex construction processes (p. 1328).
Presented at the Proceedings of the 28th conference on Winter simulation, IEEE
Computer Society.
Ioannou, P., & Martinez, J. (1996b). Comparison of construction alternatives using matched
simulation experiments. Journal of construction engineering and management, 122(3),
231-241.
Ioannou, P. G. (1989). UM-CYCLONE user’s guide. Dept. of Civil Engineering, The University
of Michigan, Ann Arbor, Mich.
Jain, R. (1991). The art of computer systems performance analysis (Vol. 182). John Wiley &
Sons New York.
Jones, R. E., & Deckro, R. F. (1993). The social psychology of project management conflict.
European journal of operational research, 64(2), 216-228.
Kalk, A., & Douglas, S. A. (1980). INSIGHT: Interactive simulation of construction operations
using graphical techniques (No. 238). Stanford University, Dept. of Civil Engineering.
Kartam, S. (1995), "Reengineering construction planning systems." Ph.D. Thesis, Department of
Civil and Environmental Engineering, University of California, Berkeley, CA.

260

Kartam, S., Ballard, G., and Ibbs, C. W., (1997), "Introducing a new concept and approach to
modeling construction," Journal of Construction Engineering and Management, 123(1),
89-97.
Kauffman, S.A. (1995). Antichaos and Adaption. Scientific American.
Kirkwood, C. (1998). “System Dynamics Methods - A quick introduction”.
Klimoski, R. J., & Jones, R. (1995). Suppose we took staffing for effective group decision
making seriously? Teams and Groups, 291-332.
Koskela, L, & Center for Integrated Facility Engineering. (1992). Application of the new
production philosophy to construction. Citeseer.
Koskela, L. (2000). “An exploration towards a production theory and its application to
construction.” VTT Publication 408, VTT Building Technology, Espoo, Finland,
http://www.inf.vtt.fi/pdf/publications/2000/P408.pdf
Koskela, Lauri and Howell, Greg (2001). Reforming project management: The role of planning,
execution and controlling. - Proceedings of the 9th International Group for Lean
Construction Conference. Kent Ridge Crescent, Singapore, 6 - 8 August 2001. Chua,
David and Ballard, Glenn (eds.). National University of Singapore (2001), pp. 185-198.
Koskela, Lauri, & Howell, G. (2002). The underlying theory of project management is obsolete
(pp. 293-302). Presented at the Proceedings of the PMI Research Conference, PMI.
Kozlowski, S. W., & Bell, B. S. (2003). Work groups and teams in organizations. Handbook of
psychology.
Kozlowski, S., & Ilgen, D. (2006). Enhancing the effectiveness of work groups and teams.
Psychological Science in the Public Interest, 7(3), 77.
Law, A. M., and Kelton, W. D. (2000). “Simulation modeling and analysis.” McGraw Hill, New
York, NY.
Law, A. M., & Kelton, W. D. (1991). Simulation modeling and analysis (Vol. 2). McGraw-Hill
New York.
Law, A. M. (2009, December). How to build valid and credible simulation models. In Simulation
Conference (WSC), Proceedings of the 2009 Winter (pp. 24-33). IEEE.
Leach, D. J., Wall, T. D., Rogelberg, S. G., & Jackson, P. R. (2005). Team autonomy,
performance, and member job strain: Uncovering the teamwork KSA link. Applied
Psychology, 54(1), 1-24.
Lee, S. H., Diekmann, J. E., Songer, A. D., & Brown, H. (1999). Identifying waste: applications
of construction process analysis. In Proceedings of the Seventh Annual Conference of the
International Group for Lean Construction (pp. 63-72).
261

Lee, Z. W., Ford, D. N., & Joglekar, N. (2004). Effects of resource allocation policies on project
durations (Vol. 25). Presented at the International Conference of the System Dynamics
Society.
Lee, S., Pena-Mora, F., and Park, M. (2005). "Quality and change management model for large
scale concurrent design and construction projects." Journal of Construction Engineering
and Management, 131(8), 890-902.
Lee, S., Pena-Mora, F., and Park, M. (2006a). "Dynamic planning and control methodology for
strategic and operational construction project management." Automation in Construction,
15(1), 84-97.
Liberda, M., Ruwanpura, J., and Jergeas, G. (2003). “Construction productivity improvement: A
study of human, management and external issues.” 2003 Construction Research
Congress, ASCE, Reston, Va.
Lichtig, W. A., Holland, M., & Allen, P. (2005). Ten Key Decisions to a Successful Construction
Project. Presented at the Proc. American Bar Association Forum on the Construction
Industry.
Liker, J. (2004). The Toyota way: 14 management principles from the world’s greatest
manufacturer. McGraw-Hill Professional.
Liou, F., and Borcherding, J. D. (1986). "Work sampling can predict unit rate productivity." J.
Constr. Engrg. and Mgmt., ASCE, 112(1), 90-103.
Liu, M., & Ballard, G. (2009, July). Factors affecting work flow reliability—A case study. In
Proc. of the 17th Annual Conference of the International Group for Lean Construction
(pp. 657-666).
Love, P., Holt, G., Shen, L., Li, H., & Irani, Z. (2002). Using systems dynamics to better
understand change and rework in construction project management systems. International
Journal of Project Management, 20(6), 425-436.
Madachy, R. J. (2008). Software process dynamics. Wiley-IEEE Press.
Maloney, W. (1983). Productivity improvement: The influence of labor. Journal of Construction
Engineering and Management, 109(3), 321-334.
Manz, C. C., & Sims Jr, H. P. (1987). Leading workers to lead themselves: The external
leadership of self-managing work teams. Administrative Science Quarterly, 106-129.
Martínez, J. C. (1996). STROBOSCOPE: State and resource based simulation of construction
processes. University of Michigan.
Martinez, J. C. (2001). EZStrobe: general-purpose simulation system based on activity cycle
diagrams. In Proceedings of the 33nd conference on Winter simulation (pp. 1556-1564).
IEEE Computer Society.
262

Martinez, J., & Ioannou, P. (1999). General-purpose systems for effective construction
simulation. Journal of Construction Engineering and Management, 125(4), 265-276.
McGrath, J. (1984). Groups: Interaction and performance (Vol. 14). Prentice-Hall Englewood
Cliffs, NJ.
McGrath, J. E., & Altman, I. (1966). Small group research: A synthesis and critique of the field.
ILLINOIS UNIV AT URBANA DEPT OF PSYCHOLOGY.
Medsker, G. J., & Campion, M. A. (2007). Job and team design. Handbook of Industrial
Engineering: Technology and Operations Management, Third Edition, 868-898.
Merry, U. (1995). “Coping with uncertainty - insights from the new sciences of chaos, self
organization and complexity.” Westport: Praeger publications.
Mills, A. (2001). “A systematic approach to risk management for construction.” Structural
survey. 19, 254-252.
Monden, Y. (1983). Toyota production system: practical approach to production management.
Norcross, GA: Industrial Engineering and Management Press, Institute of Industrial
Engineers.
Moore, D. 2002. Project management: Designing effective organizational structures in
construction, Blackwell Science, Oxford, U.K.
Morecroft, J. D., & Sterman, J. D. (2000). Modeling for learning organizations. Productivity
Press.
Morgan Jr, B. B., Salas, E., & Glickman, A. S. (1993). An analysis of team evolution and
maturation. The Journal of General Psychology, 120(3), 277-291.
Morgeson, F. P., & Campion, M. A. (1997). Social and cognitive sources of potential inaccuracy
in job analysis. Journal of Applied Physics, 82, 627-655.
Morgeson, F. P., & Campion, M. A. (2003). Work design. Handbook of psychology.
Morgeson, F. P., Delaney-Klinger, K., & Hemingway, M. A. (2005). The Importance of Job
Autonomy, Cognitive Ability, and Job-Related Skill for Predicting Role Breadth and Job
Performance. Journal of Applied Psychology, 90(2), 399.
Morgeson, F. P., Reider, M. H., & Campion, M. A. (2005). Selecting individuals in team
settings: The importance of social skills, personality characteristics, and teamwork
knowledge. Personnel Psychology, 58(3), 583-611.
Morgeson, F., Medsker, G., & Campion, M. (2006). Job and team design. Handbook of Human
Factors and Ergonomics (3rd Ed), Hoboken, NJ: John Wiley & Sons, 428-457.

263

Morgeson, F. P., & Humphrey, S. E. (2006). The Work Design Questionnaire (WDQ):
developing and validating a comprehensive measure for assessing job design and the
nature of work. Journal of Applied Psychology; Journal of Applied Psychology, 91(6),
1321.
Morgeson, F., & Humphrey, S. (2008). Job and team design: Toward a more integrative
conceptualization of work design.
Morgeson, F. P., & Humphrey, S. E. (2008a). Job and team design: Toward a more integrative
conceptualization of work design.
Morgeson, F. P., & Humphrey, S. E. (2008b). Job and team design: Toward a more integrative
conceptualization of work design.
Mumford, E. (2006). The story of socio technical design: reflections on its successes, failures
and potential. Information Systems Journal, 16(4), 317-342.
Nasirzadeh, F., Afshar, A., & Khanzadi, M. (2008). System dynamics approach for construction
risk analysis. International Journal of Civil Engineering, 6(2), 120-131.
Niepcel, W., & Molleman, E. (1998). Work design issues in lean production from a
sociotechnical systems perspective: Neo-Taylorism or the next step in sociotechnical
design? Human relations, 51(3), 259.
Nieva, V. F., Fleishman, E. A., & Rieck, A. (1985). Team dimensions: Their identity, their
measurement and their relationships. ADVANCED RESEARCH RESOURCES
ORGANIZATION BETHESDA MD.
Oldham, G. R. (1996). Job design. International review of industrial and organizational
psychology, 11, 33-60.
Oglesby, C., Parker, H., & Howell, G. (1989). Productivity improvement in construction.
McGraw-hill.
Ogungbamila, B., Ogungbamila, A., & Adetula, G. A. (2010). Effects of Team Size and Work
Team Perception on Workplace Commitment. Small Group Research, 41(6), 725-745.
Ogunlana, S. O., Li, H., & Sukhera, F. A. (2003). System dynamics approach to exploring
performance enhancement in a construction organization. Journal of construction
engineering and management, 129, 528.
Ono, T. (1988). Toyota production system: beyond large-scale production. Productivity Pr.
Page Jr, E. H. (1994). Simulation modeling methodology: principles and etiology of decision
support (Doctoral dissertation, Virginia Polytechnic Institute and State University).
Park, M. (2005). Model-based dynamic resource management for construction projects.
Automation in construction, 14(5), 585-598.
264

Park, M., & Peña Mora, F. (2003). Dynamic change management for construction: introducing
the change cycle into model based project management. System Dynamics Review,
19(3), 213-242.
Parker, H. W., & Oglesby, C. H. (1972). Methods improvement for construction managers (pp.
178-203). New York: McGraw-Hill.
Pasmore, W., Francis, C., Haldeman, J., & Shani, A. (1982). Sociotechnical systems: A North
American reflection on empirical studies of the seventies. Human relations, 35(12), 1179.
Pasmore, W. A., & Sherwood, J. J. (Eds.). (1978). Sociotechnical systems: A sourcebook. La
Jolla, CA: University Associates.
Pasmore, W. A. (1988). Designing effective organizations: The sociotechnical systems
perspective. Wiley.
Pidd M. (1998) Computer simulation in management science (4th edition). John Wiley & Sons
Ltd, Chichester
Pennanen, A., & Koskela, L. (2005). “Necessary and unnecessary complexity in construction.”
Presented at the Proceedings of First International Conference on Built Environment
Complexity, University of Liverpool.
Pena-Mora, F., Han, S., Lee, S., and Park, M. (2008). "Strategic-operational construction
management: A hybrid system dynamics and discrete event approach." Journal of
Construction Engineering and Management, 134(9), 701-710.
Pritsker, A, O’Reilly, J., LaVal, D. (1997). Simulation with Visual SLAM and AweSim”, John
Wiley & Sons, USA
Radzicki, M. J., & Taylor, R. A. (1997). Introduction to system dynamics: a systems approach to
understanding complex policy issues. US Department of Energy's.
Richardson, G. P., & Pugh III, A. I. (1981). Introduction to system dynamics modeling with
DYNAMO. Productivity Press Inc.
Robinson, S. (2004). Simulation: The Practice of Model Development and Use. Wiley,
Chichester, UK.
Rogge, D. F., and Tucker, R. L. (1982). "Foreman-delay surveys: Work sampling and output." J.
Constr. Division, ASCE, 108(4), 592-603.
Sanvido, V. (1988). Conceptual construction process model. Journal of Construction Engineering
and Management, 114(2), 294-312.
Sanvido, V, et al. (1990). An Integrated Building Process Model, Tech. Rep. 1, CIC Research
Program, Arch. Eng. Dept., Pennsylvania State U.
Sanvido, V. E., & Medeiros, D. J. (1990). Applying computer-integrated manufacturing concepts
to construction. Journal of construction engineering and management, 116(2), 365-379.

265

Sanvido, V. (1984).
Designing Productivity Management and Control Systems for
Construction Projects. Ph. D. Thesis, Department of Civil Engineering, Stanford
University, Stanford, California.
Saysel, A. K., Barlas, Y., & Yenigun, O. (2002). Environmental sustainability in an agricultural
development project: a system dynamics approach. Journal of Environmental
Management, 64(3), 247-260.
SENGE, P.M. (1990). The Fifth Discipline (New York, Currency Doubleday).
SENGE, P.M., KLEINER, A., ROBERTS, C., ROSS, R.B. & SMITH, B.J. (1994) The Fifth
Discipline Fieldbook (New York, Currency Doubleday).
Senghore, O., Hastak, M., Abdelhamid, T., AbuHammad, A., & Syal, M. (2004). Production
process for manufactured Housing. Journal of construction engineering and management,
130(5), 708-718.
Shannon, R. E. (1975). Systems simulation: the art and science (Vol. 975, pp. I80-l97).
Englewood Cliffs,, New Jersey: Prentice-Hall.
Shea, G.P., & Guzzo, R.A. (1987). Groups as Human Resources. In Rowland KM, Ferris GR
(Eds.), Research in hhuman resources and personnel management (Vol. 5, pp. 323-356).
Greenwich, CT: JAI Press.
Shingo, S., Shing, S., & Dillon, A. (1989). A study of the Toyota production system from an
industrial engineering viewpoint. Productivity Pr.
Skyttner, Lars (2006). General Systems Theory: Problems, Perspective, Practice. World
Scientific Publishing Company. ISBN 9-812-56467-5.
Smith, S., Osborne, J., & Forde, M. (1995). Analysis of earth-moving systems using discreteevent simulation. Journal of construction engineering and management, 121, 388.
Srisuwanrat, C., & Ioannou, P. (2007). The investigation of lead-time buffering under
uncertainty using simulation and cost optimization. Proceedings of the 15 th International
Group of Lean Construction.
Steiner, I. (1972). Group process and productivity. Academic Press New York.
Sterman, J. D. (1992). System dynamics modeling for project management. Unpublished
manuscript, Cambridge, MA.
Sterman, J. D. (2000). Business dynamics: systems thinking and modeling for a complex world
(Vol. 53). Irwin McGraw-Hill.
Stocks R K and Male S P (1984) “An investigation into the client's perceptions of contractual
form and procedures: the instigation of good practice,” Proceedings, CIB W-65
Organization and Management of Construction, Waterloo, Ontario, Canada 291-299.
266

Sundstrom, E., De Meuse, K., & Futrell, D. (1990). Work teams: Applications and effectiveness.
American psychologist, 45(2), 120-133.
Sundstrom, E. (1999). Supporting work team effectiveness: Best management practices for
fostering high performance. Jossey-Bass Publishers.
Sweetser, A. 1999. A Comparison of System Dynamics and Discrete Event Simulation. In
Proceedings of 17th International Conference of the System Dynamics Society and 5th
Australian & New Zealand Systems Conference.
Tocher, K.D. (1963). The Art of Simulation. London: English University Press.
Thomas, H. R., (1981). "Construction Work Sampling," Construction Management Research
Report No. 14, The Business Roundtable, Pennsylvania State Univ., Dept. of Civ. Engrg.,
University Park, Pa.
Thomas, H. R. (1992). Effects of scheduled overtime on labor productivity. Journal of
Construction Engineering and Management, 118(1), 60-76.
Thomas, H. R., and Daily, J., (1983). "Crew Performance Measurement Via Activity Sampling,"
Journal of the Construction Engineering and Management Division, Vol. 109, No. C03,
Paper 18218, Sept.
Thomas, H., Guevara, J., & Gustenhoven, C. (1984). "Improving productivity estimates by work
sampling. Journal of Construction Engineering and Management, 110(2), 178-188.
Thomas, H., & Yiakoumis, I. (1987). Factor model of construction productivity. Journal of
construction engineering and management, 113(4), 623-639.
Thomas, H., Horner, R., & Sanders, S. (1990). Modeling construction labor productivity. Journal
of construction engineering and management, 116, 705.
Thomas, H., & Sakarcan, A. (1994). Forecasting labor productivity using factor model. Journal
of construction engineering and management, 120(1), 228-239.
Thomas, H., & Napolitan, C. (1995). Quantitative effects of construction changes on labor
productivity. Journal of Construction Engineering and Management, 121, 290.
Thomas, H.R., & Raynar, K. A. (1997). Scheduled overtime and labor productivity: quantitative
analysis. Journal of construction engineering and management, 123, 181.
Thompson, James D. (1967). Organizations in Action. New York: McGraw-Hill.
Thorsrud, E. (1972). Job design in the wider context. Design of Jobs, 451-459.
Tommelein, ID, & Weissenberger, M. (1999). More just-in-time: location of buffers in structural
steel supply and construction processes (Vol. 7, pp. 26-28). Presented at the Proceedings
of IGLC, Citeseer.
267

Tommelein, ID, Riley, D., & Howell, G. (1999). Parade game: impact of work flow variability
on trade performance. Journal of Construction Engineering and Management, 125(5),
304-310.
Tommelein, I. D. (1998). Pull-driven scheduling for pipe-spool installation: simulation of lean
construction technique. Journal of construction engineering and management, 124(4),
279-288.
Trist, E. (1981). The evolution of socio-technical systems. Occasional paper, 2.
Trist, E. L., & Bamforth, K. W. (1951). Some Social and Psychological Consequences of the
Longwall Method.
Trist, E. L., and Emery, F. E. (1960). Report on the Barford conference for Bristol/Siddeley,
aeroengine corporation, July 10–16. TIHR, Document No. 598, Tavistock, London.
Wageman, R. (1995). Interdependence and group effectiveness. Administrative Science
Quarterly, 40(1).
Walker, A. (1985). Project Management in Construction. Collins P. and T. Books.
Walker, G. H., Stanton, N. A., Salmon, P. M., & Jenkins, D. P. (2008). A review of
sociotechnical systems theory: a classic concept for new command and control
paradigms. Theoretical Issues in Ergonomics Science, 9(6), 479-499.
Walton RE., (1972). How to counter alienation in the plant. Harvard Business Review, 50(6), 7081.
Watson, N. (2004). Background paper on socio-technical theory. Retrieved July 25, 2006 from
https://www.sonoma.edu/phpbb/viewtopic.php?p=332&sid=d04e94067bded6533e60723
2a556a433
Williams, T. (1999). “The need for new paradigms for complex projects.” International Journal
of Project Management, 17(5), 269-273.
Williams, T. (2002). “Modeling complex projects.” John Wiley & Sons, New York, NY.
Womack, J., Jones D. & Roos, D., (1990), “The Machine that Changed the World”, New York :
Macmillan, 1990
Womack, J. P., & Jones, D. T. (1996). Beyond Toyota: how to root out waste and pursue
perfection. Harvard business review, 74, 140-172.
Wood, H., and Gidado, K. (2008). “Project Complexity in Construction.” The Construction and
Building Research Conference of the Royal Institution of Chartered Surveyors, Dublin
Institute of Technology.

268

Zabelle, T. R., & Ballard, G. (1999). Flow Driver - A System For Reducing Fabricator Lead
Time. LCI White Paper, 6. Retrieved from http://www.leanconstruction.org/pdf/WP6FlowDriver.pdf
Zayed, T. M., & Halpin, D. (2001). Simulation of concrete batch plant production. Journal of
construction engineering and management, 127, 132.
Zayed, T., and Halpin, D. (2002). Simulation as a tool for resource management (Vol. 2, pp.
1897-1906). Presented at the Simulation Conference Proceedings, 2000. Winter, IEEE.
Zayed, T. M., & Halpin, D. W. (2004). Quantitative assessment for piles productivity factors.
Journal of construction engineering and management, 130(3), 405-414.
Zhao, W., Ren, H., & Rotter, V. S. (2011). A system dynamics model for evaluating the
alternative of type in construction and demolition waste recycling center - The case of
Chongqing, China. Resources, Conservation and Recycling, In Press, Corrected Proof.
doi:doi: 10.1016/j.resconrec.2011.04.011

269