“I -:"
"an:

 

 

MICHIGAN STATE UNIVERSITY

Q LIBRARIES
a " EAST LANSING, MICH 43824-1048
0’

r“? C C'f/O

This is to certify that the
thesis entitled

ASSESSMENT OF PRODUCTION PLANNING PROCESS IN
RESIDENTIAL CONSTRUCTION USING LEAN
CONSTRUCTION AND SIX SIGMA

presented by
THANVEER M MOHAMMED

has been accepted towards fulfillment
of the requirements for the

MASTER OF degree in CONSTRUCTION
SCIENCE MANAGEMENT

fl fi
’-
J

/ Major Professor’s Signature

 

04/20/05

Date

MSU is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN BOX
to remove this checkout from your record.

TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

 

‘ up , 3 DATE DUE
M7

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ASSESSMENT OF PRODUCTION PLANNING PROCESS IN RESIDENTIAL
CONSTRUCTION USING LEAN CONSTRUCTION AND SIX SIGMA

By

Thanveer Moosa Mohammed

A THESIS

Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of

MASTER OF SCIENCE

CONSTRUCTION MANAGEMENT PROGRAM
SCHOOL OF PLANNING, DESIGN AND CONSTRUCTION

2005

ABSTRACT

ASSESSMENT OF PRODUCTION PLANNING PROCESS IN RESIDENTIAL
CONSTRUCTION USING LEAN CONSTRUCTION AND SIX SIGMA

By
Thanveer Moosa Mohammed
The homebuilding industry in the United States is at its peak with the increase in demand
for homes. With the projected demand for homes in sight, many companies are preparing
themselves to be able to build more and more homes. Hence, it is vital for companies to
adopt the best practices of production planning and control. The research aims at
providing a production planning process performance assessment tool for the
homebuilding industry while exploring the possibilities of using a combination of Six
Sigma and Lean Construction principles to achieve the goal of zero-defect in a process.
Interviews with professionals from the homebuilding industry were conducted to study
the prevalent practices. Data was collected from a homebuilding site and a set of analysis
tools was used to analyze the data and find areas of improvement. An assessment tool and
a metric for performance assessment of production planning process were developed. A
production planning process model was developed based on this study. The model is a
combination of the Six Sigma DMAIC methodology and the Last Planner System. The
study shows that homebuilding companies generally perceive a higher production yield
than actual. Data analysis revealed that task pre-requisites, coordination with suppliers,

and overestimating productivity were major causes for unreliable production planning.

I wouldn’t have been producing this thesis if not for the support and
encouragement I got from my father. I dedicate this thesis to him.
I also dedicate this thesis to my mother who compromised her own happiness by
letting me away from her, so that] could achieve this.
I extend my dedication to my lovely sister and my brothers. And to my two little

nieces who always kept me in high spirits.

iii

ACKNOWLEDGEMENTS

I would like to express my sincere gratitude to my mentor and major advisor Dr.
Tariq Abdelhamid for his constant support and brilliant guidance. I thank him for being a
father figure to me for the past two years. He sincerely dedicated his time and effort to
help me accomplish my goal. Without his incessant encouragement and appreciation, this
thesis would not be ‘complete’ in its true sense.

I would also like to thank the members of my advisory committee, Dr. Matt Syal
and Dr. Jon Burley for offering their valuable time and guidance. Their insightful
comments and suggestion were significant for the successful completion of this thesis.

My special thanks to all the homebuilding company professionals I have
interacted with, for the purpose of this thesis for providing me with necessary data and
insightful expertise without which this work would not have been possible. I also thank
the people in the company who let me collect data on their site and who shared their ideas
and expertise to help me achieve my goal.

I would like to thank Pooja for her absolute support and inspiration throughout
my thesis. I also thank my friends VJ, Guru, Hyder, Abhi and Jimmy for all the good
time they shared with me and for making this a great experience. I also thank my seniors
Bhavin, Bhavani, and Afshan for guiding and helping me right from when I joined this
program.

Many thanks to all the professors and the support staff in the Construction
Management Program for all their assistance and support throughout my graduate

program.

iv

TABLE OF CONTENTS

LIST OF FIGURES viii
LIST OF TABLES ix
1. CHAPTER ONE: INTRODUCTION
1.1 Motivation .................................................................. 2
1.2 Need Statement .......................................................... 5
1.3 Research Goal and Objectives ....................................... 7
1.4 Research Scope ......................................................... 8
1.5 Repot Overview .......................................................... 9
2. CHAPTER TWO: LITERATURE REVIEW
2.1 Residential Construction ................................................ 11
2.1.1 Production Planning Process .............................. 11
2.1.2 Prior Research ................................................. 13
2.2 Lean Production System ................................................ 16
2.2.1 Lean Production? .............................................. 16
2.2.2 Lean Construction ............................................. 17
2.2.2.1 Even Flow Methodology .......................... 18
2.2.2.2 Last Planner System .............................. 19
2.3 Performance Measurement in Construction ........................ 23
2.3.1 Total Quality Management .................................. 25
2.3.2 Six Sigma ........................................................ 27
2.4 Summary .................................................................... 28

3. CHAPTER THREE: METHODOLOGY

3.1 Methods and Tools for Objectives ..................................... 31
3.2 Mapping the Production Planning Process .......................... 31
3.2.1 Interviewing ...................................................... 32
3.2.2 Field Visits ........................................................ 32
3.2.3 Creating a Process Map ...................................... 33
3.3 Production Planning Process Performance ......................... 33
3.3.1 Performance Measurement Tool ............................ 34
3.3.1.1 Percent of Plan Complete (PPC) ................ 34
3.3.1.2 The Six Sigma Metrics ............................. 35

3.3.1.3 Percent of Plan Complete (PPC) vs.
Rolled throughput yield (YRT) ............ 38
3.3.2 Measurement of Planning Performance .................. 39
3.3.3 Data Analysis .................................................... 40
3.3.3.1 Visual Tools for Data Analysis ................... 40
3.3.3.2 Root Cause Analysis ............................... 44
3.4 Summary ...................................................................... 46

4. CHAPTER FOUR: DATA COLLECTION AND ANALYSIS

4.1 Overview ..................................................................... 48

4.2 Interviews ................................................................... 48

4.2.1 Interview Response ........................................... 49

4.2.2 Production Planning Process ............................... 50

4.2.3 Summary ......................................................... 56

4.3 Data Collection on Site ................................................... 56

4.3.1 Site Overview .................................................... 56

4.3.2 Data Collection .................................................. 56

4.4 Data Analysis ............................................................... 59

4.4.1 PPC Calculations ...................................... 59

4.4.2 PC vs. Rolled PPC .................................... 61

4.4.3 Sigma Value ............................................ 64

4.4.4 Statistical Analysis of Data ......................... 65

4.4.5 Pareto Analysis ........................................ 70

4.5 Summary .................................................................... 73

4.6 Production Planning Process Model(PPPM) ........................ 74
4.6.1 Six Sigma Methodology for Process

Improvement ............................................ 74

4.6.2 Last Planner System .................................. 77

4.6.3 Production Planning Process Model Description.79

4.6.4 Validation ................................................ 83

4.7 Summary ................................................................ 85

5. CHAPTER FIVE: SUMMARY AND CONCLUSION

5.1 Overview of Report ........................................................ 87

5.2 Research Goals and Objectives ....................................... 87

5.3 Contributions ................................................................ 88

5.4 Research Limitations ...................................................... 90

5.5 Discussions .................................................................. 91

5.6 Research Conclusions ..................................................... 93

5.7 Recommendations for Future Research ............................. 94

5.8 Summary ..................................................................... 95

REFERENCES ........................................................................... 97

vi

APPENDICES

Appendix-A ........................................................................ 1 00
Appendix-B ......................................................................... 105
Appendix-C ......................................................................... 108
Appendix-D ......................................................................... 124
Appendix-E ......................................................................... 147
Appendix-F ......................................................................... 1 S4
Appendix-G ......................................................................... 1 S7

vii

LIST OF FIGURES

Figure 2.1: Construction Planning vs. Last Planner System ................. 21
Figure 2.2: Evolution of Residential Construction .............................. 24
Figure 3.1: Sample Format for Data Collection .................................. 39
Figure 3.2: Typical Pareto Chart ..................................................... 41
Figure 3.3: Format for p-chart Data ................................................. 43
Figure 3.4: Sample p-chart ............................................................. 44
Figure 3.5: General Structure of a Cause and Effect Diagram ............... 45
Figure 4.1: Flowchart Representing Production Planning for Small
Homebuilding Company .................................................. 52
Figure 4.2: Flowchart Representing Production Planning for Medium
Homebuilding Company .................................................. 53
Figure 4.3: Flowchart Representing Production Planning for Large
Homebuilding Company .................................................. 55
Figure 4.4: Format for Data Collection .............................................. 58
Figure 4.5: Spreadsheet Showing Daily PPC and Reasons .................. 58
Figure 4.6: Graph Showing Weekly PPC for Building l ......................... 60
Figure 4.7: Graph Showing Weekly PPC for Building ll ........................ 60
Figure 4.8: Graph of PPC and rolled PPC for Building | ........................ 64
Figure 4.9: Graph of PPC and rolled PPC for Building ll ....................... 64
Figure 4.10: p-chart for Tasks Incomplete for Building | .............. 67
Figure 4.11: p-chart for Tasks Incomplete for Building || ........................ 69
Figure 4.12: Pareto Chart for Reason Codes ..................................... 71
Figure 4.13: Pareto Chart for Trades ................................................ 72
Figure 4.14: Graph of PPC vs. Weekday ............................................ 73
Figure 4.15: Last Planner System Model ............................................ 78
Figure 4.16: Production Planning Process Model ................................ 8O

viii

LIST OF TABLES

Table 1.1: Projected Demand for Housing ........................................ 6

Table 4.1: Sample Data for a Week ................................................. 62
Table 4.2: Special Reasons Causing Variations in Building | ................. 68
Table 4.3: Special Reasons Causing Variations in Building II ................. 70

ix

Chapter 1

Introduction

Introduction

1.1 Motivation

Construction is as old as any civilization. Every civilization in this world considers
construction of structures as a symbol of their prosperity. Construction of structures has
been executed for shelter, transportation, defense, storage as well as for public
entertainment and meetings. In the past few centuries the world has seen amazing
structures like the Pyramids of Egypt, Taj Mahal in India, and Great Wall of China. It is
remarkable that these were built in an age when technology was way behind what we
have now.

These constructions projects had so many challenges to tackle; one can imagine
the logistics involved, the different materials requirements, specifications of the owners
(generally a king or a ruler), labor productivity, etc. How did they manage to achieve
such a task? Most of such projects didn’t have a time restriction and labor force was large
and money also was abundant. Yet, history of the construction of such structures does
show some sort of formal planning (Patrick, 2004). One can imagine the complexity of
the production planning process involved.

As civilizations grew, the variety and quality of construction improved (Allen and
Thallon, 2002). In the last century, technology had an exponential growth and this also
had an influence on construction. The construction systems became more and more
sophisticated, the construction materials improved, the labor force started to receive
formal training, and formal management techniques began to be recognized (Allen,
1999). The industry has recognized management has an important and distinctive

requirement of a successful project (Patrick, 2004). Today, the construction industry

follows traditional management techniques. The main functions of project management
are planning, organizing, staffing, coordinating, directing and controlling (Patrick, 2004).

The first of the many steps involved is Planning. This task basically means
establishing the course of project execution. Project planning involves outlining a course
of action, deciding on what is to be done, who will do it, how it will be done, when will it.
be done, and how much it will cost (Syal 2003). A good and reliable plan will exude
confidence in the project team. Highly complex projects have been executed successfully
with the aid of systematic and skilled planning. Flaming being the first step of the project
management functions, it decides the success of all the following functions. If planning is
not done meticulously, execution and control becomes difficult. The success of a project
is determined on how well it is executed in comparison with the plan. When plans change
due to unexpected events and uncertainties, the project team develops recovery plans to
stay within the broad objectives of the project as previously captured in initial plans.

The planning process differs across the different categories of construction. For
example in residential construction companies building few homes a year have less needs
for detailed planning compared to larger homebuilding companies. It is the author’s
opinion that a systematic planning process as well as a system that will measure the
effectiveness of the planning process is necessary to increase its production.

The typical construction process for small size homebuilding companies, is a very
ad-hoc approach where one builder finds a good site, decides to build a home, hires local
subcontractors and starts the process of construction. All the planning process is done
mentally and informally with many details left to chance. When this builder is building

about 3 to 5 homes simultaneously, everything is in chaos. This typically leads to

confusion, less productivity, monetary losses and delay. Medium companies have a more
systematic approach, wherein they have proper written agreements, qualified
subcontractors, planned production goals and deadlines to meet. Most of the larger
companies follow a system with bids, contracts, planned production, well-maintained
cash flows, performance analysis, quality considerations, safety and environmental
considerations, and requirements for customer satisfaction.

To successfully compete long term, a company must satisfy its customers by
providing superior quality which is rapidly becoming the criterion which differentiate
companies from their competitors (NAHB, 1993). With increasing demand for homes, it
is evident that more and more construction companies will participate in the competition.
Thus, it is clear that companies have to plan their operations to reach and maintain the
highest level of customer satisfaction.

Planning being the first step towards the successful execution of a project, it needs
to be as representative as possible (Hinze, 1998). Equally important is site production
planning done by homebuilding companies. Site production planning is the planning done
for day-to-day site activities. It is planning at a smaller level that includes all tasks
performed on site for completing the project. Present site production planning methods
used in the construction industry have evolved out of experience and knowledge. Site
production planning is necessary to breakdown the whole construction process into
smaller tasks done on a daily basis and achieve the production goals. Thus, site
production planning is critical for successful execution of tasks. Construction companies
need to assess the current practices and improvise on its techniques if needed. This could

be done by accepting and implementing some changes, even small ones. In order for

companies to identify areas needing improvement, 3 system of performance assessment is
the key.

Performance assessment is a very vital tool to improvise on any process (Evans
and Lindsay, 2001). Assessment means to evaluate how effective a process is. For this, a
framework has to be developed such that the process can be studied, its outcomes can be
measured and areas needing improvements can be identified. Such a framework will have
to be developed for all processes that considerably affect the execution of site
eoperations. Therefore this research developed an assessment framework for residential
construction site production planning process. This framework will allow homebuilders,
and possibly contractors from other sectors, to find improvement opportunities in the

production planning process.

1.2 Need Statement
In residential construction, a homebuilder needs a system of planning, to manage and
control the entire construction process. It is the author’s opinion that a good system of
planning will be one that can establish the firture course of actions correctly as to when
what work will be completed, how can the construction process finish on time, provisions
to tackle crisis, and at the same time produce a high quality product, in the most efficient
and profitable way. With the projected growth in demand for housing (Bureau of Census
2000), as shown in Table 1.1, the homebuilding industry needs a system, which will
address both the short term and long term goals and thus keep the entire process under
control and it should help in the process of continuous improvement.

Most homebuilders follow a system of planning where the whole process is

broken down to the activity level such that it can be monitored and controlled and each

activity is sequenced following technical, structural and regulatory considerations.

Sometimes, the duration and timing is decided such that they save on time without even

considering whether it is practically possible. Later, during the construction process when

things do not go as planned, crashing and other crisis management techniques are

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

followed.
YEAR TOTAL NUMBER - - J, -___ _ _,____.._ __ -_L_ _ _ _ _ _ A _
OFHOUSEHOLDS Prp ctodDemaio lb h
2000 105,480,101 }
l 118
2001 106,478,583 1 r-
1 116 _.__..--...-._....W.-__.m___.__-..______._wfl». L.
2002 107,690,202 1 "4 r--~----- ~---—~----~~~ —- -‘ -
l ._ 112 9- 1.4 .2, E. l-1 a
2003 108,800,265 ‘ 2 110 ._._.___-_--__.-_.- “Ln. L“ ._H ...- F” H
108 — ~—~ ——~ 1———~ _~ “.4 L... as
2004 109,937,037 1 " f ”Tm“
. 35106 -—1 1—- r—F ”-4 ~—~—1 ~— r—-—- ~—1 »—»< r-
2005 111,052,797 l g 104 _. ~ ,-~ ~ -~ L-a __ -~ —1
i 102 u — - ”fir r—H F—y F-~ R r-T —-—- m '1
2006 112,216,108 ‘ 100 __- .__. __. L... L... .4 p“ L, -
2007 l 13’396’397 i 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
2008 1 14,596,986 1 Y-
2009 1 15,802,105
2010 1 17,059,577

 

 

 

Table 1.] Projected demand for housing

(Source: Bureau of the census 1996 projections updated with 2000 census results)

 

According to Lauri Koskela (1999), a production planning and control system

should follow three principles. “The first principle is that the assignments should be

sound regarding the prerequisites... The second principle is that the realization of

assignments is measured and monitored... The third principle dictates that causes for

non-realization is investigated and those causes are removed. Thus, in fact, continuous,

in-process improvement is realized.”

The Last Planner System proposed by Ballard (2000) is a system of production
planning in which any assignment has to be well defined, sequenced in such a way that it
is constructible, should be sound, and should be “sized to the productive capability of the
crew”.

A careful consideration of the planning procedures suggested in prior and state-
of-the-art research raises the question of whether homebuilders plan their production
process accordingly? Few questions that arise are:

' Do the homebuilders have a systematic production planning method? If
yes, then do they measure and monitor the process effectively?

I Do homebuilders investigate the causes behind production plan failures?

I Are these causes eliminated and better methods applied in future projects?

These questions underscore the need for a detailed study on the production
planning system commonly used and its effectiveness in the residential construction
industry. Improvising on the production planning system promises a process that is in
control and more reliable. The residential construction industry will benefit from an

effective production planning system with a pr0per monitoring and analyzing tool.

1.3 Research Goal and Objectives
The goal of this research is to assess the effectiveness of production planning process in
the residential construction industry. To attain this goal the following objectives were
achieved:

I Documenting the production planning process prevalent in the industry

This objective involved the following steps:

- Interviewing of homebuilders to understand the production planning
method currently in use.

- Field observations on project sites to record how the production planning
process is conducted.

' Developing a framework for production planning performance assessment and
identification of process improvement opportunities.
This objective involved the following steps:

- Conducting a literature review and evaluation of tools and techniques used
for measuring and analyzing production performance to find a suitable
tool for measuring production planning performance and for identifying
process improvement opportunities.

- Measuring, through field visits, the production planning performance
using the selected tools and techniques.

- Analyzing results to identify process improvement opportunities.

1.4 Research Scope

The scope of this research is limited to the assessment of the current production planning
system used in the industry. This research will measure the performance of a production
planning process and analyze it to find opportunities for improvement. The goal of this
research is not to generalize any of its findings; rather, it develops a framework for
assessing the production planning system and for identifying opportunities for
improvements. The framework is evaluated on two example buildings and the production
planning process of one homebuilder is studied in detail. The homebuilder falls into the

category of a ‘medium’ size homebuilding company. The company operates in the

Lansing area (Michigan). This is due to accessibility, time and budget constraint. The tool
that was used to measure and analyze the performance is selected after the study of tools

commonly used for performance measurement.

1.5 Report Overview

This proposal is divided into five chapters. Chapter 1 discusses the importance of a
planning process in construction. A need statement was formulated and research goals
and objectives were presented along with the scope of the research.

In Chapter 2, a background of residential construction is discussed including
previous research in the area. This chapter also discusses the various performance
measurement and analysis techniques commonly used, of which one technique or system
is selected for use in this research.

Chapter 3 outlines the methodology that is adopted to attain the research goal.
Detailed discussion on action items that are undertaken to achieve the various objectives
is presented. A hypothetical case study demonstrates the research methodology and the
expected results.

Chapter 4 presents the data collection and analysis procedures. This chapter also
presents, in detail, the Production planning Model developed in this study.

Chapter 5 concludes this report with a summary, research contributions and
conclusions. This chapter also includes a discussion on research outcomes and

opportunities for future research.

Chapter 2

Literature Review

10

Literature Review

2.1 Residential Construction
Growth in the housing industry is an indicator of a strong national economy (Willenbrock
et al., 1998). The housing industry in the US is a fast growing industry and is expected to
grow astronomically in the coming years (Bureau of Census, 2000) as explained in the
‘Introduction’ chapter. Large homebuilders currently anticipate construction of up to
30,000 houses a year across the nation to meet the huge demand for houses. And the rate
of production is going to increase proportional to the demand for houses. To
accommodate this, the housing industry has adopted production building as a process for
constructing homes. Houses are built in mass numbers and customization of houses is
also made possible with various options for the customers to choose from. The planning
involved in a production process is a meticulous task and it is a key to successful
production.

This section describes the production planning process and gives an
understanding of production planning process in residential construction indusny. Prior

research in this area is presented and its limitations are stated.

2.1.1 Production Planning Process

The meaning of the term ‘Production’ at its most generic sense is synonymous with
“making” (Ballard, 2000). In the US, the residential construction sector generally repeats
a small number of models several times in a relatively confined area using specialized
trade contractors to complete each phase of each home (Bashford et al, 2003). The
homebuilders make a large number of homes in a specified time while allowing

numerous customization options to the customers. This essentially is production of

11

homes, similar to the production concepts used in the manufacturing industry. A
production process can be defined as the method or the steps involved in making a
product. In residential construction, production processes includes site preparation,
construction of homes, and finishing and delivering the end product.

The project planning process in residential construction includes defining and
organizing the work to be accomplished; the most common way of doing this is by
developing a list of activities and determining relationships between each activity and
sequencing them. A Work Breakdown Structure (WBS) is generally used to develop a list
of activities and a Gantt chart (Bar chart) and a network drawing is used for sequencing
and showing the interrelationships. Next, a quantity takeoff is performed for each activity
and the productivity of the crew is determined, from which activity duration is calculated.
All this information is integrated into a ‘plan’ with inputs of additional information for
responsibility, cost, constructability and availability of resources. For the purpose of this
research, the planning of construction activities in the field for a period of time is referred
to as ‘production planning process’ and is different from project planning. Production
planning is the process of organizing and developing a plan of daily actions to be
executed to complete a production process. This process depends on a number of factors
and variables as well as a number of key people who have been or will be associated with
the project (Peterson 1993).

Residential construction could be viewed in two different ways: a single project
view or a multiple project view. According to Bashford et a1 (2003), in a single project
view it is assumed that start of an activity depends on completion of its predecessor

whereas, in a multiple project view factors such as availability of subcontractors, their

12

response time and communication between numerous parties has to be also considered
while planning. Thus, production homebuilders view their project as a factory type
setting and not as small individual projects (Bashford et al 2003). Flaming for such a
setting has to be systematic, accurate and reliable.

In addition to the process of planning explained above for residential construction,
planning for a production type setting would include coordination of trade contractors,
planning of material supply chain, continuous availability of work and contingencies for
possible uncertainties involved in completing a task. Production planning should provide
superior organization, control throughout the project, and workflow reliability such that

work starts and finishes according to the plan.

2.1.2 Prior Research

This section reviews studies performed in the area of planning in residential construction
that are relevant to the research. Previous research efforts in this area have focused on
scheduling for repeating activities, comparison of production building and
manufacturing, and evaluation of production planning systems in construction.

In his book “Scheduling Residential Construction,” Love (1995) mentions that
“...many builders might be able to guess with a fair degree of accuracy. But can you
afford to guess when there are ways to predict accurately?” The book has an entire
chapter on the process of construction planning process to complete a residential project
successfully. Love (1995) has explained in detail the planning and scheduling of a
residential project, where planning is explained as the process where the project is
organized into different phases, developing a work breakdown structure, sequencing

activities and assigning responsibilities to perform activities to various trade contractors

13

or crews. Planning for a production style residential construction project with many
homes has not been explained in detail in scheduling books reviewed by the author
(NAHB 1993, Love 1995, Hinze 1998, and Patrick 2004).

The National Association of Home Builders (NAHB) research center published a
report in June 1993, titled, “Cycle-time reduction in the residential construction
process, ” which described an investigation conducted to find the time various builders
take to produce an average single home. The report compared the various methods used
by homebuilders for planning and scheduling to enhance their competitive position. The
report identified few approaches that defined best practices: reducing activity durations,
reducing or eliminating non-value added activities from the work schedule, performing
activities concurrently which are traditionally sequential, and creating an even job flow
(NAHB 1993). These suggested best practices require detailed planning taking into
consideration all resource limitations, constructability issues, and productivity or capacity
of resources to be used.

Harris et al. (1998), argues that existing planning systems are ‘activity-oriented’
and are of limited use for projects with repeating activity. The authors propose Repetitive
Scheduling Method (RSM) as a planning and scheduling method which provides an
uninterrupted use of resources from one repeating unit to another. This method is
‘resource-oriented’ and it recognizes the need to have continuous use of resource through
many similar units. This method, which incorporates continuous resource usage, is
suitable for homebuilding companies with many similar units to be built. However,

adoption of such techniques is not yet widespread.

14

Any planning system, currently practiced by a company, should also be evaluated
for its accuracy and efficiency. Bemardes et al. (2002) identified fourteen practices to
improve efficiency during development and implementation of a project control system.
The practices suggested includes:

I “Planning in hierarchical levels” which means that the degree of planning
details should be more when the date of performing the activity
approaches (Bemardes et al. 2002).

I Short term planning

I Explicitly documenting assigned tasks

I Creating detailed specifications for tasks

I Using measurement techniques to indicate performance

Seidel (1999) presented a production planning methodology for large
homebuilding firms. This methodology focuses on production planning as a core process
of business planning. The production planning is viewed as a process which utilizes both
internal and external resources such as management, cash, labor, and material to build
homes (Seidel 1999). According to Seidel (1999), production planning for large
homebuilding companies involves organizational planning, development schedule, and
financial planning. It is a core sub process of business planning for the firm (Seidel,
1999). In contrast, this research views production planning as a process by which the
daily construction activities are planned in the field. This view of production planning is
similar to the production scheduling as used in the manufacturing industry.

Research performed on manufactured housing with focus on production planning

is discussed later in appropriate sections. Prior research, concentrating on the planning

15

aspect of production building, which fall under the Lean Construction domain will be

discussed in the next section.

2.2 Lean Production System
The second part of this literature review introduces the lean production concept, lean
construction, principles of lean construction, and finally the Last Planner System of

production control as conceived under lean construction.

2.2.1 Lean Production?
Lean production is a design and production management system developed in the ‘508 by
Engineer Ohno at Toyota. The aim of Lean Production is to shorten the cycle time to get
products to market by minimizing waste. According to Engineer Ohno, this new
production philosophy should provide (Howell et al., 1998):

I A uniquely custom product

I Instant delivery with minimum inventory

I Production with zero waste
Unlike Craft productionI or Mass-production systemz, Lean production has the qualities
of both wherein custom-made goods are produced at lower costs. The concepts of lean
production include multi-skilled workers, stopping the production process by any worker
when faults are found, no rework area, pull system (Hopp and Spearman, 1996), flexible

and automated machines, and transparency of the production systems.

 

' Craft production system which uses highly skilled workers and simple but flexible tools to make exactly
what the customer ask for—one item at a time, and of course very expensive

2 Mass-production system which uses narrowly skilled professionals to design products, using expensive
single-purpose machines and thus higher volumes are produced at lower costs but at the expense of variety

16

Lean principles have been applied successfully worldwide in the automobile
industry. Manufacturers like Toyota have strived to achieve the ideal, which is 100%
value added work with zero (or minimum) waste. These lean principles are being
increasingly employed in many other industrial sectors with great success. The best
results can be obtained when used in a repetitive or continuous production environment.

Two primary entities have advocated application of lean philosophy to
construction: The International Group of Lean Construction (IGLC) and The Lean
Construction Institute (LCI). In recent years, a handful of construction companies have
embarked on a lean conversion path. Guided by research efforts to tailor concepts of lean
production to fit the construction industry, these companies are observing good returns on
their investments, specifically in the areas of waste reduction in on-site production

activities.

2.2.2 Lean Construction

According to Howell et a1 (1998) current project management in construction views a
project as a combination of activities whereas, Lean views the entire project in production
system terms. Lean thinking brings into attention the concept of value generation instead
of activity management. The main aim is being able to minimize waste, adopt a pull
system", and produce a custom made product that satisfies the owner. Application of the
Lean production philosophy in construction has been researched and implemented and
proved to be a success. Lean construction researchers have developed techniques and
tools, which enable the implementation of Lean Construction concepts in construction

environment (Abdelhamid, 2003).

 

3 Full System in construction implies that upstream activities should not start before the downstream work
is assured to be released (Howell and Ballard, I998).

17

Koskela (1992) introduced a new unified theory of production, termed as the
T ransformation-Flow- Value Generation (TFV). This theory is an integrated theory
developed from the three main production paradigms namely; Craft production, Mass
Production, and Lean production. The TFV theory can be defined as follows (Koskela
20001

I Transformation view — Concept of transforming inputs to outputs.
I Flow view — Materials and information flow in a production process
I Value generation view — Process where the value for customer is created
through fulfillment of his/her requirements.
Koskela (2000) states that “...the new 'TFV theory of production provides a new,
theoretical foundation for construction.” Koskela (2000) also mentions that this theory
has the potential of inducing changes in performance in the construction industry.

Lean construction ideas have been adopted in the construction industry and
companies have reaped the benefits. Homebuilding companies that have adopted the
production system of construction have the most advantageous position with respect to
application of the Lean Construction philosophy. However, application of lean principles
in residential construction is relatively new. The following section will firrther discuss the
even flow methodology and last planner system (both of which seem to be very suitable

to the residential construction industry) with related literature in the area.

2.2.2.1 Even Flow Methodology
Even flow production is based upon the theory of constraints described by Goldratt
(1992) and is linked closely to Lean thinking. Bashford et al (2003a) describes and

presents the implications of even flow production techniques in the residential

18

construction industry. The objective of this technique is to reduce the work flow
variability and thus gain efficiency. Current production management strategies frequently
do not account for variability in workflow4 between independent contractors, which
reduces the efficiency of production. The objective of even flow production strategy is to
reduce the workflow variability in an effort to gain efficiency. The study by Bashford et
al. (2003a) included a simulation of a hypothetical subdivision of 90 homes in which
even flow methodology was implemented where two types of even flow were studied,
namely; activityvbased and start-based. It concluded that activity-based even flow
involves a rigid schedule that is maintained for each specific activity such as foundation
setting or framing across the entire tract of homes and it is a good way to reduce
management effort and workflow variability. Whereas, start-based even flow is when
only the initial activity for each home in the tract is scheduled to occur at a uniform rate
and it increases management effort and increases workflow variability. The study also
indicates that theories and practices used in the manufacturing industry can be effectively

applied in the construction industry.

2.2.2.2 Last Planner System

Constructing a building element requires planning to deliver materials in time and at high
quality. There is a lot of decision making in the production process where the product
moves from one stage to the other. Who is the right person to make these decisions? The
Last Planner SystemTM (LPS) states that the last person who performs the work is the one
responsible for making execution-related decisions. The last planner could be a

superintendent, foreman, or a crew member. The Last Planner System is a lean based

 

‘ Variability in Workflow refers to the uncertainty of time taken by one trade contractor to finish and
transfer his work to succeeding trade contractor

19

production planning and control system. Application of the LPS to projects has shown
reduction of duration and costs while improving quality and safety.

Ballard (2000) presents the LPS as a system of production planning and control
which can increase workflow reliability. Work assigned to a crew is termed as ‘Should-
do’ work, the capability of the crew to perform that ‘Should-do’ work is termed ‘Can-do’
work, and the work the crew actually commits to is termed ‘Will-do’ work. In traditional
project planning systems, the ‘Can-do’ aspect is not considered and a crew is assigned to
work without making sure that the work can be done. In contrast, planning for
assignments in the Last Planner System is performed after work assignments are
subjected to a constraint analysis that ensures no obstacles will prevent execution. Figure
2.1 shows the steps involved in the Last Planner technique.

As illustrated in Figure 2.1, master schedules involve the development of logic
and sequence that helps identify the commitments throughout the project. The phase
schedule involves greater detail of planning where the project components are tested for
logic and the work is divided into phases to identify constraints or related work. This
system of division into master and phase schedules lacks a tool of detailed work
structuring.

In the Last planner system, as shown in Figure 2.1, the ‘Lookahead’ process takes
its input from traditional project planning techniques, which result in a master schedule
with project milestones and phase schedules. Usually, the lookahead involves
consideration of potential assignments for the upcoming 3-6 weeks based on the project
characteristics. The activities are exploded from master schedule or phase schedule into a

level of detail, appropriate for an assignment on a weekly work plan. This typically yields

20

multiple assignments for each activity. As each assignment appears in the lookahead
window (a 3- 6 week time period), it is subjected to constraint analysis to make sure it is

ready to be executed (Ballard 2000).

 

 

 

 

. . Master Schedule
Traditional Strategic Planning
Techniques Project Milestones

Logic and Sequence
Identifying Long Lead Delivery Requirements
Drilling down
from strategic £
scheduling into Detailed (Phase) Schedule
production More Detail
planning and Project Components
detailed work Testing Logic
assignments Identifying Related Constraint Groups

 

 

 

 

__________________ +-__--__-_____..__--

Six Week Lookahead
Constraint Analysis
Responsibility Identification
’To-Do’ List
Resource Levelling
Weekly Updates

I

Weekly Work Plan
Detail Work Assignments
Decentralized Crew Planning
Measurement of PFC
Learning
Workable Backlog

 

Last Planner
Techniques

 

 

 

 

 

Fig 2.1: Construction Planning vs. Last Planner
(Modified: Kaufman Consulting Group, LLC)

Assignments that are made ready for execution enter into a workable backlog.
The assignments entering the workable backlog are all constraint free and in the proper

sequence for execution. Ballard (2000) suggests that assignments are quality assignments

21

when they satisfy the following criteria: definitions, sequencingb, soundness7 and sizes. If
the last planner finds a constraint that could not be removed in time, the assignment
would not be allowed to move forward. The last planner

should maintain a backlog of work ready to be performed, with assurance that everything
in the workable backlog is indeed workable (Ballard 2000).

Weekly work plans are formed from the workable backlog. Such assignments
help improve the productivity of those who receive them and increase the reliability of
workflow between the production units. The Last Planner System can be viewed as a
needed supplement to traditional project management for better production. The analysis
of reasons for plan failure reveals more information regarding how the production system
actually functions and what could be done to improve it.

To assess the quality of the assignments made, a metric known as the Percent Plan
Complete (PPC) calculated as a ratio of the number of assignments completed to the total
number of assignments planned for that particular unit in a given period of time. PPC is
expressed as a percentage with a range between 0%-100%. In general, the higher the
PPC, the more reliable the production planning system. A PPC of 100% means all the
work assigned is completed as planned and it is the best-case scenario. A PPC value less
than 100% means there is a problem with the production planning system.

The reasons for the failure have to be evaluated to enable future improvement of

the production planning process. Thus, in the homebuilding industry where a production

 

5 Definition: the work should be specific enough to understand the requirement and completeness of the
assignment (Ballard, 2000).

6 Sequencing: the assignment should be sequenced in thecorrect constructability order (Ballard, 2000).

7 Soundness: the materials design and perquisites are identified and constraints removed (Ballard, 2000).
8 Size: means that the assignments should be correctly sized based on the capacity of the crew (Ballard,
2000).

22

process is adopted, a planning system like the LPS will make the process very reliable
and thus reduce variability, while increasing throughput and profits.

Chitla (2003) used the LPS to evaluate the production planning process in
manufactured housing and to identify opportunities for improvements. The PPC metric of
the Last Planner System was used to evaluate the efficiency of production planning.
Analysis of the output from the evaluation utilized the Pareto Analysis and Fish Bone
Diagrams. This analysis was used for finding reasons behind the failure of the plan and it
can also be used as improvement opportunities (Chitla, 2003).

The next section discusses performance measurement tools by introducing Total
Quality Management and Six Sigma, the quality concepts used in the manufacturing

industry to increase productivity.

2.3 Performance Measurement in Construction
The Arizona Partnership for Advancing Technology in Housing (AzPath) is a regional
partnership formed to promote research and development in the Arizona homebuilding
industry. Arizona State University (ASU) has entered into a partnership with PATH (a
White House initiative aimed at bringing new technologies to the homebuilding industry),
five major Arizona homebuilders, and the Home Builders Association of Central Arizona
(HBACA). AzPath was formed to serve as a vehicle to lead the southwest area housing
industry to embrace research, learning, and the adoption of innovation to provide housing
of higher value at a reduced cost.

Figure 2.2 shows the evolution of residential construction as compared to the
manufacturing industry. In the manufacturing industry various new philosophies were

introduced and implemented which escalated the production capability. Residential

23

construction had evolved primarily as a Craft production system and has some
improvements with the introduction of modularization, smarter materials and
management systems, information technology, and new manufacturing technology.
Homebuilding, which has been accepted as a production process, certainly needs to take
the path of manufacturing industry and explore the various techniques and tools used to
improve on production and increase efficiency and profits. The construction industry
certainly needs to be at par with manufacturing industry, and this can be achieved by
researching on the progress track of manufacturing industry and adopting and then

tailoring the philosophies to the needs in construction.

 

   
 
        

 
   
 

    
   
      
 
  
  
  
  

1* Mass Customization New\
Lean Thinking Manuf-
Quality Movement actuting
Manufacturing \‘ Process
Track Approach
to Housing

Automatic

A b1 L'
Line Process ssem y me

Industrial Technology

Panelization

Interchangeable Parts Modern Mngmnt Systems

Production Output

T

     

Power Tools

 

 

 

 

 

New Materials
Craft Based Assemblv
1825 1880-1890 1913 1950 1980 1990 7
Process Evolution Over Time
Fig 2.2 Evolution of residential construction
(Modified: Azpath Website, http://construrtinn as" ‘ ' . " " a "‘ ‘ .jpg)

 

24

Lean production, Total Quality Management (TQM), and Six Sigma are the three most
recent philosophies that had a major role in the present success level of the manufacturing
industry. The following section will briefly discuss TQM and Six Sigma philosophies,
which focused on quality and continuous improvement. Lean production and its

application in construction have been discussed in earlier sections of this chapter.

2.3.1 Total Quality Management

The primary purpose of TQM is to provide excellence in customer satisfaction through
continuous improvements of products and processes by the total involvement and
education of each individual who is in any way a part of that product/process. It is a
structured approach to improvement. If correctly applied, it will assist a construction
company in improving its performance (Ahmed et al, 2001).

Ahmed et al (2001) have explored TQM as a tool to achieve continuous
improvement in construction. The primary objective of this study was to develop a
measurement model - with tools and methodologies for the identification and
measurement of construction processes for continuous improvement and customer
satisfaction (Ahmed et al, 2001). The authors have identified some causes for client
satisfaction, and this was achieved by developing a model using TQM tools. The study
emphasized on the necessity of measurement of construction processes. The author states
that continuous improvement and customer satisfaction will in turn improve quality,
productivity, and competitiveness. For a homebuilding company it is important to have a
high level of customer satisfaction, indeed most of the home companies focus on this
aspect because it is the ‘word—of-mouth’ of satisfied customers that can increase its

popularity and standing in the industry. Hence a technique to assess and measure

25

customer satisfaction and also internal process performance is the primary need for
growth. Bassioni et al. (2004) addresses the need for research to develop comprehensive
performance measurement frameworks for the construction industry.

Process performance can be measured on the basis of any aspect that improves the
quality of the final product. Performance measurement techniques and tools are abundant
in the manufacturing industry, but it’s critical to first identify what has to be measured
and then find out how it is to be measured such that it leads to process improvement. Cox
et a1. (2003) have identified performance indicators to measure construction performance.
They termed these as “Key Performance Indicators” (KPI). The study identifies
quantitative and qualitative performance indicators as perceieved by management in
construction firms. The study identifies quality control, on-time completion, and cost as
three of the six highly significant performance indicators as perceived by management
(Cox et. al., 2003).

In other studies, the major focus in the process improvement has been reduction
of variability in a process. According to Picard (2002) “...reducing the variation or
variability in a process is considered a sign of improved quality of the construction
process.” Lean Construction also advocates reduction of waste and value generation. To
reduce waste and non-value adding activities, the process has to be studied extensively
and all non-value adding activities identified and eliminated. Thus, the objective should
be to understand the process, measure the performance, and find out the causes for
variability and then eliminate or reduce the causes. Six Sigma approach aims at this

objective and is discussed next.

26

2.3.2 Six Sigma

Motorola developed Six Sigma in the late ’80s as a system that would help them achieve
near-perfect products (Breyfogle et al. 2001). This system aims to reduce variability and
uses various statistical methods and tools to identify and reduce variability and thus
achieve a ‘closest to zero-defect’ product. Linderman et al. (2003) defines Six Sigma as
. .an organized and systematic method for strategic process improvement and new
product development that relies on statistical methods and scientific method to make
dramatic reduction in customer defined defect rates.”

The aim of a company adopting Six Sigma is to bring all its defect rate to as low
as 3.4 defects per million opportunities, which translates to a 6-6 measure. Sigma (0) is
the symbol for standard deviation in statistics. Thus a six sigma level means having all
the products produced within six standard deviations of the mean (average). Thus a
process having a six sigma yield level will have 99.99966% of its products produced
without defect. The methodology used to achieve six sigma goals is known as DMAIC
(Define, Measure, Analyze, Improve, and Control).

The following example of the sigma metrics applied to a small library will
illustrate the different sigma yield levels (Breyfogle, 2001). The example also shows that
the sigma yield levels are not linear but they are exponential.

I Sigma level 1 corresponds to 170 misspelled words per page in a book

I Sigma level 2 corresponds to 25 misspelled words per page in a book

I Sigma level 3 corresponds to 1.5 misspelled words per page in a book

I Sigma level 4 corresponds to 1 misspelled word per 30 page in a book

I Sigma level 5 corresponds to l misspelled word in a set of encyclopedias

27

I Sigma level 6 corresponds to lmisspelled word in all the books in a small library

Abdelhamid (2003) reviews Six Sigma, its methods and metrics and suggests
applying it to lean construction. The paper shows an example by extending the Six Sigma
metrics to the Percent Plan Complete (PPC) of the last planner system. The Six Sigma
metrics provides a more thorough understanding of the process and gives a true picture of
its performance. This metric will standardize the performance measure and thus can be
used to compare the performance of various processes. According to Abdelhamid (2003)
the advantage of finding the sigma quality level is that it better reflects the magnitude of
the disability in the production planning process to maintain a reliable workflow.

This research uses Six Sigma methods and metrics to measure the performance
and assess the planning process in the residential construction industry. A detailed

explanation of all methods and metrics to be used are included in the next chapter.

2.4 Summary

This chapter provided the relevant literature available with respect to this research.
Literature in residential construction that focused on planning and its importance was
covered in the first part of this chapter. The second part summarized Lean production and
its application to the construction industry. The Even flow methodology and Last Planner
System, which are two Lean-based techniques, were discussed as they are suitable for the
homebuilding industry’s production planning. Finally, TQM and Six Sigma were briefly
discussed in this section. The next chapter describes the methodology used to achieve the

goal of this research.

28

Chapter 3

Methodology

29

Methodology
While the residential construction industry may be able to meet demand, it is important
for the industry to satisfy its customers by improving its performance. Improving
performance should incorporate an assessment of its current process. This research aims
to do this by taking the first step. Planning as explained in earlier chapters has an impact
on the entire homebuilding process. Thus, assessment of the production planning process
is the main focus of this research. The DMAIC (Define, Measure, Analyze, Improve, and
Control) method proposed in the Six Sigma literature and the LPS in Lean were selected
as appropriate approaches for the assessment of the production planning process. The
following sections will explain the methodology of this research and steps taken to fulfill

the objectives and each of these steps will be discussed in detail.

3.1 Methods and Tools for Objectives
The goal of this research is to develop and implement an assessment framework for the
production planning process (PPP) in the residential construction industry.
To attain this goal the following objectives were proposed and accomplished:
I Documenting the production planning process prevalent in the industry.
I Developing a framework for production planning performance assessment and
identification of process improvement opportunities.

The following section will discuss how each of these objectives were achieved.

3.2 Mapping the Production Planning Process
The production planning process in the residential construction industry is different from

one homebuilder to the other. However, the aim in this research was to develop a

30

framework for assessing the PPP by considering a small group of homebuilders. The aim
is to assess the PPP and not to generalize on current practice.

The first objective, as mentioned in chapter 1, involved mapping the PPP.
Mapping of a process involves determining the process inputs and outputs, the value
adding sub-processes, the involvement of any other process, flow of information and all
cross firnctional activities. Once these are determined it is easier to create a process map.
A process map is similar to a basic flow chart with all functions shown in boxes and flow
of data shown by arrows. Mapping a process is explained by Pande et al. (2000) to be
analogous to the act of putting a puzzle together. While putting pieces of a puzzle in
place, firstly the puzzle box should give you a basic idea of what has to be done. Next,
one will assemble the edges of the puzzle together and then assemble the internal pieces
such that it adds to the clarity of the initial image. This step will involve some trial-and-
error and some research too. Similarly, in this research, a broad picture of the PPP was
established. Then, the details of the various process steps were determined. This data was
incorporated in the form of a map, which gave a clear picture of how the PPP is
undertaken. Mapping the PPP had the following advantages:

- gave a clear ‘big-picture’

- defined the key inputs and outputs of the process

- facilitated the understanding of cross functional activities

- showed the interface between various contributors in the process
- showed how the process flows through the organization

- highlighted the value adding sub-processes in the overall process

31

The information necessary for creating a process map of the PPP for
homebuilding activities were collected in two steps:
I Interviewing management and field personnel,

I Field observations.

3.2.1 Interviewing

The interviewing process involved the development of an interview questionnaire
focusing on understanding the PPP employed by homebuilding companies. It also
included questions intended to shed light on the interviewee’s perceptions about the
process, what was lacking in the process, and any recommendations for improvement.

The interview questionnaire is included in Appendix A.

3.2.2 Field visits

The purpose of field visits was to develop the complete process map of the PPP. After the
interviews, the author had a macro-level view of the process. Similar to completing the
inner pieces of the puzzle, the field study gave a better representation of how a plan was
deveIOped directly from the inception stage. Data on how each step of the planning
process was sequenced, how the information flowed between sub-processes, and what
inputs were used to complete the planning processes were collected. Any other process
that had an impact on the planning process was also considered for the study. The
duration of each field visit was driven by the magnitude of the process in the particular

company.

32

3.2.3 Creating a Process Map
A process map was used to show the flow between each step involved in the overall
process. Pande et al. (2000) suggested the “SIPOC” process model which includes the

following:

Supplier: the person or group providing key information or material or other
resource to the process.

- Input: the “thing” provided.

- Process: the set of steps that transforms or adds value to the input.

- Output: the final product of the process.

- Customer: the person group or process that receives the output.
Such a process map will give the overall picture of the process. All information collected
in this research was documented in this format.

The process map provided insights critical to the second objective of the research.

The map clarified what items had to be measured for an accurate assessment of the
production planning process. The next section explains the tools used for measuring the

performance of the production planning process.

3.3 Production Planning Process Performance
The second objective of this research was directed at measuring the production planning
performance with a suitable tool to identify improvement opportunities. This objective
was realized by performing the following three steps:

1. A literature review was conducted to evaluate previously developed tools and

techniques for measuring and analyzing production planning. A suitable tool for

33

measuring and analyzing production planning performance was selected and
tailored to the research needs.

2. The production planning performance was measured using the selected tools and

techniques.

3. The results were analyzed to identify improvement opportunities.

Measurement of process performance is a necessary step to identify the shortcomings
of any process. Measuring, analyzing, and implementing are the three basic steps to
improve any process. Once the process performance is measured, it makes it easier to
analyze and find out the shortcomings, if any, in the process. The identified shortcoming
should be rectified or removed and the new improved process should be implemented.
This new implementation should be measured and analyzed again for its performance and
then control can be achieved. These are the steps, that make it possible to continuously
improve a process. However, this research is limited to measuring the performance and

analysis of the observations.

3.3.1 Performance Measurement Tool

The second chapter included a section for performance measurement, wherein two tools
suitable for measuring the performance of the production planning process were
identified, namely, Percent of Plan Complete (PPC) and the Six Sigma metrics. The

following sections will discuss how the two metrics were used in this research.

3.3.1.1 Percent of Plan Complete (PPC)
The Last Planner System (LPS), as explained in Chapter two, provides a framework to
plan and control daily production assignments in a construction project (Ballard 2000).

This system uses a metric termed Percent Plan Complete (PPC) to measure the reliability

34

of production planning and that of workflow. PPC is the ratio of number of assignments
completed to those planned. The PPC is generally calculated on a daily or weekly basis.
It is reported for a particular trade or crew on a periodic basis to measure the
effectiveness of the planning process or the capability of the crew to execute the
assignments they commit to. Once PPC for a crew is obtained, an average value of the
metric over a period of time will reflect the efficiency and effectiveness of production

planning for that particular crew was.

3.3.1.2 The Six Sigma Metrics
The Six Sigma approach to improve quality of an organization advocates the use of
quantitative metrics to assess performance. A number of metrics are available depending
on the problem. Rolled throughput yield (YRT) is a metric suitable to this research,
especially when combined with the PPC metric.

Abdelhamid (2003) proposed the application of Six Sigma in Lean Construction.
In this paper, the rolled throughput yield (YRT) metric was explained and its use as a
performance measure with the Last Planner System was demonstrated. Generally the
metric ‘Yield’ (Y) represents the percentage of units that pass final inspection relative to
the number of units that were processed (Abdelhamid 2003). This metric does not take
into account the rework performed in any of the sub-processes within a process, and thus,
it masks the true output of the process. In contrast, the Rolled Throughput Yield (Y RT)
metric takes into account all the rework done in a process. The YRT metric measures the
output at every sub-process within a process to calculate the final Yield of the whole

process. The YRT can be mathematically shown as follows:

35

YRT= Hm I Yi ...(I)
1:

where m is the number of processes involved and Y, is the throughput yield of process i.
The Six Sigma metrics are obtained by converting the YRT to a sigma level so that
comparison is possible across different operations and even across industries. This can be
achieved by using standard tables or by calculations. The sigma quality level for a
process using rolled throughput yield can be calculated using the set of equations

(Breyfogle 2002).

Sigma level calculations:
- The rolled throughput yield is calculated from the yields of different sub-

processes as follows:

YRT= Hirn—l Yi ...(I)

Where, YRT = Rolled throughput yield

Y, = Yield of sub-process ‘i’

- The rolled throughput yield is normalized by finding the geometric mean
as follows:
Ynom.= M (ii)
- The normalized yield is then converted to Defects Per Unit (DPU) using

the following equation:

DPUnorm = 'ln (Ynorm) (III)

36

- To determine the sigma quality level, also called as Zbemhmrk, for the
process, the following equation is used
Sigma quality level = Znorm + 1.5 (iv)
Where, 2mm is the standard normal value corresponding to the

DPUnorm found using the standard normal table.

Modified calculations using PPC (Abdelhamid, 2003)

The modified equations used for this research using PPC value are as follows:

- The rolled PPC is calculated using PPCs of different sub-processes as

follows:
rolled PPC = 1'1““ PPC- (v)
i = 1 ‘

— The rolled PPC is normalized by finding the geometric mean as follows:

PPCnorm = W (vi)
- The normalized PPC is then converted to Defects Per Unit (DPU) using
the following equation:
Drum... = -1n (Prcnom) (vii)
- To determine the sigma quality level, also called as Zbenchmark, for the
process, the following equation is used
Sigma quality level = Znorm + 1.5 (viii)
Where, Znorm is the standard normal value corresponding to the DPUnorm found using the

standard normal table.

37

The sigma level corresponds to ‘parts per million’ (ppm) which essentially is the number
of defects per million parts. Hence, this metric emphasizes the magnitude of the problem.
It gives a more representative picture of the state of the process. A low sigma level

indicates need for process assessment and problem solving.

3.3.1.3 Percentage of Plan Complete (PPC) vs. Rolled throughput yield (Y nr)
Abdelhamid (2003) illustrated an example of using the YRT metric instead of PPC in a
manufactured homes factory setting. He suggested using rolled PPC similar to the rolled
throughput yield as explained above. Using the rolled PPC metric, as suggested in the
paper, exposed the hidden factory (rework performed to rectify defects during sub-
processes). It also standardized the metric so that it can be compared with other
production processes as well as other production processes of other companies. The
rolled PPC metric “... gives a better sense of magnitude of the process performance
failure” (Abdelhamid 2003).

This research uses the rolled PPC metric as a measurement tool for assessing the
performance of production planning in residential construction. This metric gives a true
picture of the deficiencies of the planning process for any homebuilder. In contrast to
calculation of PPC, the rolled PPC will better reflect the magnitude of the problem, and
thus it is more realistic tool for performance measurement. The rolled PPC can be
calculated by multiplying the PPC of every single sub-process as shown in the previous
section for calculation of rolled throughput yield. The PPC of every sub-process will be
useful to determine the inefficiency of planning for that particular process, whereas, the

rolled PPC will aid in finding the shortcomings of the planning process as a whole.

38

3.3.2 Measurement of Planning Performance

The second step towards attaining ‘Objective II’ was to measure the performance of
production planning process of the homebuilding companies selected. This step involved
field visits and recording of day-to-day activities. The idea was to calculate the number of
assignments completed vs. the number of assignments planned to be completed. The data
was collected from the supervisor on site as well as by direct observation. A sample form

for data collection is shown in Figure 3.1.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Crew Observed Feb 4 - Feb 8 Feb 11 — Feb 15
MIT [WIT IF MIT IwIT IF
Framing Crew REMARKS:
1. Measuring and cutting
2. Framing First half
3. Framing Second half
4. Flooring
5. Finishing and Cleaning
Completed assignments
PPC
Insulation Crew REMARKS:
1. Outer Wall Insulation
2. Inner wall insulation
3. Flooring Insulation
4. Ceiling Insulation
Completed assignments
PPC

 

 

 

 

 

 

 

 

 

 

 

Figure 3.1 Sample format of data collection

The choice of entries in the form were a ‘C’ (Complete) or an ‘I’ (Incomplete),
respectively indicating whether the assignment was completed successfully or not by the
crew. The calculations for the rolled PPC were performed for each process as explained
in the earlier section. This rolled PPC was converted to a corresponding sigma value to

determine the efficiency of the production planning process.

39

 

For this research, PPC data was collected for 1 1 weeks. This data was sufficient
for calculating the rolled PPC value and for use in analysis. The next section will explain

the analysis techniques adopted in this research.

3.3.3 Data Analysis

Once all the measurements were available, the next step was to find the causes behind
failures in the production planning process. For this research, visual tools and root cause
analysis (Breyfogle, 1999, and Evans and Lindsay, 2001) were used. The following

sections will discuss these methods of analysis and their use in this research.

3.3.3.1 Visual Tools for Data Analysis

The best way to detect a problem is when it is visual. When the data is arranged in certain
patterns, one can detect variability, catch the outliers, and follow trends that occur in the
data set (Pande et al., 2000). Visual presentation of data also helps in understanding the
process and thus makes it easier to identify the source of a problem. Some common
techniques used are: Pareto Charts Frequency Plots, Run Charts, and Correlation
Diagram (Evans and Lindsay, 2001 and Breyfogle, I999). The choice of a technique
depends on the type of data collected as well as anticipated problems. The Pareto chart
and the Run Charts were used for analysis of the PPC data collected. The following

sections provide examples of how these tools were used.

Pareto Charts or Pareto Analysis

The Pareto Charts are bar charts where the data is stratified into groups and displayed in a
descending order according to frequency of occurance. It helps in finding out the most
common cause of a problem under investigation. Pareto analysis is based on a notion that

80% of the problems are created by 20% of the causes. This suggests that for most

40

situations, major portion (about 80%) of defects or inefficiencies are caused by factors
that are about 20% of all the factors determined. Thus a Pareto chart helps in highlighting
the factors that cause most of the problems. This indicates that these factors should be
analyzed first and the root causes can be eliminated. A typical Pareto chart looks like

Figure 3.2. In the figure, categories A & B are the main problem areas.

Pareto Chart l

 

 

 

 

 

 

 

 

7o ‘
60
o 50 ,
‘U
3 40
'E
g 30 ,
E 20 ,
10 7 , , , , '
o - - - .
A B c D E I
Categories I

I I T T" I '7 a T 7 EiguiésizTTQPEII PgretoChagin— ’ ‘7 A W A V J

In this research, the Pareto chart was used by first dividing the data into categories
of production problems. Hence the type of work needing a better planning technique can
be identified. Then each of the problem areas are separately analyzed and factors causing
non-completion can be determined. This can again be shown using a Pareto chart and the
major contributors can be analyzed further to find the root causes. For example, if
framing and roofing are found to be the two major contributors causing most of the
delays, we analyze these two activities separately to determine the root causes for each of
these activities.

The Pareto chart can also be used to compare problems by days of the week to

assess whether a particular day of the week has more planning problems. For example, if

41

the planned assignments are met the least on Thursdays and Fridays, we could utilize this
finding and change the planning process such that less volume of work is assigned to the
crew on these days. Thus, the data available can be used in different formats and find
ways to improve the process. Pareto charts can also be used to identify the root cause that
creates most of the problems. This use of Pareto charts is described later during the root

cause analysis discussion.

Run Charts or Time Series Plots

The Run Chart is useful for showing degree of variations in the process over time. It
could be also used for detecting any pattern (weekly cycle or an hourly cycle) in the data-
set over time. It is a very useful tool to see how a process is responding to a change.
There are various types of run charts. The p-chart with variable sample size is the most
suitable for this research. In a p-chart, a statistical range (upper and lower limits) is set
for the data and the data is plotted. The range can be set by defining an upper control
limit (UCL) and a lower control limit (LCL). These limits can either be set by the
company or can be mathematically calculated as shown later. Any data that is outside the
limit is an indicator of an out of control process. This may be due to a special cause and it
can be detected and avoided.

For plotting a p-chart with variable sample, the data collected has to be rearranged
in a spreadsheet depending on the type of the problem being investigated. Figure 3.3 is
the format used for plotting the p-chart for the data collected in this research. It should be
noted that the PPC metric is converted to a Percent Plan Incomplete (PPIC) calculated as

l-PPC or as a ratio of Tasks Incomplete to the Tasks Plamed. This modification was

42

necessary because of interest in causes for incomplete assignments. Plotting the PPIC

made it easier to detect instances where the assignments were not successfully completed.

 

Day Tasks Tasks PPIC Std Dev UCL CL LCL
Incomplete Assigned (ratio)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3.3 Format for p-clrart data

The attributes for each day can be calculated using the following formulas (adapted from

 

 

 

Breyfogle, 2003):
. Std _ dev : PPICavg(l - PPICavg) (ix)
Tasks _ assigned
2 Tasks _ incomplete
where, PPICavg = ...(x)
Z Tasks _ assigned

I UCL = PPICavg + Std __dev ...(xi)
" CL = PPICavg ...(xii)
I LCL = PPICavg — Std _dev .. .(xiii)

Completion of the spreadsheet in Figure 3.3 enables plotting the p-chart. Figure 3.4 is
an illustration of a p-chart with variable number of assigned tasks for each day. Different

p-charts for different crews observed can be plotted to find crew-wise variations or total

43

for each day can be considered. This will depend on the type of analysis to be done. This
research includes all possible p-charts for the data-set to determine variations of different

aspects of plaming.

 

 

 

 

 

 

 

 

 

 

 

 

 

I
I
I
I
2 _ 50 .....-zm - .. ---“-.. WWW---” m- I
1 I
I 1
l 2.00 - 1
I g r _, a a _ _ _- -.- _:
a 1.50 4 I—UCL M
I 3 ‘——CL 11
C i1
’ 3 ”’0 1 17.183911139111998?
i a ,
. -
1 0.50 I
I

0.00

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3.4 Sample p—chart

3.3.3.2 Root Cause Analysis

Typical analysis of data is comprised of two major activities: detection of a problem and
finding the root cause of that problem. The Pareto chart and the P-chart as described in
the previous section helps in detecting problem area and specific problems. The next step
of finding the root cause for the identified problems can be performed by drawing a
‘Cause and Effect’ diagram. The following section explains how a cause and effect

diagram was used for this research.

Cause and Effect Diagram

Variation in a process performance can occur due to many reasons, such as materials,
methods, measures, machines, environment, and people (Evans et al., 2002 and Pande et
al., 2000). Identifying causes of such problems can be achieved by using a cause and
effect diagram. It is a simple graphical method for presenting causes and effects and for
sorting out causes and identifying relationships between variables (Evans et al., 2002). It
is also called a F ishbone Diagram due to its structure. The general structure is shown in

Figure 3.5.

Cause Cause Contributor
/ J to the Cause

Contributor
to the Cause

 

P Problem

Contributor
to the Cause /'

Caus

Figure 3.5 General structure of a cause and effect diagram
(Source: Evans et al., 2002)

The best way to construct a cause and effect diagram is in a brainstorming session
with all the participants giving various ideas that could be a cause of the problem. In this
research, a Cause and Effect diagram to find the root causes of the problems identified
using the Visual tools was constructed. The root causes having the most impact were
identified and selected as a prime candidate for further evaluation.

Both the visual method and root cause analysis aided in finding the problems

causing poor performance of the production plaming process. The identified causes will

45

lead to many process improvement suggestions, which will be discussed in the following

chapters.

3.4 Summary

This chapter described the methodology used to achieve the goals and objectives of this
research. The production plaming process was assessed using Last Plamer System and
Six Sigma concepts. The findings and LPS are used to demonstrate how a production

planning model specific to a residential housing company may be developed.

46

Chapter 4

Data Collection and Analysis

47

Data Collection and Analysis

4.1 Overview

In the previous chapter, the methodology adopted to achieve the goal and objectives of
this research were presented. This chapter explains, in detail, the steps involved in data
collection and its analysis. The analysis is performed following the steps discussed in
section 3.2 and 3.3 of the previous chapter. The outcome of the data collection and its
analysis are presented in this chapter. Data collection was conducted over two phases:
Interviews were conducted with construction professionals from four homebuilding
companies, and data of daily activities at one residential construction site were recorded.

Each of the steps involved are discussed in the following sections.

4.2 Interviews

Interviews were conducted for this research in order to understand and map the
production plaming process prevalent in the homebuilding industry. For this purpose, a
questiomaire was developed. The questiomaire contained 45 open-ended questions
categorized under six headings, Background information, Long-term plans, Subdivision
plans, Construction process, Plaming process and Opinions. Refer to Appendix A for the
complete questiomaire. The questionnaire was approved by the University Committee on
Research Involving Human Subjects (U CRIHS).

The author interviewed eight construction professionals from four homebuilding
companies. An upper management and a field management persomel were selected from
each company. The companies selected were from Mid-Michigan and Southeast
Michigan. The number of homes built by the companies ranged from 10 homes to 500

homes annually. The interview answers are summarized and presented in Appendix C.

48

The names of individuals and their companies are not disclosed for confidentiality. In the

summary, ABC 11, ABC 22, etc. are used to identify the different companies, and XY Z

11, XYZ 12, XY Z 21, XYZ 22, etc. are used to identify the individuals interviewed in

each company.

4.2.1 Interview Response

This section summarizes the responses to the interviews conducted by the author. For

detailed responses from all the interviewees refer to Appendix C. Highlights of the

interview responses are as follows:

Number of subdivisions for the homebuilding companies ranged from 4 to 32

subdivisions with a goal of building 18 to 400 homes in the year 2004.

Most companies set the goal for number of homes based on demand, market

trends, sales and historical data.

Production plaming for a subdivision is generally done 3 to 4 weeks before

construction.

Almost all companies have a schedule or a representation for a subdivision or
single home. The companies use MS Excel or any standard/tailored scheduling

software for the same.

Generally a time contingency ranging from 1 week to a month is built into the

schedule for compensating time lost.

If the interviewee has worked for another company previously, they are of the

opinion that the current production plaming practice used is better.

49

o The interviewees are of the opinion that regularly updating the schedule, and

centralized scheduling are good practices of production planning.

0 Most interviewees believe that more research is necessary and can improve
production plaming in homebuilding construction. Some mentioned that intemet
based plaming, benchmarking, Just In Time (JIT) etc are areas that could be

adopted to production planning in residential construction.

4.2.2 Production Planning Process

The main objective of the interviewing process was to understand and map the
production plaming process prevalent in the homebuilding industry. For the purpose of
this research, homebuilding companies are classified as small size (1 to 50 homes per
year), medium size (51 to 300 homes per year), and large size (301 and more homes per
year). The interviews reveal that the production plaming process implemented typically
varies with respect to the size of the company. The information obtained from the
interviews was used to construct flow charts representing the production plaming
process. Three different flow charts are produced, each representing the production
plaming process in the three categories of homebuilding companies. The following
sections detail the production plaming process in each category of homebuilding

companies.

Small Size Homebuilding Companies (1 to 50 homes per year)

In homebuilding companies building less than fifty homes per year, it is typical to have a
specific production plan for each home to be built. The company builds a home mostly on
demand by a particular customer or developer. Once the plan is selected and the permits

have been requested, the superintendent drafts a rough schedule based on past experience

50

and a general idea about the specific home plan. Mostly basic spreadsheets are used to
represent the schedule. When the subcontractors are selected, they are allotted the time
they request to complete the activities they are responsible for. The rough schedule is
then adjusted according to each subcontractor’s time input. Once an activity is started, the
superintendent keeps track of the trades to determine the exact date on which the trade
will finish, and informs the following trade contractor to start their work. In case of
delays due to weather, material and/or low productivity, the superintendent
communicates with the subcontractors and reschedules the activities.

It appears that, in small homebuilding companies, the production plaming is on
an activity-by-activity basis. The schedule largely depends on the availability of
subcontractors and this becomes a constraint to achieve the finishing dates plamed
initially. No formal recording and assessment technique is used for improving the
production plaming process. The typical average percentage of work completed per plan
(or PPC) is about 70% (this is based on the interviews conducted). Figure 4.1 is a flow
chart representing the production plaming process for the small size homebuilding
companies interviewed. It is safe to assume that this is representative of most small

homebuilding companies.

51

 

Rough Schedule

(Based on home plan)

/ Past Experience; Inputs: <3
\ Trade contractors

Legends:

 

 

 

 

 

 

 

 

 

 

Final Schedule OUIPUISI

(Based on permit approval)

/ Work Progress for
\ each trade

Activity Based Schedule
(Production Plan)

 

 

 

 

 

 

 

 

 

Figure 4.1: Flow Chart Representing Production Planning for Small Homebuilding Company

Medium Size Homebuilding Companies (51 to 300 homes per year)
In medium sized homebuilding companies, it is typical to have a general template of the
production plan. Typically, the production plan is graphically represented on spreadsheets
or using custom software outputs. The project manager for the respective subdivision will
use this template to form a new template based on the specific home plans to be built.
This new template will incorporate the time required for each activity based on the
project manager’s past experience with the selected subcontractors. This new template is
used for production plaming for all the homes in the subdivision. Once, a plot is selected
and permits are secured, the expected start date is plugged in the template to produce a
complete production plan for the home. This is repeated for each home. Some companies
will have specific software tailored for the company, which will adjust the schedules for
more than one home when the same resources are used.

The production plan for each home is handed to the various trade contractors and
suppliers. Typically, the superintendent will update the production plan on a weekly

basis. Most of the delays due to low productivity and/or weather are countered by

52

 

working weekends and/or overtime. If such adjustments are not possible, a new updated
production plan is communicated to the downstream subcontractors.

In general, the production plaming in medium size homebuilding companies
seems to follow a unit-by-unit scheme. The company may have a performance
assessment system for each subcontractor or supplier to gauge their capability and to
make selection decisions in the future. There appears to be no formal production plaming
process assessment technique. Records of daily performance are not maintained and are
not used for future plaming. The typical average percentage of work completed per plan
(or PPC) is about 80% to 90% (this is based on the interviews conducted). Figure 4.2 is a
flow chart representing the production plaming process for medium size homebuilding

companies.

Standard
Template
/ Past Experience;
\ Trade contractors Legends:

Specific Schedule Inputs: C>

(Based on home plan)

/ Dates Based on OUtPUtSI
Vermit Approvals

Production Plan for

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Each Home
/ Weekly Update;
\ Trade Coordination
Unit Based Schedule
(Update regularly)

 

 

 

Figure 4.2: Flow Chart Representing Production Plan for Medium Homebuilding Company

53

Large Size Homebuilding Companies (301 or more homes per year)

Large size homebuilding companies will typically have a fixed number of starts in each
of its subdivisions. This is predicted based on various factors such as number of
households in the area, employment growth, census, and competitive survey. The sale of
homes is generally high and this helps in keeping a steady construction rate. The
company will generally have customized software for production plaming. The plaming
is based on a standard template for each home. The software uses data regarding the
number of homes to be built for each type of home, expected start date, and permit
request submission dates to produce a production plan for the whole subdivision. The
template can be modified specific to certain conditions in the subdivision to obtain a
more realistic production plan. The specific production plan is shared with trade
contractors, typically by fax. A spreadsheet with dates or a graphical representation of the
production plan is produced. It is common to have a schedule for suppliers too such that
the required material is delivered on site per the production plan.

Weekly updates and meetings are conducted to track the progress of the
production plan. Typically, a biweekly lookahead plan is produced afier each update and
this is communicated to the trade contractors performing any task in the two weeks. This
ensures that no last minute surprises occur during construction. The large size
homebuilding companies may have a centralized scheduling system, which will track the
overall production plans for all the subdivisions. This helps in coordinating trade
contractors, equipments, and materials. This system is also used to assess the production

plaming process performance. In case of any delays, it is resolved by working overtime

and weekends.

54

Given the volume of houses they work with, large size homebuilding companies,
typically plan production on a subdivision-by-subdivision basis. The company may have
a performance assessment system for each subcontractor or supplier to gauge the
capability and to make selection decisions in future. Typically, a production plaming
process performance assessment technique is used to keep track of the progress of
construction. Records of daily performance are maintained and are used for future
plaming. The typical average percentage of work completed as per plan (or PPC) is
above 90% (this is based on the interviews conducted). Figure 4.3 is a flow chart

representing the production plaming process for large size homebuilding companies.
Standard
Template

Specifier Schedule

(One for each Subdivision)

 

 

 

 

 

 

 

 

Permit Approval

Past Experience;
< Trade contractors

£ / Subdivision Sales;

/

 

Centralized Schedule
(Compile all Subdivision Schedules)

 

 

 

 

 

. Legends:
Trade/Supplrer
Coordination; Inputs: Q
Personnel Capacity

Outputs:
Periodic Update;
Current Progress

 

 

 

 

Periodic Lookahead Schedule
(One for each Subdivision)

I

Daily Production
Plan

 

 

 

 

 

 

/\ /\

 

 

 

 

Figure 4.3: Flow Chart Representing Production Plan for Large Homebuilding Company

55

4.2.3 Summary

The previous section provides an overview of the production plaming process prevalent
in the homebuilding industry. It also documents a detailed production planning process
by means of the three flowcharts representing the process of production plaming
observed in the three broad categories of homebuilding companies. The documentation

was based on interviews and site observation performed by the author.

4.3 Data Collection on Site

To facilitate measuring production plaming performance, an assessment tool was
developed. A complete explanation of the assessment tool was presented in Section 3.3.
To demonstrate the use of the assessment tool and to identify areas of improvement, data
collection and analysis was conducted for one homebuilding project. The following

sections will present the details of the data collection process and respective analysis.

4.3.1 Site Overview

One homebuilding project was selected for the purpose of data collection. This
construction site was situated in Lansing, Michigan. The homebuilding company was
building one community with 12 apartment complexes. Two of the 12 buildings were
selected for data collection, as the work in progress in the two buildings was substantial
during the period of data collection. Each of these buildings comprised of 12 apartment
units. Most of the work was subcontracted. The site had one Project Superintendent, who

managed most of the daily tasks on site.

4.3.2 Data Collection
The purpose of data collection was to demonstrate the assessment tool developed in this

research and to help in developing the production planning process model. Data was

56

/

collected for a period of eleven weeks for two apartment buildings. The activities for
which data was collected included pre-drywall inspection, hanging drywall, plastering,
painting, flooring, trim carpentry, cabinet and door installation, plumbing, and HVAC.
These activities were in progress in Building I and Building 11.

During the period of data collection, the site was visited daily at approximately
one hour before the end of work. On each day, the site superintendent provided details of
the activities plamed for the next day. The author also walked around the site to check
what activities were completed each day in the two buildings. Any activity that was not

performed or partially completed was recorded as “Incomplete”.

Data was recorded in a spreadsheet as shown in Figure 4.4. The activities planned
to be completed for each day were recorded in the ‘Activities Planned’ row and activities
which were not completed were recorded in the ‘Incomplete Activities and Reasons’

Column with details on why the activity could not be completed (see Figure 4.4).

The data collected on a daily basis were then used to calculate the daily Percent
Plan Complete (PPC). This was done using another spreadsheet as shown in Figure 4.5.
The reason for incompletion of any activity was recorded using the following reason
codes: Productivity, Engineering, Non-Conformance, Owner Decision, Weather, Pre-

Requisite, No-Show, Trade, Supplier, Space, and Other.

57

Daily Data Collection Sheet
Bu l
15-Mar 16-Mar 17-Mar

2 units

Activities planned

I not started

Incomplete didn’t turn
activities and
reasons

 

Figure 4.4: Format for Data Collection

15-Mar 16—Mar 17-Mar 18-Mar 19-Mar
asks Planned 1 2 2 2 2

asks 1 1 0 2 1
50 0 100 50

Confirmance

Declsion

 

Figure 4.5: Spreadsheet Showing Dally PPC and Reasons

The spreadsheet used for data collection and recording has been maintained for the
eleven weeks of data collection and is presented in Appendix D. The spreadsheet for PPC

and Reason Codes is presented in Appendix E.

58

4.4 Data Analysis

The data collected was analyzed to find opportunities for process improvement. As
explained in Chapter 3, an assessment tool was developed to measure the performance of
production planning process. The Last Plamer System’s (Ballard, 2000) PPC metric and
rolled PPC (Abdelhamid, 2003) are used for measurement of performance. The detailed

steps for data analysis and its outcomes are presented in the following sections.

4.4.1 PPC Calculations

The data collected on site was used to produce a spreadsheet as shown in Figure 4.5. In
this spreadsheet, the number of activities plamed for each day was recorded and then
based on daily site visits the number of activities that was actually completed was
recorded. The daily PPC was calculated using this data. PPC can be expressed as the ratio
of number of activities completed to the number of activities plamed to be completed.
This ratio is expressed as a percentage. The average PPCs over a week was calculated to
obtain an ‘Average Weekly PPC’. This data was then used to plot a graph of average
weekly PPCs. The graphs for average weekly PPC for Building I and Building 11 are
shown in Figure 4.6 and Figure 4.7 respectively.

The overall average of the weekly PPCs for Building I is 64% and for Building II
is 75%. The PPCs for Building 11 is higher than that of Building I because all the
activities in Building 11 were performed after completing the same in Building 1. This
may reflect a ‘learning’ experience accumulated by the management and the trade crews

in the process.

59

 

Weekly PPC for Bldg l

 

09.-- ._z -2 -4 -2 ---- --_~_...__...- --t __-____ -_-_._ __.s

I
I
l
I
0.8 ._b__._ — — M i
I
I
l
I

 

 

 

 

|

I
‘ Week1 Week2 Week 3 Week4 Week5 Week6 ‘Week7 Week8 Week9 Week Week I
10 11

l
I Week '
1 Average Weekly PPC Bgildingl: 643/3 g

 

 

Figure 4.6: Graph Showing Weekly PPC for Building 1

 

F ..k i L i L V E ._ .2. ___....._.c_..,__ L” _.__.

Weekly PPC for Bldg 11 I

 

 

 

 

I
l
. l
1
1
l
l

I

I
I
o - - .1

Week1 Week2 Week 3 Week4 Weeks Week6 Week7 Week 8 Week9 Week Week 1
10 11 I

 

 

 

 

Week
_,_ I __ ___,___ __, ___ __ .__ __ __,__ __ m__A_ygrgge Weekly PPC Building H: 75%

Figure 4.7: Graph Showing Weekly PPC for Building [1

 

Figures 4.6 and 4.7 indicate that the work was completed according to plan, 64%

and 75% for Building 1 and Building II respectively. Thus, it can be concluded that

60

according to the PPC calculations, the production planning process adopted by the

company was unreliable 25% to 35% of the times.

In general, incomplete assignments could be due to over expectation from the
crew while planning and/or incapability of the crew to complete the assignment. The
underlying cause could be lack of coordination, shortage of materials, low productivity,
space constraints, etc. Regardless of the reason, the production planning process should

result in assignments that are realistic, constraint free, and achievable.

The following section will explain the use of the rolled PPC metric versus the

PPC metric as a measure of the production planning process performance.

4.4.2 PPC vs. Rolled PPC

In a homebuilding project, the construction process is divided into various activities using
a particular Work Breakdown Structure (\VBS). If an activity is not completed in one
particular day, it has to be completed the next day, thus in turn, reducing the output for
the particular day as planned. Thus, a PPC value lower than 100% for one day will affect
the output for the next day. If this scenario is considered over a week’s time, having a
variable PPC on each day, at the end of the week the overall output is much less than
planned. Thus at the end of the week, the performance assessment metric should be able

to project the inefficiency of the process such that it can be used to improve the process.

PPC measurement on a daily basis is effective, but planning for production on a
daily basis is very tedious and close to impossible considering the effort and cost that
would be required. This is especially true for the homebuilding industry, where the
competition is based on the final cost of the product, it seems impractical to require such

a tedious process. Therefore, evaluation of the construction progress and implementation

61

of the necessary changes on a weekly basis is considered more appropriate and this was
reflected in the conducted interviews. Hence, the metric used for performance assessment
should reflect the true scenario on site and should be used to improve the production plan
on a weekly or bi-weekly basis. Using an average PPC value for the period will not
reflect the inefficiency in the process correctly. This can be better explained with an

example. Consider a sample data for a week as shown in Table 4.1.

I 27 PPC rolled
asks Planned 2 3 3 3 3

asks 1 2 2 2 2
50 66.66 66.66 66.66 66.66 64

Table 4.1: Sample data for a week

 

The table above shows the tasks planned and completed for a week. It also shows the
PPC value for each day. The ‘Weekly PPC’ is an average of the daily PPCs for the week.
The data for this week shows an average PPC of 64% and this is calculated by taking an
average of all the daily PPC values.

Turning attention to the daily PPC values shown in Table 4.1 shows that on day 1,
one out of two activities were completed giving a PPC value of 50%. On day 2, two out
of three activities were completed giving a PPC value of 66.66%. The incomplete activity
on day 2 was due to the extra work performed to complete the previous day’s work. Thus,
on day 2 only 50 % of the work expected to be passed from day 1 was actually completed
and passed on. Thus, the PPC value for day 2 hides the extra work done owing to
incomplete work on the previous day. Thus, in reality, on day 2, only 66.66% of the work
passed on was completed. Multiplying PPC for day 1 with PPC for day 2 can accurately
reflect this. So, at the end of day 2 the PPC should be 33.33% (50% * 66.66%). Similarly

for day 3 the work passed on was only 33.33% completed as planned and hence the

62

output for day 3 will reduce due to the fact that earlier work had to be completed. This

can be shown by multiplying daily PPC for day3 with the previous days PPC and so on.

The multiplication procedure indicated above is the same procedure for arriving at
the rolled PPC metric described in Chapter 3. The weekly ‘rolled PPC’ is a value
obtained by multiplying the daily PPCs for the week. The weekly rolled PPC for the
sample data in Table 4.1 is 10% (50%*66.66%*66.66%*66.66%*66.66%). Thus, the
rolled PPC metric gives an accurate value for measuring the performance of the process

without hiding the rework because of incomplete tasks on the previous day(s).

The metric rolled PPC is developed from the rolled yield metric suggested in the
Six Sigma methodology. This metric focuses on revealing the hidden work or rework in a
production process. Hidden work or rework reduces the output of a process and thus it
should be eliminated. Thus, rework or hidden work should be taken into account when

assessing performance of production planning.

Figure 4.8 and 4.9 shows graphs plotted for PPC and rolled PPC for Building I
and Building II. It can be seen from the graph that the rolled PPC is much lower than the
PPC for each week. A value of zero for rolled PPC indicates that the week had at least
one day with a PPC of zero, and a value of hundred for rolled PPC indicates that the
week had 100% PPC on all days of the week. This clearly shows the effectiveness of the
rolled PPC metric in being able to measure the performance of the production planning

process in both the best and the worst cases.

63

 

 

J I—PPc I

I
I
I
I
I
I
A I--Roned PPC II
I
I
I
I
I

 

 

 

 

 

 

0 I I r r r r T r
Week Week Week Week Week Week Week Week Week Week Week
1 2 3 4 5 6 7 8 9 10 11

 

, PPC/Rolled PPC I

 

 

 

 

 

 

 

 

Week I
I__, _ _ _____ _ ____ _ __ ___________ -nr_______-___fin_r__w___I
Figure 4.8: Graph of PPC and rolled PPC for Building 1
I Weekly PPC vs rolled PPC for Bldg II I
I I
1 I
I 0.9 I
I 0.8.. __-__-_H___ ‘— I
IUO.7«——v*7i—i~w—- _.__ I
I a" I
I 3‘ 0-6 - A A“ r , ___ ,- ‘— ——EEE-—1I
E 05 I— - A A ~ ~— 7 — —— ———— ——‘——~ I II
-..- 0, I TR°"_°9EP£JI
. T v i ”' T i i + V _‘ —' fl _‘ __ _ "k I
I S 0.3+ene~AAi-+—fi——~Ai*— I
. n- I
I a. 0.2 . — e ——— — '__, u— e— I _, -—~_ I
I 0.1. w _ __,__ fin“ - I
o . - . e . . . . I
I Week Week Week Week Week Week Week Week Week Week Week I
I 1 2 3 4 5 6 7 8 9 10 11 .
I Week I
* , , _,____.-___ ___.,_.__.__- as!
Figure 4.9: Graph of PPC and rolled PPC for Building ll
4.4.3 Sigma Value

Once the rolled PPC is calculated, the sigma yield value for the process has to be
calculated as explained in chapter 3. Equations (v) to (viii) are used to derive the sigma

value for this process. The sigma yield value for the production planning process in each

64

building was calculated on two basis. Firstly, the average PPC value was used to find the
sigma yield value and then the rolled PPC value was used. The values were as follows:
- Sigma Level for Building I

By Average PPC, Sigma Yield Value = 3.24
By rolled PPC, Sigma Yield Value = 2.39

- Sigma Level for Building II
By Average PPC, Sigma Yield Value = 3.44
By rolled PPC, Sigma Yield Value = 2.85

The difference in the sigma yield value for the same building is due to the
different yield metric used to calculate the sigma level. The rolled PPC, as suggested by
the author shows a true picture of the efficiency of the production planning process. This
is due to the fact that rolled PPC metric captures the extra work done to complete the
previously incomplete activities. The difference between the sigma values in the two
buildings when we use the same metric is due to the learning effect achieved in Building
II because it always followed Building 1. As a normalized value, this sigma value can be
used for comparing different processes and can also be extended to compare between one

company to another, and even one industry to another.

4.4.4 Statistical Analysis of Data

In general, measured or collected data can be analyzed statistically in various ways to
determine areas of improvement. This is necessary so that management can identify the
aspects that need immediate attention and assign priority levels to the tasks that need to

be undertaken to improve the overall production planning process and thus improve crew

65

performance and crew-to-crew handoffs (workflow). This step of prioritizing items that
need action is advocated in the Six Sigma Methodology. Determination of high priority
actions can be achieved by performing statistical analysis of the data. Chapter 3 explained
various tools that can be used to determine the causes of incomplete assignments.

The metric ‘PPC’ can be modified to focus on the incomplete assignments by
introducing the Percentage Flamed Incomplete (PPIC) metric. The sole purpose of
developing this new metric in this research is to focus on incomplete tasks and find
causes for the same. PPIC is the ratio of total number of tasks incomplete to the total

number of tasks planned to be completed.

Run Chart or Time Series Plot
Variation is inherent in any process and it leads to unreliability in the process. The aim of
Lean Construction and Six Sigma is to reduce the variations to the extent possible. The
variations could be attributed to either common causes or special causes (Abdelhamid
2003). Common causes are those causes that are built into the process and can be
eliminated by means of reengineering. Whereas, special causes are those that create
sudden variations in the process and can be eliminated without changing the overall
process. Run Chart or Time Series Plot is a popular method to determine the special
causes and it is used here to statistically determine such causes and find means to
eliminate the cause.

In Chapter 3, run charts were discussed and a brief discussion on p—chart and its
appropriateness for this data was presented. Thus, for the PPIC data a p—chart is plotted

for both buildings in the study to determine special causes that lead to incomplete tasks.

66

 

Task Incomplete
- N
8 8
'J

<>'
-3
-3
:I
‘1 j
i :3
i

IT

9 8I

r—I

I

I

A”, 4|

 

 

.3
'o
O

 

 

 

 

 

 

 

 

 

 

 

o
01
o
.;
:==-I
— I
I-II-nn_~q
I
II.
In:
a:
x
Ia:
Io
I3
”2
'9.
I‘m

 

 

 

 

 

 

 

 

 

 

 

 

Figure 4.10: p-Chart for Task Incomplete in Building I

Figure 4.10 shows a p-chart plotted using data of incomplete tasks for Building I
and the equations listed in Section 3.3.3. The plot shows an Upper Control Limit (UCL)
line, a Center Line (CL), and the Tasks Incomplete plot. The LCL limit is zero in this
case. The UCL value for each day is varying because the number of tasks assigned for
each day is varying. On any given day, the tasks incomplete should not be more than the
UCL value.

In Figure 4.10 we can clearly see that on some days, the tasks incomplete were
more than the UCL. This is an indication of special cause variation. Such days can be
identified and reasons for incomplete tasks can be evaluated and reduced or if possible
eliminated, such that similar conditions does not reoccur. This is the main aim of
monitoring and controlling using statistical analysis. If this process is repeated on a

weekly basis or any fixed period of time, the past causes for high number of incomplete

67

activities can be determined and measures can be taken to avoid such instances in the
future. Table 4.2 shows the days having incomplete assignments more than the UCL and

reasons that caused them.

 

Day ”Flamed lnoonplete PlannedWork lBeasonsforinoorrpletim

 

 

 

 

3 2 2 D'ywdl 2 units (Pre dwelling aaprovd not grated)
Inspection Inspector approvd not obtained
Drywdling not stated
6 3 2 Engineering citrus
anssnepa'r anssrepa’roorrpletedbutnaaaprwedbyinweaor
Inspection Inspector aaprova not obta'ned
8 3 2 Insulation
Drywdl unit 1 Wanted insulation tofinish forspaoe
D'ywall unit 2 Wanted insulation to finish for space
44 3 2 Stockfioors

 

lnstalldoorsin3mits Stockingtookmostoflheday
InstdlCourtenops tirneenamtoinadlinonlyonemit

 

 

 

 

45 2 2 Instdldoorsiniium‘ts 'Lessmanpoaer
InstallComtertops Instaled in onlyonemit

46 2 2 Installdoorsin3mits Lessnmpower
lnstdlComtertops lnstdledinonlytwouit

48 2 2 lnstalldoorsinZuits appliercidr'twrplyriglttypes
lnstdlCourtenops Stoplierddr'tamyrighttypes

49 2 2 inadldoorsianits SippIierdm’tsrmynglttypes

lnstdlComtertops Sipplierddr'tsutplyrigfltypes

 

 

51 2 2 lnstalldoorsinamits aewdlaomdthebuildngoorruefingsrralljobs
Install Courtedops Iorew ail a'omdthe buildng corrpleting sn‘dl jobs

 

52 2 2 lnstdldoorsinZuits Lessrrmpower
InstdlComtertops LessrramAer

 

 

54 4 2 Plurrbing 3

..I... I
Electriw 3 Nmpmer
lnstalldoorsinZunits Desigépmblem

Table 4.2: Special Reasons Causing Variations in Building 1

 

 

 

 

 

68

 

Similarly a p-chart is plotted for Building 11 and a table is produced showing reasons that
led to incomplete assignments. Figure 4.11 shows the p—chart plot for incomplete
assignments in Building 11 and Table 4.3 shows the reasons that led to the incomplete

assignments.

t 7. .... .7 2 . ,7 fig, 7 . _ _ _._

 

p-Chart for Task Incomplete Bulldlng ll

 

 

 

Tasks Incomplete

 

 

 

 

 

,,,,,, ,sz, , _L_ m______ ___,,,,.______,LI

Figure 4.11: p-Chart for Task Incomplete in Building II

The reasons listed in Table 4.2 and Table 4.3 here can be evaluated to find and eliminate
their root causes and thus improve the process. A Cause and Effect diagram approach can
be used to arrive at the root causes. This was explained in Chapter 3, Section 3.3.3.2. The
following sections will describe various other analyses that were performed on the data to

find areas that need improvement.

69

 

   

 

 

 

 

  

 

DayslPIanned Incomplete Planned Work Beasons for incompletion
4 4 4 Truss repair completed with fault have to be redone
Building Inspection Inspector No Show
Insulation Interior Insulation could not proceed wiflrout inspection
Drywall Stocking Drywall Stocking could not proceed without inspection
6 4 2 Truss Bracrng
Building Inspection Inspector No Show
Insulation Interior
Drywall Stocking Supplier didn't respond
—"/ 3 2 Building Inspection Inspector No Show

 

 

 

 

 

 

 

Table 4.3: Special Reasons Causing Variations In Building [1

4.4.5 Pareto Analysis
Pareto Analysis is a tool used to stratify the data into groups and Pareto Charts are used
to present this data. Pareto analysis is used for this research to stratify the data into
specified groups and identify areas that need improvement.

The reasons for incomplete assignments are stratified into various reason codes
and are used to identify the reasons that caused most of the incomplete assignments. A
Pareto chart is used to present this data graphically. Figure 4.12 shows the various
reasons grouped into eight different reason codes and the number of times each reason

code has caused an incomplete assignment.

70

I Pareto chart Overall I

 

% of times

 

 

 

 

 

Figure 4.12 Pareto Chart for Reason Codes

The chart shows that ‘Pre-requisite’ is the reason code that caused most of the incomplete
assignments in the data. ‘Pre-requisite’ means work that needs to be completed before the
particular assigmnent can be undertaken. ‘Supplier’ is shown as the second highest
reason causing delays to the project. ‘Productivity’ stands at the third highest reason
causing incomplete assignments. The reason code ‘productivity’ is used when an
assignment could not be completed due to overplanning for a certain crew, or when the
crew couldn’t complete the assignment in the given duration. The Pareto chart for this
data almost conforms to the 70-30 rule, which states that 70% of the inefficiencies are

caused due to 30% of the factors.

71

The results obtained from the Pareto analysis can be used to identify the areas that
need immediate attention. In this case, the homebuilding company should focus attention
on making all the prerequisites available before making assignments to crews. The
second highest reason code ‘Supplier’ also needs immediate attention. The homebuilding
company can use such data to conduct a brainstorming session with the personnel
involved to develop a Fish Bone Diagram and find the root causes for the incomplete

assignments as explained in Chapter 3.

PPIC for different Trades

 

Overall PPIC

 

 

 

Figure 4.13 Pareto Chart for Trades
The Pareto Analysis was also used to stratify the data based on trade type and
weekdays. Figure 4.13 shows a Pareto chart that identifies the number of incomplete
assignments associated with each trade. The figure shows that ‘Inspection’ had the most

number of incomplete assignments, which could either mean that the Inspector did not

72

show up, or the inspection could not be passed most of the times. The company should

look into the reasons for such performance and improve on the process.

Another chart was constructed to show the production performance on site based
on the day of the week. Figure 4.14 shows a graph for PPC vs. Day of the week. The
nature of the graph shows that Tuesday and Wednesday is the days when the production
is higher compared to other days, and Friday shows the least production. This result
matches with a graph cited in Oglesby et a1. (1989) for production rates on different days
of the week.

#7 ,7 L, A 7, 7., __. __,”.____ ___w,fl

PPC by Day of the Week

 

 

 

 

Wed
Weekday

 

 

 

Figure 4.14: Graph of PPC vs. Weekday

4.5 Summary

The above sections detailed the steps followed for data analysis. The interviews
conducted were used to understand the production planning process prevalent in the
homebuilding industry. Homebuilding companies can use the interview data to improve
the technique used for production planning. The section also illustrated various tools
suggested for data analysis to obtain information on the performance of the production

planning process. The statistical data analysis presented, would be useful for determining

73

the shortcomings of a process and can be used to prioritize and focus on problems that
need fixing.

The next sections describe a production planning process model suitable for the
homebuilding industry. This model is developed on the basis of data acquired in the

interviews and analysis performed on the data.

4.6 Production Planning Process Model (PPPM)

This section presents a production process planning model. This model was developed
using inputs from the interviews and the data analysis. The planning model is based on
the Last Planner System (Ballard, 2000) and Six Sigma methodology. The following

sections will describe the model in detail.

4.6.1 Six Sigma Methodology for Process Improvement
As explained in Chapter 2, Six Sigma methodology is a method for process improvement
that relies on statistical methods to reduce defect rates. This methodology has been
widely used in a variety of industries. The aim here is to bring a process to a zero—defect
condition. Chapter 3 of this report explained how Six Sigma methodology is used in
conjunction with Lean construction principle in this research. Data collection was
conducted using interviews and site visits to explain the use of this methodology, as
presented in previous sections.

Six Sigma methodology is based on a 5-step process. This process is illustrated in
Breyfogle (1999), and Pande, et a1 (2000), and was briefly explained in Chapter 2 and
Chapter 3 of this report. The 5-step process is also known as DMAIC: (Define, Measure,

Analyze, Improve, and Control). The development of the PPPM follows the DMAIC

74

methodology and LPS by superimposing both on the conventional planning processes

depicted in Figures 4.1 through 4.3.

Define

The first step in process improvement in the Six Sigma methodology is to define the
process that needs improvement. All possible processes that require improvement can be
identified, and then prioritized to select the processes that need immediate attention. For
this research, the production planning process was selected for study and improvement.
Once, the process is selected as a candidate for improvement, the second step in the

methodology, which is ‘Measure’, can be performed.

Measure

The Six Sigma methodology is based on statistical methods to improve process
performance. Afier the process needing improvement is defined, the next logical step is
to measure the performance of the current process. Performance measurement can be
performed in various ways. It is important to understand the process completely and
select a measurement method and metric that is most suitable for the process.

In Chapter 2, the author reviewed the existing literature in various fields for
process performance measurement and identified the PPC metric suggested in the Last
Planner System (Ballard, 2000) as a relevant and suitable technique to achieve the goal of
this research. Also, a study presented by Abdelhamid (2003), advocating rolled
throughput yield as a metric for production planning performance assessment was
studied. The author adopted and implemented the new metric ‘rolled PPC’, which was

based on these two studies. This new metric for performance assessment was adopted for

75

this research based on its simplicity and its ability to assess the true performance of a
production planning process.

The measurement of a process metric makes it easy to understand the current
scenario and thus can be used to identify the areas that need improvement. This research
also advocates the use of the sigma level metric, as explained in the Six Sigma
methodology (Breyfogle, 1999, and Pande, et al, 2000). The conversion of the rolled PPC
to a sigma level helps in quantifying the process capability, and also helps in comparison
and benchmarking as explained in Chapter 2 and Chapter 3. The aim should be to reduce
the unreliability of the production planning process to a state where the process will yield

a 6-sigma or higher value.

Analyze

This step is used to analyze the current production planning process to identify the areas
that need improvement. Chapter 3 of this report explains the step-by-step process of
measurement and analysis. The data collection process included interviewing
professionals from 4 homebuilding companies and data collection from one residential
construction site. Previous sections of this chapter presented the data collection process
adopted in this research and its analysis. The different steps used for data analysis are
suitable for performance assessment of production planning process in a homebuilding

company.

Improve
Once the measurement and analysis phases are completed, the next step is to improve the
current production planning process based on the analysis results. This research has

developed a production planning process model suitable for homebuilding companies

76

based on the data collection and analysis. The model is explained in detail in a later
section of this chapter. The model developed intends to improve the production planning
process and it includes a system for performance measurement and analysis such that

continuous improvement can be achieved.

Control

The production planning process was selected for improvement. A model was developed
for the production planning process with the ‘rolled PPC’ measurement metric
incorporated along with a system of analysis. Once this model is implemented for
production planning, it is very important to keep the process in control. Continuous
measurement, analysis and improvement cycle has to be achieved to keep the process in
control. The measurement and analysis steps help in identifying the unreliability of the
production planning process. It is important to find solutions to reduce this unreliability.
The model for production planning can be modified to adapt any inconsistencies and
achieve a higher sigma value for the process. Control is thus necessary in the overall

improvement process.

4.6.2 Last Planner System
The Last Planner System® (LPS) provides a framework for management and workers to
plan and control daily production assignments (Ballard 2000). Chapter 2 explains the

LPS system in detail.

Should-Can- Will-do Tasks
The LPS model, as shown in Figure 5.1 has 4 categories of tasks: Should, Can, Will, and
Did. The ‘Should’ tasks are those tasks identified on the Master/Phase schedules and

must be completed at a specific time. This schedule is based on structural, technical,

77

regulatory considerations, and sequencing. This step also takes inputs from current status
and forecasts. The LPS model suggests a Lookahead process, where all the ‘Should’ tasks
that are planned for a given period of time are scrutinized, such that all the work is

declared constraint free.

     
 

& Phase
Schedule

  

.0... .'e.. seaplane-*eee-eeee: c N
.°° Current '0, : Selecting, :
5 status '; ' sequencing, & “man...”
'g & 5 : sizing work we 3'.“ L 00th e a d 3..
’0, forecasts ,o' i think can be 5 '. :
. ’° done :

Mlfi WILL

O ...
...I... III-BIIIIIIIIIIIIII'

 

IIIIIIO*IIII‘ 5 Selecting,
5 Make : , Workable :- g sequencrn

III.
.....II II.....
0’ '0

 

 

 

 

 

. . . work : ‘ ' g
.. loofndtn1runncn:III‘.:"§ reaKty by, E‘IIEELECkJO%a: : gisfiZdrug :: VVEN1(
IIIIIIIIII E screenin E eeenneee. 5 work 5... Plans
: g, E 5 we know 5 0.... ee’
5 & 5 W3 5 E can be 5 '
: g : done 5
Construction
Figure 4.15: Last Planner System Model DID

(©Lean Construction Institute)

A constraint analysis is performed to find any constraints in completing an upcoming
task. These tasks are termed as ‘Can’ do tasks. All constraints should be removed for the
tasks before they enter into the workable backlog. The workable backlog is a set of tasks
that are constraint free and doable. These tasks are sized properly and all prerequisites for

the task should be available at the time of its execution. The foreman responsible for

78

performing an activity should only commit to tasks. These are known, as ‘Will-do’ tasks.
The tasks that are actually completed are known as ‘Did’ tasks.

The LPS system uses the PPC metric for performance measurement. The PPC
value can be calculated by dividing the number of tasks actually completed by the
number of activities committed to. The number of assignments completed to those that
should be done is also tracked.

The production planning model developed in this research is based on the LPS
Model and the Six Sigma Process Improvement Model. The following section describes

this Production Planning Model in detail.

4.6.3 Production Planning Process Model Description
One of the objectives of this research was to develop a model for production planning
process. This model is based on Six Sigma Process Improvement methodology and the
Last Planner System. Both of these are explained in the previous two sections. The
production planning process model is a combination of both these systems, and is tailored
to meet the needs of homebuilding industry. Information of the production planning
process prevalent in the homebuilding industry was necessary to develop the new model.
This was facilitated by the interviews conducted. Opinions and concerns received in the
interviews were addressed and have been incorporated in the Production Planning
Process Model.

Production planning in most homebuilding companies starts from a standard
template for a particular type of home (Single home, apartment, condominium, etc.).
Once the home design and plans are selected for construction, the standard template can

be modified into a specific template for that particular community or set of homes. The

79

company then submits plans and details for permits and approvals for each home. This

determines the start date for construction. The availability of resources (trade contractors,

 

 

 

 

 

 

Design

Standard Template /
Home Plan; Sales;
pt \

 

 

Specific Template for

Each community
Art Dates Based on Permits;
\ Resource Availability

Specific Schedule for I Should I
Each Home/Unit ' . .....................

I Specific Schedule I
% , from all Communities :

I Centralized Schedule System :
---------------------- /rade Contractor, Supplier,
4 Inspector, Superintendent,
PM Involvement

 

 

 

 

 

 

 

 

 

 

 

 

 

Short Term Lookahead CAN
Plan

 

 

 

 

 

 
    

Constraint Analysis and
Make Work Ready

Current Weekly
Progress Update

  
 

\/

 

 

 

 

 

 

 

 

 

   
 

 

 

     

. I WILL I f
Weekly Production e ME A SU RE)
Plan rolled PPC/ I
Sigma Value
, U 7 .
@ Tasks Completed Data Analy 515/
Charts/ Pareto

 

 

 
 

 

 

   

Analvsis

 
 
 

DID
ANALYZE

  
 

 

Brainstorming/
Reengineering

 

   

 

 
 
 

IMPROVE

  

 

Figure 4.16: Production Planning Process Model

80

suppliers, etc.) has to be considered also for obtaining a specific schedule for each home
or unit. This schedule represents the Master schedule in the LPS. Thus, this can be termed
as the ‘Should’ do tasks. In case of large companies having multiple communities it is

advantageous to compile a centralized schedule based on inputs from all communities.

This helps in scheduling resources centrally and avoids resource conflicts between
production plans of different communities. Thus, some constraints are removed right at
the beginning. The specific schedule for each home should be scrutinized in a Lookahead
Plan. This plan is a breakup of the master schedule and it is for a shorter period of time.
This process involves inputs from the trade contractors, suppliers, superintendents, and
project managers. The trade contractors and suppliers agree to certain tasks they think can

be done. These tasks form the ‘Can’ do work.

The project manager’s, and superintendent’s aim should be to make the ‘Can’ do
tasks doable. The set ’of ‘Can’ do tasks have to be further analyzed to identify the
constraints to completion of tasks as planned. A complete Constraint Analysis should be
performed so as to make the work ready for the trades. This involves making available
the prerequisites for each task, and proper sizing of tasks, thus making the work doable.
Only tasks that are free from constraints can pass through this process as assignments to
be completed. The trade contractors, superintendents, and the foremen should commit to
these assignments, and these forms the Weekly Production Plan. These are tasks that
have been committed to, and thus are identified, as ‘Will’ do tasks.

The weekly production plan should be passed on to each and every supplier
involved. These tasks are then completed during the week. As the day for performing

each tasks approaches, the superintendent should make sure that all constraints identified

81

are completely removed. In case of delays or weather calamities, the resources should be
assigned to new production tasks that had been identified as doable but not assigned yet.
Such a workable backlog should be kept available with the Weekly production plan.

At the end of each day or week (depending on resource availability) a planned
versus completed sheet should be filled out. This could be in the format shown in Figures
4.4 and 4.5. This forms the ‘Measure’ phase of the model. The simple form can be used
to keep track of the daily progress, and to measure the PPC for each day and hence the
weekly rolled PPC. This can be converted into a Sigma level. The reasons for
incompletion of tasks should be categorized to identify major causes of delay. A
complete data analysis can be performed as explained in Chapter 4 of this thesis. This
forms the ‘Analyze’ phase of the model. The results from the analysis can be used to

conduct a brainstorming session and identify the root causes for the problems.

The analysis and brainstorming sessions could lead to a complete re-engineering
of the process or modifications to the process. For example, if productivity is the highest
reason causing delays, either the sizing of tasks could be faulty, or the trade workers do
not perform well, or it could be that some unidentified constraint is restricting the trade to
be more productive. A brainstorming session with the trade workers, suppliers, and
superintendent could lead to the root causes for low productivity. Other statistical
analysis, as explained in Chapters 3 and 4, could lead to various problem areas that can

be identified and eliminated or reduced. This forms the ‘Improve’ phase of the model.

It is important to keep the whole process under control such that variations do not
occur, and thus reduce defects in the process. Superintendents are well positioned to do

this. Once, the whole process is under control, the superintendents should maintain the

82

whole process of developing a doable weekly plan, measuring, analyzing, and improving.
The superintendents in turn can be motivated by rewarding them with some kind of
recognition on a regular basis for continuous improvement. Larger companies can create

competitions between the superintendents by comparing the sigma value for each site.

4.6.4 Validation

A ‘light’ validation was conducted with the objective of getting inputs from
homebuilding companies with respect to the applicability, adaptability, and completeness
of the model. Personnel from two homebuilding companies validated the Production
Planning Process Model developed in this thesis. The validation was done by an
interview with the personnel from each of the two companies. The following section
summarizes the result of the validation process.

Both the interviewees appreciated the layout of the model. According to them, the
model could be readily applied and put into use with minor additions, rather than
changes. Market sampling, and production absorption studies were recommended as an
additional input to develop the specific template for each community. Sales being one of
the foremost drivers of the whole process for most homebuilding companies, it would be
a good idea to include the sales department in the process. The sales input could come at
a stage before the specific schedule for each home or unit is developed. The various
options provided by the companies to the customer is communicated through the sales
department. Hence, ‘Sales’ should have inputs into the process at various levels, like the
short term lookahead plan development phase. Both the interviewees were of the opinion
that the sales personnel represent the buyer’s viewpoint and thus it is a key to achieving

an overall value.

83

The format proposed for data collection on site was well received due to it’s ease
of use and simplicity. The reason codes list could include few more reason for delays
such as the sales change or customer’s option changes. both interviewees are of the
opinion that the model is ready to go on a general basis and then can be modified by
different companies based on the company’s organization. The practice of daily PPC and
rolled PPC calculations is possible only if the format is simple and if it is not time
consuming. On seeing the format developed by this research, the interviewees accepted

that it is “as simple as it could get”.

Neither of the companies does the Analyze, Improve, and Control phase of the
process. But they are in agreement that this is a vital portion of the whole process and that
it would “complete the circle”. The brainstorming session suggested in the model could
be incorporated in the weekly meetings. This is possible only if certain agenda or reason
codes were to be tackled at a time. The model is considered to be a control tool and
would thus help reduce the amount of resources spent on traditional “fire-fighting”

approach.

The model seems to be complete in most terms, except to including inputs from
the sales department for the planning process. The charts developed for the case study
was shown to illustrate the use of the analyze phase in the model. Both the interviewee
were of the opinion that if such tools were made available to the superintendent, it would

ease the communications between the office management and the field personnel.

The validation process gives insightful comments on the applicability and

adaptability of the production planning process model in homebuilding companies. The

suggestions received in the interviews could be applied to the model and this process of

improvement of the model should continue throughout the model’s life.

4.7 Summary

This chapter describes the Six Sigma methodology and the Last Planner System. The
above sections describe in detail the steps involved in the production planning process.
Homebuilding companies can use the model developed and aim for continuous
improvement. The Production Planning Process Model is presented and explained in
detail. The validation interviews were summarized in this chapter. The next chapter will
summarize the whole research report and present conclusions, contributions, and

limitations of this research.

85

Chapter 5

Summary and Conclusion

86

Summary and Conclusion

5.1 Overview of Report

Residential construction is booming and the projected growth of the industry calls for a
more scientific approach to develop tools for production planning. As indicated in
Chapter 1, residential construction companies in the United States are building homes
with a manufacturing-like production or what is commonly known as production
homebuilding. The goal of this research was to develop a tool for performance
assessment of the production planning process in the homebuilding industry. Literature
review in this area was available, but mostly for automobile and other industries.
Literature for production planning in the construction industry was limited to general
construction and there were no formal production planning tool for the residential
construction sector.

Interviews were conducted with professionals from the homebuilding industry to
understand the prevalent production planning process. Data collection was conducted to
demonstrate the methodology developed in the research. The data collected was analyzed
to find areas of improvement. A production planning process model for homebuilding

industry was developed using the results of the research.

5.2 Research Goal and Objectives

The goal of this research was to assess the effectiveness of the production planning
process in the residential construction industry. To attain this goal, the production
planning process prevalent in the industry was documented by interviewing homebuilders
and through site visits and observations. A framework for production planning

performance assessment and identification of process improvement opportunities was

87

also developed through a literature review and adaptation of tools and techniques used for
measuring and analyzing production performance. The developed framework was
implemented using data collected during field visits. Implementation results were

analyzed to identify process improvement opportunities.
5.3 Contributions

Process Maps

This study contributed to the understanding of the prevalent practices for production
planning in the homebuilding industry. Three different production planning processes
were documented based on practices in companies of different sizes. This documentation
will be a good resource for future studies in this area to understand the production

planning process in homebuilding companies.

Performance Assessment Tool

Another outcome of this research is a tool for production planning performance
assessment. An easy-to-use metric, rolled PPC, was developed for production planning
performance assessment. The ‘rolled PPC’ metric can be modified and tailored to

measure performance of various other processes.

A format for data collection was developed and is shown in Figure F .1 (Appendix
F). This format was developed for data collection and was modified with inputs from site.
The format is simple and easy to use. Figure F2 (Appendix F) shows a format for
calculations of PPC and ‘rolled PPC’. This also includes categories of reasons that caused
incomplete activities. These categories can be used as is, or can be modified to
accommodate any factors that are not included or if it misrepresents the scenario on site

for any homebuilding company.

88

The set of tools mentioned can be used for analyzing the data collected on site for
a homebuilding company. These tools can be used separately or together to find areas
needing improvement. These tools can be implemented and modified to adapt the specific

needs of the company.

Production Planning Process Model

A production planning model was developed in this research based on the knowledge
assimilated through literature review, interviews, data collection, and data analysis. The
Production Planning Model is shown in Figure 4.15 of this report and also presented in
Appendix G for easier reference. This model accommodates the production planning

performance assessment metric and the analysis tools.

This model focuses on continuous improvement. The process of Constraint
Analysis included in the model can be improved based on the analysis conducted and this
cycle has to be followed continuously to keep the production planning reliable and
efficient. The model can thus be used to achieve a reliable production plan. This model
can also be used for improving other processes in the company after specific

modifications.

Synergies between Lean Construction and Six Sigma

Six Sigma and Lean Construction are both powerful tools. Lean Construction seeks to
achieve reliable workflow by eliminating inefficient work and Six Sigma focuses on
reducing variability in a process using scientific methods to identify areas that are root
causes for the variability. The combination of both tools can lead to a very useful
methodology to improve any process. This integration was first suggested by Abdelhamid

(2003). This research further developed the ideas in Abdelhamid (2003) by applying the

89

power of both methods to the production planning process in residential construction.
Both Lean Construction and Six Sigma were used in conjunction to develop the
Production Planning Model. The aim was to develop a system that helps in identifying
defects in a process, eliminating it and thus create a zero-defect process. Continuous
improvement of the process is also important to keep the process in control and make it

efficient.

5.4 Research Limitations

This research is focused on homebuilding companies. This is due to the lack of research
conducted in the areas of production performance for homebuilding companies. Although
the results can be useful for different industries, it is primarily suitable for the residential
construction industry.

Interviews conducted for this research were limited to five homebuilding
companies due to time and budget restraints. All of these companies were in the vicinity
of Michigan State University, Lansing, MI. The small sample size in the research is
consistent with the research goal, which was not to generalize any of the findings. Rather,
the research was concerned with understanding the current production planning process
prevalent in the homebuilding industry, developing a framework for assessing the
performance of the production planning system, and identifying opportunities for
improvements. There was no case study conducted to implement specific improvement
opportunities to find root causes.

The following section presents some pointers that are the by-products of this
research. It is an attempt to discuss some information gathered, or developed during the

whole process of doing this research.

90

5.5 Discussions
This section presents information gathered by the author during the course of this
research other than the conclusions. The author intends to share this knowledge

accumulated.

Research Area

The author started off this research process by doing literature review in the field of
interest. Being a novice in the field of research, while reviewing various research work
the author would very often be tempted to use a particular methodology or tool used to
solve a problem and then look around for problems that could be solved using the
particular method. Later on, with much needed guidance, theauthor learned to curb the
temptation and followed the reverse by first setting a goal and then developing a list of
objectives to achieve the goal. Literature review was in the specific area made it easier to
find similar research and methodologies used to achieve similar goals. The author shared
this knowledge with some of his fellow graduate students who were new to this
experience too and they found it useful. Thus, the author intends to share this with new

researchers reading this report.

Data Collection

For the purpose of this research the author conducted interviews and site visits. The
author interviewed two personnel each from five homebuilding companies, one from the
home office and other from the field. This was done with a purpose to capture ideas from
people involved at a macro level and at a micro level. The author feared a biased

response from the interviewees either to project a good image of the company or due to

91

the fear that upper management would see the report. To avoid a biased response, the
author had to convince the interviewee about the purpose of the interview.

For data collection the author performed site visits over a period of time. During
these site visits the author gathered information from the site superintendent. As the
upper management directed this, the superintendent was under the impression that it was
an audit and the reports would go back to the upper management. One way to tackle this
problem was by making it clear that it was for a research the author was doing as part of
his graduate program. The author spent time bringing the superintendent into confidence
by encouraging informal discussions. Besides, the author had to verify the information
given by the superintendent by visiting the work location where work was in progress or
where it was planned. Talking to site crewmembers also helped to verify the information.
The reasons for incomplete activities generally were explained more in detail by the
crewmembers. The superintendent tried to hide his faults, if there was any, that lead to

incomplete activities or work not performed as planned.

Analysis

The goal of this research was to develop a framework for assessment of effectiveness of
production planning in residential construction. Hence data collected on site was mainly
analyzed with the objectives in mind. The data could be analyzed in different ways to
find other facts. One such output was the overall PPIC for different trades. As shown in
Figure 4.13, the highest PPIC was for Inspection. Inspectors from the building
department are generally contacted at various stages for inspecting the code compliances.
The site superintendent generally schedules the inspector to arrive based on the progress

of the particular trade before the inspection. Superintends are not confident of the

92

production plan to an extent where he or she can schedule an exact date for the inspection
much ahead of time. Thus, this is generally done on a short-term basis because the site
superintendent would not want to arrange an inspection and cancel at the last minute
because the job was not complete. This hampers the progress because generally on a
short-term notice it is difficult to schedule an inspection, and as the inspection is
necessary the trades cannot proceed until it is done and satisfied. If on the other hand the
superintendent schedules an inspection and then have to cancel it, the inspector can be
rescheduled only after a longer time. Sometimes, as a deadline is set before the
inspection, the trade contractor might tend to ignore on quality to complete that portion of
work before inspection. Thus, it is a difficult decision to make for the superintendent. If

the production plan were more effective this would have been easier.

5.6 Research Conclusions

This study sheds light on the necessity of more research for the homebuilding industry.
The study acknowledges the effort and work done by various researchers in this field, but
the requirement of more research in production planning is addressed.

Interviews conducted reveal that companies generally perceive higher production
yields than actual. This is due to the traditional approach of looking at productivity. As
explained in this thesis, the rolled PPC and sigma level approach to assess production
planning reliability captures the extra work required to complete previously incomplete
work. The aim should be to reduce and eliminate the root causes leading to an incomplete
activity. The process maps also supports the need for more a robust planning model in

the homebuilding industry.

93

The site visited by the author for purpose of this research was firnctioning at a low
sigma level and there was certainly a need for major improvement in the process.
Identifying such problems early could save a lot of resources. The analysis revealed that
task pro-requisites, coordination with suppliers, and overestimating productivity were
major causes for unreliable production planning. For the site visited, Inspection,

Carpentry, and Engineering trades caused the most incomplete activities.

5.7 Recommendations for Future Research

Research plays an important role in developing any field of study and related industries.
The residential industry in the United States is growing at a very fast rate and this growth
can be leveraged with proper research supporting it. It is very important for homebuilding
companies to implement the right processes to attain the maximum returns for clients.
Future research is necessary for the residential construction sector for developing
processes that are efficient and reduce overall waste.

This research focused on the production planning process in residential
construction. Future research can focus on different processes. Some areas that can be
studied for firture research in residential construction sector includes: Design Process,
Pro-Construction Process, and Construction process. Generally, in residential
construction, sale of homes controls the whole cycle of construction. Improvement in the
sales process for residential companies can be an area of research that will be of interest
to all homebuilding companies. In addition, customer satisfaction is a very important
factor for popularity of homebuilders. The Warranty Process, which is a part of all
homebuilding companies, could be another usefirl area for future research to improve

process.

94

The ‘rolled PPC’ metric used in this research can be further studied and analyzed.
This metric can be modified and future research could include improvement of this
metric for production planning process performance assessment. Future research could
identify other processes and tailor the metric to suit the necessities of these processes.
The metric can also be modified and tailored to be used in other sectors of the
construction industry. Researching more areas that uses different metrics for

measurement can develop new metrics for process performance assessment.

Research can be conducted to implement and improve the Production Planning
Model developed in this study. New models can also be developed based on this model
for other processes. This research has initiated a study to use Six Sigma and Lean
Construction in the residential construction industry. Future research can adopt more
tools from the Six Sigma Methodology and Lean Construction. The combination of the
two methods promise to be a very powerful instrument in improving processes in the

construction industry.

5.8 Summary
This chapter concluded the thesis by giving an overview of results, conclusions and
recommendations. It also reported the limitations of the research. This chapter also
includes future research areas that can be inspired by this study.

The importance of research to improve production planning process in residential
construction was first described in the literature review and was re-emphasized in the
interview responses. This thesis concludes by describing the contribution of the

production planning process model as follows:
0 Use of the Production Planning Model for a more reliable production plan

95

0 Use of Six Sigma tools in Lean Construction Systems to obtain a powerful
production planning process
0 Use of ‘rolled PPC’ metric for process performance assessment

a Possible improvement in the production planning process in homebuilding

96

References

Abdelhamid, T. S. (2003). “Six Sigma in Lean Construction Systems: Opportunities and
Challenges. ” Proceedings IGLC-l 1, Aug. 2003, Blacksburg, Virginia.

Ahmed, S. M., Mitrani, J. D., Azhar, S., and Roldan, J. (2001). “Development of a Risk
Management Model for Florida Contractors and Subcontractors. " Department of
Community Affairs, State of Florida.

Allen, E. (1999). “ Fundamentals of Building Construction Materials and Methods.”
Wiley, New York, NY.

Allen, B, and Thallon, R. (2002). “Fundamentals of Residential Construction." Wiley,
New York, NY.

Ballard, H. G. (2000). “T he Last Planner System of Production Control" University of
Birmingham.

Bashford, H. H., Sawhney, A., Walsh, K. D., and Kot, K. (2003a). “Implications of even
flow production methodology for U. S. housing industry. ” Journal of Construction
Engg. And Management May/June 2003.

Bashford, H. H., Walsh, K. D., and Sawhney, A. (2003b). “Relationships between system
loading and cycle time in a residential construction production system. ”
Proceedings of Construction Research Congress, March 2003.

Bassioni, H. A., Price, A. D. F., and Hassan T. M. (2004). “Performance Measurement in
Construction. ” Journal of Management in Engineering, ASCE, April 2004.

Bemades, M. M. S., and Formoso, C. T. (2002). “Contributions to the evaluation of
production planning and control systems in building companies.” Proceedings
IGLC-lO, August 2002, Gramado, Brazil.

Breyfogle, F. W., Cupello, J. M., Meadows, B. (2001). “Managing Six Sigma: A practical
Guide to understanding, Assessing and Implementing the strategy that yields
bottom line success.” Wiley, New York, NY.

Breyfogle, F. W. (1999). “Implementing Six Sigma: Smarter Solutions Using Statistical
Methods.” Wiley, New York, NY.

Chitla, V. R. (2003). “Performance assessment of planning process during manufactured
housing production operations using Lean production principles.” Michigan State
University, E. Lansing, MI.

Cox R. F., Issa R. A., and Ahrens D. (2003). “Managements Perception of Key
Performance Indicators for Construction. ” Journal of Construction Engineering
Management, March/April 2003.

Evans, J. R., and Lindsay, W. M. (2001). “T he Management and Controls of Quality-
Fifth Edition ” South-Westem, Cincinnati, OH.

Goldratt, E. (1992). “T he Goal, 2"" Edition, Revised. " North River, Great Barrington,
Mass. '

97

Harris, B. B., Ioannou, PG. (1998). “Scheduling Projects with repeating activities. ”
Journal of Construction Engg. And Management July/Aug 1998.

Hinze, W. J. (1998). “Construction Planning and Scheduling. ” Prentice Hall Inc., NJ.

Hopp W., and Spearrnan, M. (1996). “Factory Physics: Foundations of Manufacturing
Management. " Irwin/McGraw Hill, Boston, 668 pp.

Howell, G., and Ballard, G. (1998). “Implementing Lean Construction: Understanding
and Action. " Proceedings IGLC ’98, Guaruja, Brazil.

Koskela L. (1992). “Application of the New Production Philosophy to Construction.”
Tech. Report No. 72, CIFE, Stanford University, CA.

Koskela, L. (1999). "Management of Production in Construction: A Theoretical View.”
Proceedings of the 70. Annual Conference of the International Group for Lean
Construction, University of California, Berkeley, CA. p. 241-252.

Koskela, L. (2000). “An exploration towards a production theory and its application to
construction.” Helsinki University of Technology, Finland.

Linderman, K., Shroeder, R. G., Zaheer, S., and Choo, A. S. (2003). “Six Sigma: A Goal-
theoretic Perspective. ” Journal of Operations Management, Elsavier Science,
21,193-203.

Love, A. Thomas (1995). “Scheduling Residential Construction. ” Washington: Home
Builder Press, WA.

NAHB (1993). “Cycle-time reduction in the residential construction process. ” National
Association of Home Builders Research Center, June 4, 1993.

Pande, P. S., Neuman, R. P., and Cvanagh, R. R. (2000). “The Six Sigma Way: How GE,
Motorola, and Other Top Companies are Honing their Performance.” McGraw-
Hill, New York, NY.

Patrick, C. (2004). ”Construction Project Planning and Scheduling. ” Pearson Prentice
Hall, Inc., NJ.

Peterson D. (1993). “IS-464”, Course Notes, Colorado State University, Fort Collins,
CO.

Picard, H. E. (2002). “Construction Process Measurement and Improvement. ”
Proceedings IGLC-10, Aug. 2002, Gramado, Brazil.

Seidel, A. W. (1999). “Production Planning Methodologies for Large Homebuilding
Firms.” Michigan State University, E. Lansing, MI.

Syal, M. (2003). “BCM-823: Advanced Project Management. " Course Notes, Michigan
State University, E. Lansing, MI.

Willenbrock, J. H., Manbeck, H. B., and Suchar, M. G. (1998). “Residential Building
design and Construction. " Prentice Hall Inc., NJ.

98

Appendices

99

Appendix A

Interview Questionnaire

100

INTERVIEW QUESTIONNAIRE

Interviewee’s name, Position:
Number of years with the company:

Background (demographics)

Company name, Location:

Brief about the company:

Number of subdivisions currently work undergoing:

Company long term plans

Goal for the company in 5 years:
Goal for the company this year:
On what basis is the goal set?
How do you predict the demand?

Did the company fulfill demand/ goal for the previous year?

Subdivision plans

Name one subdivision you would like to discuss regarding planning:
Goal for this subdivision:

When do you start planning for one subdivision?

When do you start planning for a single house?

Do you have a schedule (representation) for a subdivision/single home?

Who/which department is responsible for planning?

101

Construction process
0 What work is done by the company and what work is subcontracted?
o On what basis is a supplier selected?
0 On what basis is a subcontractor selected?

0 How many subcontractors for one single task (eg: framing) in this subdivision?

Planning process

0 What is the procedure for planning the activities for a single home?

a Who are involved while planning these activities?

0 How are trades coordinated when working on multiple homes?

0 How is the plan passed on to the members involved?

0 What softwares are used?
a Who decides the day to day activities to be completed?

0 What parameters are considered while assigning activities on daily basis?

0 In general, what percentage of assigned work per day is completed as per plans?
0 Do you have a format for recording planning performance?

0 How is the plan changed in case of an incomplete assignment the previous day?

102

0 Do you maintain records of day-to-day activities?

0 Do you use these records for future planning?

0 Are progress meetings conducted on a regular basis in this subdivision?

a What are the major causes of delays in this subdivision?

0 If subcontractor’s performance is the cause how do you handle it?

o If supplier delays are the cause how do you handle it?

o What time contingency is given in a plan for environmental delays?

0 What time contingency is given in a plan for inspection and permit delays?

What are the incentives of completing a house in time?

Opinions
0 Have you worked in other homebuilding companies? (Yes/No)
o What was good/bad about other companies with regards to production planning

process?

0 What according to you is a good practice for planning in a homebuilding

company like yours?

103

o What factors influence the planning process? (Also most critical and why?)

0 Is it better to have a general plan for all homes or specific plans for each home?

0 At this juncture, do you think innovation/research will help to improve the

production planning process? What steps are your company taking?

104

Appendix B

Calculations for Sigma Quality Value

105

Calculations for Sigma quality value
The sigma quality level for a process can be calculated using the following set of
equations. For detailed discussions on these equations see Breyfogle (2002). These
equations are essentially for a process using rolled throughput yield metrics. The

modified equations used for this research using PPC value is shown thereafter.

Sigma level calculations:

- Firstly, the rolled throughput yield is calculated from the yields of different sub-
processes as follows:

m

YRTzl—I. lYi
[:

Where, YRT = Rolled throughput yield

Y, = Yield of sub-process ‘i’

- The rolled throughput yield is normalized by finding the geometric mean as
follows:
Ym= W
- The normalized yield is then converted to Defects Per Unit (DPU) using the
following equation:
DPUnorm = -ln (Ynorm)
- To determine the sigma quality level, also called as Zbenchmark, for the process, the
following equation is used
Sigma quality level = Znorm + 1.5
Where, Zlnorm is the standard normal value corresponding to the DPUnorm

found using the standard normal table.

106

Modified calculations using PPC

Firstly, the rolled PPC is calculated from the PPCs of different sub-processes as

follows:
rolled PPC = 1.1m PPC-
i =1 '

- The rolled PPC is normalized by finding the geometric mean as follows:

PPCnorm = W
- The normalized PPC is then converted to Defects Per Unit (DPU) using the
following equation:
DPUnorm = -ln (PPCnorm)
- To determine the sigma quality level, also called as Zbenchmark, for the process, the
following equation is used
Sigma quality level = Znorm + 1.5
Where, 2mm, is the standard normal value corresponding to the DPUnorm

found using the standard normal table.

107

Appendix C

Interview Summary Table

108

 

  

 

00636030 £5 0_ 0:00:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 

 

 

     

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ym 000E000 .000: 0000 02 N 0m000>0 00 00 N 53 0.00“ 0605 000 00“— 002000000030 >00E 30: cm
200030.00 $5030 .02me .0000. 003000 .0000 00.00.00 0200000003 0 0_ £000 .003 00 or
90000020.. 5:030 003.00 .0000 02200 .0000 0200.00 02.0030 0 m_ £000 “003 00 or

02000000030 m_
02000000030 0_ 0:0; 05 _0 305. 02000000030 2 0:03 05 n.0 .005. 0:03 .003 000 >000E00 05 >0 0000 m_ 0003 300>> 5
“0000900003 «0000200005 050003 .2 0.9000000. m_ «00800000 003E003 9

l4 0E0; 0. 5002220030
m0> 00> 0 02 3000000000000 2300000 0 0>00 30> 00 2.

020300030

“€000 €800 £00E r - 9.095 >__00_0>._. 0000 m>00 218 00300 0603 0 ..2 m0_000_0 00% 30> 00 000>> 3
020.3 0000300000 2 .000 m>00 8.? 0232003 000 02 m0_000_0 00% 30> 00 000>> m...
mmEoc 0 000.00 0 02220030 £5 02 _000 me

5000. 0:00 00

5.4. 52 $3020 2 00.: 0.303 30> 0003603 000 0E02

000 03039
030E000 «00 50 00> 00> 05 02 0090000000 :52 >000E00 05 ED 3
an: mz<z co 2,3
20:02 £05 002003006 00 00:005Eoo 00:03:60 .5305 300E>0_0Em 000E00 05 02090 30> 00 30: m
0009000

00000000 003090000 .0200 000.000 .0000: “00:05. 560000 m.>000Eoo Em .00m 05 m_ 07.00 3003 00 w
09000 cm mmEoc m: 000 £03 00 E00 00:8 .000 n
moEoc 8.3 000.00 cm 900 m 0_ 00 E00 05 02 .000 0
002036030 v 002220030 v 9690003 0:03 >=00030 002220030 00 009032 N
>00 E00 05 «3000025 V
jg u: om< 3”: RE 8:83 .95: :0 :50 U
00> 9m 00> n >00 E00 05 £05 200 .6 $00.32 w
0000300000 n_> d> 000500 .9000 0.0032020. F
NV 5? S N>x 00030030 a.

 

 

 

 

30%

109

 

 

.0..0.0E

000.0080 0.00:30

00.03.0030

 

 

 

 

 

 

 

 

 

 

 

.050 0...... 0.00.50.00.00 .00030 ..0500>> .00.00>> .03 9.3000 .00 0000. .. 0.0.0. 0>0_00 .0 000300 .0..0E 00. 0.0 .00>> vm
00.0.>.0030 0.0.0. 0.000
>200; >300; 0.3.0. 0 00 00.030000 0m0..00E 00050.0 0.4. mm
02 oz m0.000.0 0.0.3. .0. 00.000. 0000. 003 30> 0n. mm
mm; 00l> 00.._>..00 00-0.- 00 .0 00.000. 0.0.0.08 30 0n. 5
. .00 0 00 .003. . 0.00 05.0 2% 00 03030. 00. .000.0 .000
0030 05 00.00.. 0.0050300 000 0:00 0.03.00 .00000.0..003M— 0.0.0.0000. 00 .0 0000 0. 00m0000 00.0 00.0. .50... on
00005000
02 02 m0.000.0 00.0.80. .0. .0E.0. 0 0>00 30> 0n. 0N
000. .0 00 00.0. E00 0. 00
$3 $00 .00 0.0.5 0000.000 .0 0m0.000.00 .003 ..0.000m 0. mm
.0E00 0E00 05 0. 0000.. .050 .0 0.000 0.00 00 00...>..00
02.00.0800 000 00000 0000.. 9.008%— 0000.0..0.0. .00000 9.09000 0:03 00.00.0000 0.0 0.0.00.0.00 .00>> R

00.0.mE00_

 

 

 

 

 

 

 

 

 

 

 

00030 00.0.0E0. .n.> 00... 05.0 0. 2., 00002000030 00 o. 00...>..00 >00 0. >00 00. 000.000 00>> mm
02 02 0003 0.0 00.05.00 .00>> mm
00030 0.0..

0050.0 .0 0.000. >:0n. .>:0000.00 00030 00290.,
0.00 000000 0.0 00.00 00.0 0.300002 .00.0> 0.00.000. 00.0.00 000000 00.0 00. 0. 30... vm

.0030 0. 90:0... 0000.. 00:300. 00 000.. 0. 0:... 000.00 0.0...3E
.0. 00.0.0000 0.0 00.000 05.0 .005. 000. 020 .0000>00 0. 0.003 0 0.300000 00 00.000; 0002. 00.00.0500 0000.. 0.0 .50: mm
005 000 00> 00002000030 00...>..00 0000. 9.0000 0:03 00>.0>0. 0.0 00.5 mm

0.500.000 0:588 0:00.230 0:588,

0.0.00..000030 .0 030.305. .0000..00x0 0.0.00..000030 .0 030.305. 000000008, 0E00 060.0 0 .0.
00 00000 0.000 0000 .0. .0000000.00 ...u. 00 00000 00.00 0000 .0. .0000000.00 ...“. 000.500 05 m0.000_0 .0. 0.300090 00.0. .00>> _,N
N. N>x S N>x 002.0030 ...

 

 

 

: om<

 

 

110

 

     

 

 
  

 

 

 

 

 

.000..00E 0. 0000300.0 00000.0 .00E0>0.00.. 03000000 00.0.0. >000E00 .30> 0.0 000.0 .00>> 0000090
00.0000 00 0... .30 .00..0>000. .0 00000.0 0. 00.000. >..0300. 00. 0. 0005000 000000 00003020 00. 0>0.0E. 0.0.00 ...3
.0000. 0 30:0. .00 0000 >0000.00 00> .000 00. 00:00:00.0 00000.0 00. 00.0000 00.0000..00..0>000. 0.0.5 30> 00 0.3.003. 0.0. .< 0..
00.00 0000 .0. 000. 00.00 0
00300 >0 0030... . 00.0000 00.0000 .0 000.00 :0 .0. 00.0 .0.0000 0 0>00 o. .0000 .. 0. 3
.0030—
0E00 .0 >._.00.03000< 00000.0 02.00.00 000000.00:
0.0030 00.0... >00. 0. 0000000000 0030 0.0030 .0003 .0 00000 000.00 0.0.030 .0. 00000.0 05000.0 00. 00002.0. 0.0.00. .00>> 00
0000000000 .00.. 0000.00_.E 00.003 .003000030 .0. 00000 00.0 0.30> 0.... >000E00 00.0..300Eo0 0 0. 05000.0
0. 00.0.0000 000.090 0.00.. 0000.00..E .00 0.000 0.003 00 00.0003 00.30000 00: .0. 00000.0 0000 0 0. 30> 0. 00.0.0000 .00>> «0
05:00.0
00000 00003020 .000.>o.0 002.0000. 00000.0 05000.0 00003090 0. 00.000.
$2 020.00 0.00. 030. ”>000E00 .000.m 5.3 00.000080 .050 .3000 0000000 003 .00>> Eu

 

 

 

 

 

 

 

 

00.00 E00
02 00> 00.0..300E00 .050 0. 00.003 30> 00... 9.
00...
0002 0002 0. 00300 0 0000.00.00 .0 000.0000. 05 0.0 .00>> mm
0% 0 0.00 ..E.0 000 00000 00.
0.0.00 0>00 c: 02.00000. .00.... .02 >00000..000 0>00 on 00. 0. 0003.00. .0. 00.0 0 0. 0020 0. >00000..000 0E.. .00>> mm
.0030 00 0 00 YN 00>. 0.0 0.0.0.. 0 0.00 .0.:00.00..>00
000 0:03 0.00.0000 0.... 0003000 .00....0 .005 0. ...30 >00000..000 .0 0>00 on .0. 00.0 0 0. 0000 0. >00000..000 0E.. .00>> mm

 

006090 0.0 0.0000. 00 00 .00500 ..
E9030 .0. 0.0 0.00.090 .0..0030 >..00_2 0>0.0E. .00 0000 000 000.. 000000 0.0000 30> 00 300 00300 00.0.0 0>0.00 .0..0030 ..

 

 

  
 

 

....03 .0 :03 000 .. 0.0000 30
.. 000000 .05000 0>.0 .0030 5.3 00000.2 00030 5.3 0000005. 00 300 00300 00. 0. 0000E.0..00 0..o.00..000030 ..
N. N30 3 EC. 002.0030 0

 

 

 

:ofi

111

 

 

 

 

00.03.0030 0.0. 0. 0.E0..

 

 

   
 

 

    

 

 

 

 

 

0030.0 0.0...3E 0..3 00E..0E00. 0030 _. >__0303 _. .00. 0.00. 0.050 000 .0. 0.0.00..000030 >000. 30... ON
..03
..03 .0 >...03c 02.10% .>.0.0.0 0.300000 05.005 .>.00..03 ..000 00.00.00 5800000030 0 0. 0.000 .003 00 m.
..03
0.03 .0 >502. .0000 .>.0.0.0 0.300000 05.000. _>.00..03 ..000 00.00.00 .0..0030 0 0. 0.000 .003 00 0.
02000000030 0.
00800000030 00.0.3910. 00.000000030 00.03.05 ...03 .003 000 >0000.00 00. >0 0000 0. ..03 .002. t
05.300000 000..0..000 00000 000..0..0000000 0000300000 05000.0 .0. 0.0.000000. 0. .000...0000 00.03.95? 0.
00.00 0. 50.00.03.0030
00> 00> 0 .0. 0000.00.00.00: 0.300000 0 0>00 30> on. 0.
0000300000 0. .000 0.003 «N 00.0. N 00300 0.050 0 .0. 05000.0 00.0 30> o0 000>> 0..
.0>o.000 00.0 2.0 .0..< 00.0. 0.... 00.03.0030 000 .0. 05000.0 00.0 30> 00 000>> m.

 

 

 

 

 

 

 

 

        

 

 

  

 

  

 

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

000.00 0: 0.00E.000.00 N Tm 00.03.0030 0.0. .0. .000 N.

0.000. 0.0.0 0.

.... ..0 00300.0 0. 00... 0.303 30> 00.03.0030 000 00.02

.00 03030.
ON >0 .0>0 00> 00> 00. .0. .009000E00 .....3. >000E00 00. 0.0 or
00000 0.8100000. 0000300000

.000 00000005. .0. 0.30.0 .000.>o.0E0 00.000039. .0 .02 000E00 00. .0.00.0 30> 00 30... 0
0000.020 00.00 0.0. 0000.0>0o >.o.0.0 00.00 .00 0. >000. 0.00.0.0 .00 .000 00. 0. 0.000 .003 00 w
000.00 02. .00 0.0. 00 .000 00. .0. .000 x.
0.00 0 0. 00 0.00 00. .0. .000 0
000.03.0030 N. 000.03.0030 : 00.00.0003 0.03 0.00030 08.03.0030 .0 .00E3z .
.V
:2 .32 .NN 00< ..2 ,302 .NN om< . u
0.> o. 0.> 0 00 E00 00. 0..3 0.00 .0 .00E32 N
.20 .2000 n.> .0.00M_ 000.000 00.00 0.0030.>.0.0_ F
NN N>x .N N>x 00000030 a

 

 

 

 

NN Om<—

112

 

co_m_>_un:m

 

   

 

 

 

 

 

 

 

 

55mm; 55$; .uwccma mm Eon «o: 9.50: 25 E 32% be $38 BEE 9: 2a 6&5
533623 25 E £me
9.83 N 9.me N 5:52 a co umfiaucoo $580.: $295 2< mm
mm> mm> 9:553 232 .8 3809 39: mm: 30> 00 um
«and» a mm.=_>=om «v-2- mu *0 $809 EEEmE so on. 8
.mbcoo E55 ucm Emu : «mafia an 3032 9.: E95. Emu
8E .mucoxmoz, :0 £95 g 959m fo>> mocmxmmz, :0 a: mme diva: EmanoE an be $8 c_ 39.20 53 9: m_ 261 on
855.05
mm; mo> @5563 9.5.82 ..2 “9.58 m m>mz 30> oo mm
mam. 6 mm BE Eoo m_ mu
{com cofiszwcoo $8 5.62.3 $3 an x83 vocmfimm Fo 32:82 fizz, ..m..0cmm 5 mm

 

Emma >=mu co 92:26?

 

 

  

 

 

 

 

 

 

 

 

 

 

 

magnum .8385 8:338 .050 .28 3E3 528 mc_cm_mmm .223 36228 9m EBmEEmQ «93> nu

uflm. E8
25 5E 2. 2 8258 >3 2 Eu 2. 820% 95> om
Euxm ms. Emstom E295 EEms. :uoi .880 ms. vow: 2m 853:8 «95> mm

9.23:.
x3 ucm 9:25;. =me BwaEmE m5 2 co uwmmma 53 9: m_ Bo: vm

umSumcow mmEoc 29:2:
ucm mmEoc c328 umxé mum: comm—I 9:328 _mo_tm> new :35on co 92.63 ...ch 885908 $8.: 05 26: mm
Unzw E0: 935 SE .n.> Emu cozoazmcoo mmE>=um wmwé mEccmE BE; .8202: En 95> NN

mama >._m_u m :0 mmfinasdfiomcum memama ogm Ema mm 8:69: 089.. 29.7.. m .2
62% tmfi m5 5 Sn .598 cc 229:3 ucm uwm: m_ SmaES ”3.8 cc nommm mszzum m5 mcEcmE ..2 2380.5 9: m_ 5&5 3..
mm N>x rm N>x mcozmoso &

 

 

 

mm 0%

113

 

    

.00E00_0>00
0000 8 0800.000 0:... 000 >000E00 00.
0_ 000.00 000 50200008 .0.800Eoo 00>

0000.. 0..3 05250
0.0000000. ..0E80. 50200008 0.05.

000.8 €000.00 50> 0.0 0080 .00>> «00000.0
050000 00000090 00. 0>0.00: 8 000 :05
08000000000500. 0.0:: 00> 00 0.0001 02. .<

   

 

  

 

00. E0.

 

 

00.00 0000 .8 000_ 0500 0

 

mv

 

 

 

 

 

 

 

 

 

 

 

   

 

 

  

 

 

 

 

 

000 0000 .8 00000 00. _00_.. .000
0.003 r .0800 00.0.50 .00 .8 £000. _

__02$.0 .000 000 00000005 .000 ..0”.

 

5000 9.800 .00 .080 0300000

    

005E00 0 E08 0000_0>00 0000 05000.0 00_0 050000 .0 00E00 =0 .8 00_0 _0.0000 0 0>00 8 .0000 8. 0_ 3.
.00 0
:8 .800005 .8000 .050 800008000000 0E0_00.0 0:003 .0200
..08003 0.8008000 000.. .0 0000.58.00 000 0000.00 00.00000 .8030 00.0000 00000.0 050000 00. 0000000. 0.800. .002, 00
0.0 0. 00. .0 0080
.0000 0:2, 800.00EE00 0.0080000 80. 0.0. :0 00000_00EE00 ”000.080 0
:0 00200. 0.0000 8 80:0 0:00 000 0000.. 00.3000 00000. «00.05. 00:000000 0.00> 0x: >000E00 00.230008: 0 0_ 00.0000
500 0000 00. 80000080. 0500000 05 0>_._ 00208000 0300000 30: 00>m_ .8 00:00.0 0000 0 0_ 00> 8 050.0000 800>> Nv
00088.85 00.003 000. .0 0000000
00000; g 000 00_0 00000080 00x:
0 000 50 .0000 003 0000> 000 0. 000 00000.0 050000 00000080 8 00.000.
05000 >03 000 005000.000 E00. 80 N (2 £05 00_000E00 .050 .0000 0000000 003 000>> S.
0200 E00
0 00> 02 052500800 .050 0. 000003 00> 0>0I 00
0E:
0002 0. 00000 0 000200.00 .0 002.0000. 00. 0.0 «00.5
$000 0 000 8.8.0 000 00000 00_

    

.8 00_0 0 0. 0020 0. 500000000 05: .00>>
0 000 .800E00._>00

 

 

0002 0. 00.0 0300 .80_0E8 £00. 0_ 0002 .8 00_0 0 0. 00>_0 0_ 300000000 0E: .00>> ...m
.00000 g
000 0 000000 .0000 .3 0000 0000000 E0>0 0000. 0000E.8.00 0 .2003 __0Ew_0_0000 00> 00 .50... 00000 00. 0.0 0.600 .2003 2 mm
._ 0_0000 00
000 0. 000000 .0000 .3 0000 0000000 ..0n.0> 00:00. .000 00 0000000 00 .500 00000 00. 0_ 00008.8.00 018008000000: mm
00000000 %

 

 

 

NN N>x

m E?

114

 

 

 

NN Om<

 

 

 

 

00.03.0000 0.5 0. 0.E0..

 

 

 

 

 

 

  

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

  

 

 

 

 

  

 

  
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.050 .8 F .F 005 0.0E 00500.“... 0000 F .00. 0.00. 0.00.0 000 .8 0.8005000000 5000. .50... 0N
55.00000 000.0..0>> 50.000 003.00 .0050 0050.0. .000 55.00000 00500080 .0050 0800.00 .8005000000 0 0. 0.000 500.5 00 0.
55.00000 000.0..0>> 50.000 003.00 .0000 0050.0. .000 55.00000 00.5.0000 .0000 0800.00 .0..0000 0 0. 0.000 .005 00 9

.05000 .05000 08005000000 0.
50.000 .0098 08005000000 00.55.05 50.000 .0090 08005000000 00.55.0>m. 000.5 500.5 000 5000500 05 50 0000 0. 0.0.5 .00>> 5.
0050000 00500080 8.0 00 500008000000 00.5.0.0 .8 0.0.000000. 0. 5000000000 00.05.0025, 0.

00.00 0. 50.00.03.0000
00> 00> 0 .8 3058000800.. 0.000000 0 0>00 005 00 3
0050050000 0.800 00.005 0 8 n 0050050000 0.0.00 5000. 5 00000 0.00.0 0 .8 00000.0 000 005 00 000>> 3

05000 080

...00 00 8 0E00 50.5 0. .050 5000. 5 5080.0 05 0. 0050050000 0.0.00 05000. 0 00.03.0000 000 .8 00000.0 000 005 00 000>> m.
8.00 0N 00E00 0N 00.03.0000 0.5.0. .000 N5

0.000. 0.0.0 0.
0m. 0m._ 00000.0 8 00... 0.00.5 005 00.03.0000 000 00.02 E

0>000 00.5 .0 000 .00 00030.
00500090 ..000 05.50.00 00.5 00.00 050000 00.5 00.0 0. 5.50.0 00.. .Oz 05 .8 .000.000E00 ..5.0. 5000000 05 0.0 or
0.000. 0.005
00030.0 ”05.05 .0E8000 .0000 00.00 05.03 8.0 50200 055000.00 5.80.... 0000.00 05 6.00.0 005 00 .50... 0
0.00..0>0 000000 ..8w 00.00 500520.00 . 05.0.. ”005.80 .00 .000 05 0. 0.000 500.5 00 w
000.00 00. 0 ...E 0... 08000 RH .00 0.5 00 £00 05.0. .000 5
...E 0N5 0 0 ...E o: 0.00 m 0. 00 0.00 05 .8 .000 0
.5000E00 .8 ..0.0>O. m 520 .0000. 0 002030000 0 00.00.0000 0.05 5500500 000030000 .0 .00502 W
00 £8 05 .0000 .000 V
..2 0...... 0.05E00_m .mm 09.. ..>. 0...... 0.050.020 Hm 0m.< 005000.. 0E0: 00 0.00 H
0.005 m 0.5 ...N 00 E00 05 5.5 0.00 .0 00502 “
005050.0_E0<

.000000 5080.0 000 00000.“. .0 .000005. .0.0000 005.000 .0500 000503.80. 5
N0 N>x 00050000 a

 

 

 

 

 

 

..m N>x
mm 00(—

 

115

 

    

  
  

   

 

 

 

 

     

 

   

 

     

     
   

  

  

 

    
 

 

 

 

 

 

 

 

      
 

 

 

 

   

 

 

 

 

 

116

 

 

 

 

 

00.03.0000
.0500>> 00.00 0500.5 0.50. 050.00 .0 000000 .800. 05 0.0 .00>>
05080 00.03.0000 0.50. 0.000
0005 0 0000 N. .0000.E0.0005 .5. 550.005. .0000..20 .0300. 0 00 08000000 0005000. 000.020 0.< mm
0.000000.
.000 00 00000 ...00 0.0 000.03.0000
.500 .8 080.058 0.000000 .02 $2 00.000.0 0.0.0. .8 00.000. 0005 000 005 on. N0
0053500 00-8- 00 .0 00.000. 0.0.0.0E 00 on. .0
0.00 00.0. 0.00.5 8.0 00 00030.. 05 5000.0 ..000
0005 .0000000 05 005500 00008050000 00 .0000 0000000 .0050. 080000 50.00>> 80.0E000. 00 .0 0000 0. 0000000 00.0 05 0. .50...
00005000.
02 02 05000.0 00.0.80. .8 .0050. 0 0>00 005 00 mm
000.. .0. 00 08070.00 0. 00
05.00-00 $00 .00 0.05 0000.000 .0 008000.00 .005 ..0.0000 0. mm
0005.008. 0.000 5:00 00 0053500
500 .00500 .00000 .0500>> 0500000080 .0 050.00 0000.000 205 00.00.0000 0.0 0.80E0.00 .00>> ...N
080. E00
0.000000 00 00000 500008000000 500008050000 00 8 0053500 500 8 500 05 000.000 00>> 0N
0.05500 00.0.5 00.3.0 00000< 0.2 . 00..0.80m 0.05500-0_.0m. 0000 0.0 00.05500 .00>> mm
00>.0>0.
0.000000 05 .00 0.000000 x00 0.00E00. 05 8 00 000000 00.0 05 0. 50... ..N
o 000 0 .8
05.00 0005.00 00500.0.000 .0 0.00 000.0. 0.000000 000.00 0.05.0E
00.05 0.000000 0000020 0.0500 00 P .050. 00x0. 0. .8005000 0005 000m. 00 00.0.05 0005 08050.08 00005 0.0 50... mm
00>0. 0 .20 .00. 8
.20 50 0000 0. 00008050000 0.0.. 0.000. 0000.05 .20. 00008050000 0053500 0005 00000.0 0.05 00>.0>0. 0.0 00.5 mm
80000 50.005 00.000000 .00 x0. 000.00 080000 50.00>> 00.03.0000,
.0 005. 0000 .8 80.0E8 F 0.00.0.0 00 00000 0500050050.... ..000. 0. 800 0.000 0.00.0 0 .0.
.000 .0..0..0 00 00000 ...00 80.008 500 < 00.0 0000.0 0. 0.05500 0.0.. 80.00.00. 0053500 05 05000.0 .8 0.000020 05 0. .00>> .N
Na N>x Fm N>x 008.0000 0
mm 0E

 

 

 

 

 

0050050000
000.0000 ..8000 00E0... 00.8059
00000.00 0.800.000 005000 5.5000 00>

0.00000 .00.05.0 0050.080. 00 00.0.05
5000500 00.05500 .o 00: 00030000.
000.0... 0230. 0.0000 00.00000 00>

 

 

0000 Ow wwmo

  
 

000.0. 5000500 .005 0.0 000.0 .00>> «00000.0
050000 00500090 00. 0>0.0E. 0.0.00 ...5
08000050050580. 0.0.5 005 00 80.000. 0.5 .<
00.00 0000 .8000. 05.00 0

0».

 

 

 

 

 

 

00.00 E00

000.0 0.0000. 000000 000 0.0.0E8 0 0.0.. 000 8.000. .0 000.00 ..0 .0. 00.0 .0.0000 0 0>00 8 .0500 .. 0. vv
0.050.0 500. __0 00000.0,
00800.0 0.00 080.0000 00 85000.2 005000. 00.0-0.0 .00500.00E0.00 00E0I 00000.0 05000.0 050000050. 0.0.00. .00>> 0..
00>.0>0. 0E.. 000 002.8800

... a . a ... . : . L E8050 0.005 0.... 50000.00 00.0..000500 0 0. 00.0000
00. 0080.000: .000 00.00.00 000.00 .00.. 0500...... .0. 00500.0 0000 0 0. 005 8 00.0.0000 .00>> 00

5005080 000. 00000.0 00.000.0 00500090 0.00.000.
00.0.0. 0 Oz 50080.0000 .0 .000 .0500 05 .800. 00.000000 00000 00.. h 5.5 00.000E00 .050 .0000 000.0000 005 .00>> S.

   

 

 

   

 

00.03.0000 0. 00.03.0000 0.0.. 0.05.0

000.03.0000 00 0000000

0 0.00 ..E.0 000 00500 00.
.0. 00.0 0 0. 0020 0. 5000005000 00... .00>>

  

 

  
 

 

0 0.00 .0.00E00.300

  

00> 00> 00.0..000E00 .050 0. 000505 005 0>0I o...
0E..
0002 0002 0. 00000 0 0050.00.00 .0 0050000. 00.0.0 .00>>

 

08.0 000 00.0000 . .2
00000.00 5o..0> 8.0 00 ”00.0. 0.0000
8 0050.0 05 00.0.0.0 00.000 .0.00

 

 

500.000 .200. 00500 558.000

 

 

 

No. N>x

 

0002 .0. 00.0 0 0. 005.0 0. 5000005000 05.. .00>> 50
0000000 ..
m 0000.. 00.0.00. .0. .0000000.00 .o 003 0.0000 005 00 500 00000 00. 0.0 050.00 .0..0000 .. mm
a .. 0.0000 00
m 0000.. 000.00. .0. .0000000.00 .o 00.... 00 500 00000 05 0. 000058.00 058005000000 .. mm
5 N>x 00050000 a

 

 

 

00 0E

 

117

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

00.03.0000 0.0. 0. 0.0.0..
0005.00 5 5500.2 0005.00 000 .00. 0.00. 0.00.0 000 .0. 0.2005000000 5000. .50... om
003.00 000 0000 50.000 00.050 000080.00 00.00.00 .2005000000 0 0. 0.000 .00>> 00 0.
003.00 000 0050 50.000 00.050 000080.00 00.00.00 .0..0000 0 0. 0.000 .00>> 00 or
08005000000 0.
0000. .. _.< 0000. H ..< .003 .00>> 000 50000.00 00. 50 0000 0. 0.0.0 .00>> t
00.000.0 000 0050050000 0050050000 00.000.0 .0. 0.0.000000. 0. 00050000 00.020003
00.00 0. 50.00.03.0000
00 > 00> 0 .0. 0058000800.. 0.000000 0 0>00 005 on. 0.
08.0000 0.0 8.0000 0000 0050050000 0. .000 00.00E «Va 00000 0.00.0 0 .0. 00.000.0 00.0 005 00 000>> 3
.00E0o.0>00 8.0 0050050000 0. .000 00.000. m-~ 00.03.0000 000 .0. 00.000.0 00.0 005 00 000>> m.
0..00 3; 0..00 3.. 00.03.0000 0.0. .0. .000 m.
00.000.0 00.0.000.
m5 m0 00000.0 0. 00... 0.002. 005 00.03.0000 000 00.02 ...
.005 00030.0.
00> 00> 00. .0. 000000000 :50. 50000.00 00. 0.0 o.
0.00 .9505. 0.50.00 .0 000.02, 00>0.000 0.00E00.0>00 .0 000.02 000E00 00. .0.00.0 005 00 .50... 0
0.00. 050.0... 00.00 00.000.0 0.0.00 .96 .000 .00 .000 00. 0. 0.000 .00>> 00 w
0..00 oov 0..00 co». 5
.00 .0..00 005 0..00 80.. o

 

08.000000 9

000.03.00.00 0.

00.00.0000 .003 5500000 08.03.0000 .0 .00E02

 

00 E00 00. .0000 .25

 

 

..2 60.0001. .3. Om<

:2 00.0000 .3. om<

005000.. .8000 00 E00

 

0.00510

0.005 m

 

 

 

500 0.00 00. 0..3 0.00 .0 .00E0z

\VllV ll)

 

 

.00000.2 60.0.0

00.000800 .0 0050050000 .0 0>

005.000 00.00 000303.85,

 

 

Nv N>x

 

 

00050000

 

 

 

 

0V N>x
3. 00"...—

118

 

0 0.00 0.0.005000000

 

 

0000.0 .0 0. E.

00.03.0000

 

 

 

 

 

     

 

 

 

E00. .0. 0.00..0>0
.00 0. 0503 .. 005500 0.0 0000.. .00000000,

0.05 0. 0000 8000500500.

  

00 00030. 00. .000.0 .000
0.0.00.000. 00 .0 0000 0. 0000000 00.0 00. 0. .00...

 

000000000

    

 

 
 

 

.80. 0>5.0 0.050.0.2 ..00.00>> 02.00.000.000 .0 0.00.. ..00.00>> 0.0. 0. 050.00 .0 000000 8.00. 00. 0.0 .00>> 00
00.03.0000 0.0. 0. 0.000

50.003 50.00>> .0.000. 0 00 00.000000 0005000. 00800.0 8< mm

02 oz 00.000.0 0.0.0. .0. 00.000. 0000. 000 005 00 mm

000000 5..00 00> 00l> 0053500 00-0.- 00 .0 00.000. 0.0.0.00. 00 on. ..m

 

  

 

00.0. 0.00

02 0.00.5 .00.0 5.0>0 00.0003 00.000.0 00.0.000. .0. .0000. 0 0>00 005 on. mm
000. .0 00 00.0. E00 0. 00
$00 {ea .00 0503 0000.000 .0 000.000.00 .00>> ..0.0000 0. mm
00500000. 000 000.002, 00580.2 0053500,, 0.000 5..00 00 0053500.
5....00..0>< 0.000000 .0.05800. 500000080 .0.000000 00 00000 005350< 00.00.000 0:0... 00.00.0000 0.0 0.80E0.00 .00>> 5N

   

 

 

 

 

 

 

 

 

 

 

 

.0000805000m._ 0000805006.! 00 0. 0053500 500 0. 500 00. 000.000 00>> mm
.0880 0.2 .0880 0.2 0000 0.0 08025.00 .00>> mm
0.0.005000000 , 00>.0>0.
000.00 00.00.50.0 0.0 00.000090; 3.0.00.5-.0. 0005080 008080 0. 800.505 0.00E00. 00.0.00 000000 00.0 00. 0. 30... vm
00000 .0 x0. 0.> 00.000000 0.0 500. 0005000. 000.00 0.05.0.0
0003 005500 0.0 0000.. .00000000 00 0. 000 00900305 .00005 0..3 .00 .. 00050.5 00 00050.5 000>> 0800.0.000 0000580 .50... mm
0000.. 000 00008050000 ..20 0000.. 000 00008050000 ”.29 0053500 0000. 00.000.0 0..03 00>.0>0. 0.0 00>> NN
000000000 000 00:00.00 0... 000. 000 .0.00. .00 00 0000500. 00000000 00000.0 0E00 0.00.0 0 .0.
0.000000 02.00000. ..00. ..E000 0000. .. 000 0..0E00 0000.. 00.. 050.5 .0 0.000000 0053500 00. 00.000.0 .0. 0.000080 00. 0. .00>> ..N
«v N>x ..v N>x 000..0000 ..

 

 

 

 

119

 

00.000.0 .0. >000 .. 00.00.
0. 0000 00 0.0000 080.5500 80.2 00>

.0.05 00. 0. 00>.0>0. 0.8.0.0 =0 00
0.. 0. 0000.008. 80.2 .0.0000. 00. .0. .5
..000 0 005 0. 080.0x0 00 0.0000 08025.00
000000000 0. 00.0008 80.2 00>

 

 

 

000.0. >000E00 .00.. 0.0 000.0 .0ES 0000080
00.000.0 00500080 00. 0>0.0E. 0.0.00 :55
0800080050500. ..00. 00> 00 80.000. 0.0. .<

 

 

 

 

00. 05.00 0 0 0.8.00 0. 0000 0800 0000 .0. 000. 05.00 0
00.0 056002 00 0.0000 00.0 .0.0000 < 00500.0.000 < .0 000.00 :0 .0. 00.0 .0.0000 0 0>00 0. .0500 .. 0. 3.
33500080 000.. ..00.00>> 0050N.000.0 000 000500.00EE00 00000.0 00.000.0 00. 0000050. 0.0.00. .00>> 0..

0060000000. ,

005 0. ..000 0050050”. 0050050000 0. ..00 0050000 0.00.. 0.... €000.00 00.0..000E00 0 0. 00.000.0

0 0.0 00>> 0.00E0E ..0 .0 E85235, 000 0052.000. ..00E0>_0>0_ 000. ... .0. 00500.0 0000 0 0. 00> 0. 00.0.0000 .00>> mv
00.000.0 0.

00>.0>0. 0.000.0E. .000.0>_o>0. 000.. oz 00.000.0 00500080 000080 00.000.0 00500080 0.00.000.

00o..0> 0003.00 00500.0.000 0000 o. 0050~.000.0 000 .00000. 0000 0...... 00.000800 .00.0 .0000 000.0000 00.5 .00>> 3

 

00>

00>

00.00 E00
00.0..000Eo0 .00.0 0. 0850.5 00> 0>0I

 

  

 

00.....0.000008 800. .00000

 

 

00000
E0030. E8. 00000 00

    

.00030. 0.

        

       

0:...
0. 00000 0 0050.088 .0 002.0000. 00.0.0 .00>>
0 0.00 ..000 000 00500 00.

 

 

         

 

 

 

  

 

 

0. ...00 0. 00. < .. .0.00 .00....0 00. 0. 0020 0. .005 < .0. 00.0 0 0. 0020 0. 300005000 00... .00>>
0 0.00 .0.00800..>00
0002 0002 .0. 00.0 0 0. 0020 0. 500005000 055 .00>> ..m
0E00 00. 0005000 >00. ..
.. 0000000 000. 000 005500 0. .0..0000 0E00 0.0000 00> 00 300 00000 00.0.0 0>0.00 .0..0000 ._ 00
0000. 8 000 00388 0.0.0.0 .. 0.0000 00
00000. 0. 0..0500 0.00502 0.0.080 ._ 00000. 0. 000500 0. 00502 00 300 00000 00. 0. 0000000500 0..0.00..000000 .. mm
0v N>x EV N>x 000..0000 ...

 

 

 

 

 

vv om<

 

 

 

120

 

 

 

 

 

 

 

  

 

.0 0005. 00.035 E8. 00300
000 0.05 0.0.. 0.000. 0..2. 00.000800

        

 

 

000 .00 00.03.0000 0.0. 0. 0.E0..
00_0 .0.0.00..000000 00.0.0.0 N 0.0.0000 000 >535. .00. ..00. 0.00.0 000 .0. 88005000000 >000. 2.0... on
...o>> .0 >.._00O 000 0050 .0000> 0.00000000 002.005.0008... 00.00.00 .0.00..000000 0 0. 0.000 .002. 00 0.
000E00 0 >.0000 00.0000 .000 0.0000 ..00.... 50000> 0.00000000 00.00.00 .0..0000 0 0. 0.000 .002. 00 0.
08000000000 0.
08000000000 0. 00.0.>.0>m 08005000000 0. 00.0.>.0>w 050.5 .002. 000 >000E00 00. >0 0000 0. .502. .00>> ...

 

00.000800 5000005. 00.00800 00.000.0 .0. 03.000000. 0. .00850000 00.05002.
8000 0. 50.00.03.0000

 

 

 

 

 

 

 

 

 

00> oz 0 .0. 2050.000808. 0.000000 0 0>00 00> 00 m.
0000 0. .502. 0.05 000>> .0005. .0. ..0000 .00300 mom— 00000 0.00.0 0 .0. 00.000.0 00.0 00> 00 000>> 3
00.000.0 000.. 00500 00...0.000< 000.. 00500 00.03.0000 000 .0. 00.000.0 50.0 00> o0 000>> m.

      

 

alga o. 0...... 3 8.05.330 0.... .o. .80
00.000. 0.0.0 0.
00 00 00000.0 0. 0.... 0.002. 00> 00.03.0000 000 0E0z
.00 00030.
000000.000 000000xm. 8.03 3. >0 000000xim; 00. .0. .000.000E00 __5_0. >000E00 00. 0.0 0.
526.0—
.0.505. ..00> 000380 00.00 3.0.0.1 0050.000< .00080 00:00.00 .8505. 000800 00. .0.00.0 00> 00 30... 0

 

.00080... .00> .000 00.00

 

00.00 .800 00500080 .00 .000 00. 0. 0.000.002. 00

 

 

0..00 cow

 

0..00 cow

 

5050500 000 Too:

.00N.0..00 oomw

 

 

000.03.0000 mm

000.03.0000 00 00.00.0000 .502. >..00000 000.03.0000 .0 000.02

 

 

 

:2 $0.... .00 00.1

 

 

.5. .xo... .0... om< 005000.. .0E00 00 0.00

 

 

 

lV\qul

 

 

 

 

 

 

 

 

0.00> N 0.00> N >00 0.00 00. 0...... 0.00 .o .00E0z
00008050000. .000005. 00.000800 005.000 00.00 0002.03.80. , F
mN>x E N>x 00050000 ..

 

 

 

008

121

 

00.0.>.0000

 

 

 

 

 

 

 

 

 

 

 

0.0. 2.0. 0. 0.00.- .0.0>> .00.00>>, .00.00>> 0.0. 0. 0>0.00 .0 000000 5.00. 00. 0.0 .00>> X.
00.03.0000 0.0. 0. 0.000
.00.00>> .00.00>> .0.000. 0 00 00.000000 000..000. 000.080 0.< mm
00> 00> 00.000.0 0.0.0. .0. 00.000. 0000. 000 00> 0n. mm
0..0 0. ..00 00> Wyn 00...>..00 00-0.- 00 .0 00.000. 0.0.0.00. 00 on. .0
000000.. 00000.00; 00 0502. 00 00030. 00. .000.0 .000
00.0 000 0. ...00 000... 00. 000 0.00000M. 0>0.00 .050 o. 0. ...00 000... 00.. 0.0.0.0000. 00 .0 0000 0. 0000000 00_0 00.0.2.0... om
0000E.0..0
.0000000.00 .0800 00 0. 00.000. .00000.0..0000 00.000.0 00.0.000. .0. .0000. 0 0>00 00.. on. mm
000. .0 00 00.0. 0.00 0. 00
$00.. 0. 000.0 *8. 00 0.0000 .00 ..03 0000.000 .0 000.000.00 .002. ..0.0000 0. mm
000.080 0.000 ....00 00 00...>..00
000. .. 000 .00.00>> .2...00..0>< .0520... 00 00000 000.000 .00000.050000 0.0.0 00.00.000 0..02. 00.00.0000 0.0 0.0.0.0800 .00>> 0N

 

00.0.9000.

 

 

 

 

 

 

 

 

 

 

.00000.0..000fl .00000.0..000M_ 00 o. 00...>..00 >00 0. >00 00. 000.000 00>> mm
08.0 0.>. .00xw 0.>. 0000 0.0 00.020.00 .00>> mm
00>.0>0.
0.0”. 000 ..0EmL x00 000 ..0Eim..F 0.00.000. 00. 0. 00 000000 00.0 00. 0. 2.0... vw
000.00 0.00.00.
0:00 00000 00-2.2.0. 000 x00 .00 00 00.0502. 0002. 00.00.8000 0000.. 0.0 2.0... mm
.00000.0..0000 .00000.0..0000 0..0 000000.). 00.000809 00...>..00 0000. 00.000.0 0..02. 00>.0>0. 0.0 00>> NN
02.0..0. .00. 00. .00.000000 0. 00.0.0.0 .0>0 0800 0.00.0 0 .0.
000 000000000 08.0000 0.0 0..0..00 .0..< 000.0. 0.0.... 08.0000 0.0 0..0..00 0000 00...>..00 00. 00.000.0 .0. 0.000080 00. 0. .00>> .0
mm N>x E N>x 000..0000 ...

 

 

 

 

mm om<

 

122

 

«mans: 25:5 can 2:: co 8805

839508 5:03:95

039.2 ”2mm; 8:8. 2 3282

.._.=. w :8... mc:EoEooc_ 8526: 95.8..
.83: 5.3 9:8th BEEBESE >9

EmEEoo 2 moan: 222.9 5:5 mEEwoh 5:885 8995 2 963 mEmmomm< .93

95.9 >cmano 50> 9m 8.2m EES «889a
mcEcmE 5:03:05 9: w>an_ 2 9m: ___>>
cohmmmotco:m>occ_ x55 30> on .9383 25 :<

 

  

wEo: comm 8: man. 0500 m

mv

 

 

 

 

 

 

0:6QO gm 52; 5550 .o meo: __m .8 22¢ .983 m o>mc 2 5:8 = m. we
mum: mEEE... $583 \mgctm‘n: _8_cmcoms_ BEES“. 8805 9:55;. 9: 853$ 920$ 6&5 mv
28> ox: >cmano mEESnmEoc m E @5553
25:8 2 co:cm:< :coEwmmcmfi Emmh mm:_>:om m~:o._o_._n_ ”maxw co:mN_cm90 ..8 8:85 :85. m m_ 30> 2 96808 HES Nv
8885 $28 2 88905
8 ancm 3688 86 60:03:05
go :88 “.8288 8 2:93 mcmE $805 @5553 5:03:05 2 8.89
.ww_mm :0 :88 can. 828: ct: mg: (\2 5:5 memano 550 Sean umnBoom was “95> 3
925 E8
mo> 02 96:39:05 .650 E veto; 30> m>mI ov

 

 

 

 

Emucmfitmazm 8: wgcomlr

mzcom

 

mcoz

  

ocw c:coo
oc om .mzctma :o :88 mtmfi 0:6me

2:

 

      

E 830: m mc:m_ano :0 82252: 9: 9m fic>>
m m6: EEm ucm 8:8 9:
‘6: 53 m c. :96 w_ 3:85:50 9:: 6&5

 

c_ :5: m_ 8.83 :0 29.8 «mag—m :0 mm:

9.83 N

m 98 _mEoEc9_>cm
.8 cma m E :86 m_ 5:82:50 9:: 55>

     

mm

hm

 

8:02 85,50th .52

8:02 85,50th .52

=
90cm: 30> on 26: 8:8 9: 05 36.8 .233 :

 

 

umamfl m_ 8:0:
mocmEStmaéoc m :9: .cm>_m m_ mEEm>>

8:0: 8cmE5t8éoc m 88. o>>

z 222 8
w: 26: 8:8 9: w_ 8cm§ot8 {Bogcounzm :

 

mm N>x

 

Fm N>x

238:0

   

 

 

 

mm 0m<

 

 

123

Appendix D

Data Collection Summary Table

124

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

35050 2030:
:0: 35:000. 880%:— ::m 822:8
0:29:85
€25.00: €20Sw :0: a:
$8: :0 wctquwcov 139000 @50an 0:: E0: :66 8:00:25
0:29:02: :80: 82:. 00:29:00 :< 023% :0: wE—EBDQ 0:29:02: 00:809.— :0EEE00 :<
BEE—a
95:08 RE :33: 00>: 038%:— cocuommE mo_:_>:0<
:30: 33:. 8:00:80 BE 8E: m :SSEQ mat—02m =~>fDD wc§00:m :EBDQ
52.2 52.»: 5.2-2 82-:— 32-2

 

_ “52:5

125

 

 

 

 

3:530 ,

 

 

 

 

 

 

0 , 20000:
:0: _:>05 0 580%:— 05 8:23:00
0:03:50:—
5:00:35
3 0055:: :0:
b38200:— mm0q 0:05:50 =< 0:05:50 :0: 5:23:— _0>0500 :0:
:23 0:05:50 =< 3:03:50 :30: 33.:
m :c: __m>»bo :25sz 55003:. c0520
v :E: :03>...Q wEwa N :E: :Ssbo wEwcai _ :E: 55:00: 550%:— EEE mmzc. m0:_>_:0<

 

m :5: :SFCQ wEwSNI,

 

_ :E: :Esbo wEwcmz

 

EDEN—Swat

 

:30: mmEP

 

$3: :0 w:t00:_w:m

 

 

 

52-3

52-3

52.:

557mm

SSTNN

 

 

_ 052:5

126

 

 

 

 

 

 

50.5509

 

0.508.. :,:0_0 ..0:.E:m,,

m5...»
E Ease: .0:&=m

5: 0030...: :0: w:_:00:

0:03:50 =<

22:88 .2

$5000.
0:0 m0::_>:00
0:29:85

 

 

 

"3:00.: 5: 0w._:..<

_ 05%:-

0 :E: :NBEQ wEwE:

 

N: :E: =03ba mEMEI

o. :E: :3350 3:98:

w 0:: =03bo wEwENI

m :E: =0>2ba wEmSN:

 

@500: :0: 0wcmt<

_ _ :5 =0BED mecaz

o :E: :0::an m:_w:0:

n :E: =m>>ba mEme:

w :E: :03bo wEwENI

00:55
8:01:06.

 

 

.:<._

.02- _ m

 

 

 

E<-N

.02-om

 

 

52-0w

 

. 050......

127

 

 

 

 

 

 

 

 

£580.
5:: 8225:

 

 

 

 

 

 

 

32:: o .2... a... :_ 32:: o .2... 2... :_ 5:553 Em: 22:88:.
a... .& ..525 .2535
a: 58.85 a: 58.85 229.50 :< 22an0 :<
25255. 5: 9.2.8: 23285 5: 9:82
.0328 @5225... 8:52:
5.. wciooa 3:222 3:323».
2E: 0 8E: 0 8:5 0 why—ME. wcumoc wEEuc 5,. owcnt<
8.32352. 2: wEmfl. m2mcsm.:_o_. :m mEmmF 5:: $50.21. wcfimh @5595. 59. ow:m..<
5<-© 5<-w 54..-: 5<6 5<-m

 

 

 

 

 

 

. «52...:

128

 

 

 

 

 

 

 

9.8:...
5:: 85350:

 

 

 

22an5:—
mcsfifi 5:55 ..o
22550 :< 252 229.50 :< 22950 =< 229.50 =<
8 9:5 .355 52.5 52
w:_5::m 3.5: 5 m2mc: 50::2:
5:: £55.: :: 35:... 85.3.0...

 

m:_5::m

:: mEfimcc .:8 55

SE: 5 @525. .85 .m—

m::: 5 3.52:... .80 .m.

@585
:5... .8 &::..<

 

 

 

52-2

 

53-2

 

.93.

 

53-2

 

53-2

 

 

. «52...:

129

 

 

 

 

 

 

 

050:0.
5:: 85350:

 

 

 

 

 

 

 

 

 

 

 

 

0.2::505
tam
0.2:..50 =< 0.805 0.2::50 :< 0.2:..50 :< 0.2:..50 =<
tam 0550 x... o. 5::
505.2:
.82 25> 322.3.
5:: 52: 15> 5:: 5:: 5::
SEE b.:0:.:0 ...E. v .50 .05.: w5.5:: v .50 .05.: 9.55:: v .50 .05.: M5553,
55-..: .93: 5..-: ...E: ...E:

 

. 052...:

130

 

 

 

 

 

 

 

58:0.
5:: 0053.0:

 

 

 

 

 

 

 

0.0—9.50::
m0:o
Em: 0:. 5: 55 500.05%
55:8. .555 .0::::m , 55:00. .555 .0::::m 5:o:m0. .555 .0::::m 3055:“. 0.2::50 =<
5:: 0555 M55)
5:: 0:00 5:5: b.:0:.:0 b.:0:.:0 b.:0:.:0
5:: _ 5.5555: 5F .5: _ .5555: ...E. 5:: _ 5:55.55: 55:. 505.2:
2005 2005 v.55 £055.00 5:: 2005 xi 50:58 0052.0...
:5 5:5. Em: 0:. 0.08m ,5 5:5. Ew: 0:. 202m 5 5:5. Ew: 0:. 202m 5:: £25 0.02m . . .
5:: 5:: 5:: 5:: 5::
5:. 5.88:0 E... m: .... 5.85:0 .5.. :3: 5.85:0 5.; 2...: 5.58:0 .5... 2...: 55.058 ...E
...E: 5.5.: 55-..: 55-8 52-0:

 

 

 

 

 

 

. ”52...:

131

 

 

 

 

 

 

 

 

288.
E: 33.28:

 

 

 

 

229:8:—
uquEoU :< 803800 :< 0.0.6.500 :< 82an0 :< 33.5.;
E ...E H. .0» :5»
3E: v 9.3.58,
a 35.2:
E.“ 5.5.. 8.23.”?

 

9.5. v 33:28

3...: v mic—.28

m..:: v @2588 EFF

 

b.5950

b.:o&mu

£5. v onEmU

3E: v onEmU

0%.“. E Econ.

 

 

 

.32-..

 

>326

 

32%

 

32%

 

.826

 

 

_ “52...:

132

 

 

 

 

 

 

 

2800..
0:: 00...>..00

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ 052...:

0.0.9505
...E. 0:0 .25 E 00:89: ...E
. ..SE. 8 ...wso:0 0E...
$39.58 $04 >00 0... mo 69: wcioofi 0.0.950 :< 0003,50 =< 0.0.950 :<
30:02:50 :53.
38:02:00 :52. 3E: m E Eco: 203:. 30:58 0055...
0:: 32 00.550 0.00% 3.0.3.3:
3.5. m :. Eco: ..Smc. Econ. 0.0on 0.80.. ..NBED :80: :0st w:zm....o::n.
3.2.3 32-2 32-9 32-. . 32-2

133

 

 

 

 

 

 

 

 

5030..
0:: 823:0:

 

 

 

203500:—
.:: 03: 3:0 5 00:85.
803500 :< 093 Em: 09a Em: 303500 =< 532.505 304
.2995 “.50.: 5:995 .2995 .50.: 5.39%
85.0.:
30:05:00 :92. 80:25.0 0 :32. 30:25.0 0 :83. 80:05:00 :83. 30:05:00 :55. m0...>:0<

 

35: N 5 5000 :85.

3.5. N 5 500.0 :52.

$5: N E 200: :23.

3.5. N 5 £000 :03:—

m..:: m 5 2000 :32.

 

 

 

32-. N

 

€2-00

 

502-2

 

$2-0.

 

.82-» .

 

 

_ 052...:

134

 

 

 

 

 

 

 

20000.
0:0 $23.00

 

 

 

 

 

 

 

0.0.0508.
00.0 03050.2 30.. :0:;
00.0 $30.50.). 50.00.... 3.000 .50.. ”0.53 03030:. 830.505 304 9503500
@555 0:. .0550 ..0
2.5. N 5 E000 ..0.m:.
m .00.:00.m. m .0252: .8502:
00.50.:
m 0<>I m 0<>I 0<>I 58:05:00 20.2.. 50:05:00 20.5.. 00:550.:
m wEnEEA. m MEEEEL 9.5520. 35: N .... 0.00: :32. $5: m :. £000 :30... . . .
32-00 .22-: 202-00 202-3 00.2.00

 

 

 

 

 

 

. 022...:

135

 

 

 

 

 

 

 

0:00.00.
0:0 00...>..00
0.0—0500:.

 

 

 

 

 

 

 

 

005.0...
0052.0...

 

 

 

52-0.

.02-: .

.02-..

.026.

.02-m .

 

 

: 022...:

136

 

 

 

 

 

 

 

 

20000.
0:0 m0...>..00
0.035005

 

 

 

 

 

 

 

 

005.0...
m0...>..0<

 

 

 

 

 

137

 

 

8.8095.

...03 9.00 0 :0...

 

505.3 00000...
.0: 0.2.8 8.8.3:.

0.05 00 0.0...5005
.:0> m 0.00.0

 

 

 

 

 

 

 

w:..00:.m:0 08000.
0000: ..000: 00..; 0:0 00...>..00
30.5 oz 8.00:8. w:..00:.w:0, 0.07.508.
008.9.0 .0: .00... 0:N 0:000. 0.. 0. 0>0: 0..0..0. 00:... 505.3 5.3
:0..00.-...0.0 0.05 0500
008.9.0 .50. ....3 0.2.002. .0 000: 00.0...500 ..<
.00: .0. 8.82.5: 00.0...500 ..000. 00:..- .0: 8.8005. 8.0.5.... 0... .0. 0.07.5005 00.....-
w...0.00.m .8328. 50> m. 50>: 0.00.0
8.85. 80050:: ..000». 00...... ..000: 00...: £00050.
8.80.5. 0<>I 0080...
00...>..0<

8.82.5. w...0..:m.

8.80am... w:.0.5m

..03 ...-n. 8.8.3:.

...: 8.80.05 0<>:

w:..00:.m:m. 00....-

 

..0..0~. 00..-...-

..0..0~. 00:...

8.80%... w:.0..:m

0:03 ..n. 80050:.

2.03 55 8.82.5.

 

 

 

...<-N

 

...<-.

 

.02-. m

 

.02-om

 

.02-0N

 

 

.. 022...:

138

 

 

 

 

 

 

 

mcowao.

 

 

 

33.00020 30.0 0:0000. ....0.0 .0..000m 0:0 00...>..00
0.0.0500:.
m:.:0>0 0... :. ...:0
0000000.. 50802.
>000 0.0.0500 ..< 000000.. ...0.0 30cm 02 58000:. 300m 02 .0.000m:.
c0. 30.0 Hon 32m .,:0.0 ..03b0 m:.w:0:
m:.0.00.m ..0330
. :5. :03bn. w:.w:0: m:.0.00.m .5350 .0..0.:. 02.0.35
0 .5: :03bn. w:.w:0I N .5: ..03bn. w:.m:0: ..0..000m:. w:.0....m .0..0.:. :0..0.:m:. :0..000m:. 0.50:3. 00.5.0.0
m00.3.3.~

 

m ..0: ..03EQ w:.m:0:

. ..E. ..03bQ w:.m:0..

.0..0.:. 0000.005

:0..000m:. w:.0..:m.,,

m:.00.m. 3.0.0

 

 

 

.0....mV

 

.0...-»

 

.0<-..

 

.0<-0

 

.0<-m

 

 

.. 0500....

139

 

 

 

 

 

 

 

mCOmmo.
USN m0_u_>_aom

 

 

 

 

N. .5: :03bm. 050:0:

o. ...... ..03bn. w:.w:0..

w ..5. ..03bn. w:.w:0...,

m. .5: ..03bn. w:.m:0I

v ..E. =03bn. w:.m:0I

0.0.0500:.
000 0:0 :. 000 0:0 :.
. M05 00...>..00 00E. . M05 0.0.0500 ..< 00...>..00 00....
:. 0: 00.. 000.0003. ..0 0.0.0500 .,:0.:00 ... 0: 00.. 000.0003. ..0 0.0.0500 .,:0.=00
N. .5: :03bn. .0530: . 3.00% 0 .5: :03bn. 3.9.0:
. ”1.5000. . . ...... ..03bn. w:.w:0I 0 .5: ..03bn. w:.w:0I .. ...E ..03bn. m:.w:0.. m ..5. ..03bn. w:.m:0.,. 00:50.0
m0...>..0<

 

 

0..0.

 

...}.

 

 

0..-:

 

0..-:

 

0..-”.

 

 

: 052...:

140

 

 

 

 

 

 

 

0:800.
0:0 00...>..00

 

 

 

 

 

 

 

 

 

 

 

 

0.0.0500...
0.0.0500 ..< 0.0.0500 ..< 0.0.0500 ..< 0.0.0500 :< 00.5.00. ..03b0
.0. m0: ..000. .0:...
0.5: 0 m0.w:0

0:0 05.0.. :0 0.5.000 00::0.0

mc..00... 0m:0..< m0...>..0<
m..:: 0 9.5.5.. .000 .m. 0.5: 0 9.0.0.5.. .000 .m. m..:: o m0.w:0 0..0: o m0.w:0 wc...m.:_..
0:0 0.0.0.0 ..0 0.500% 0:0 05.0.. ..0 m:.000. .01. m:.0.00.m .0..0.02

a .930 ...E. aim .330 5..-...

 

= 052:...

141

 

 

 

 

 

 

 

000000.
000 00...>..00

 

 

 

 

 

 

 

 

 

 

 

 

0.0.0.0000.
0.2980 .2 220.80 =< 32.58 .2
00:00.0
.000 0.. 0000.0 m:.0:0m 0052.0».
05050 050500 =0 0:205. .80 EN
a<-om 02-00 .9.-." .3.-.” .300

 

0 050......

142

 

 

 

 

 

 

 

0:00.00.
0:0 00...>..00

 

 

 

 

 

 

 

 

 

 

0.0.0580.
0.0.0500 ...... 0.0.0500 ..< 0.0.0500 ..< 0.0.0500 ..< 0.05
000 000 0.0.0500 ..<
00000.0
00...>..0<
0.5.. 0 05.00... 35>
.0.00 ....3 0.. ..000. 0..0: 0 05.00... 35> 0..0: 0..0: 0..0:
m .000 .05.... 05.50.. m .000 .05.... 05.500 .0 .000 .05.... 05.500
1 002.0 002-0 00.2% >02-.. 00.2%

 

 

 

 

 

0 050......

143

 

 

 

 

 

 

 

 

000000.
000 00...>..00

 

 

 

0.0.05000.
.00.00. .0 000000..
.0030. 0. ..03b0v
0000.5. 0000 0.. 0.0.0500 ..< 0.0.0500 ..< 0.0.0500 ..<
0.0
0000...... 0.. 0.00.000 0.2.500 0.8.0 0:000. :03»... 00:00.0
00...>..0<

 

050000. 2030.0.

0.00Q 0.09m

.0..0.05 b50900 0.02m

05000.0. ..03bn.

w0..0.. 00:00

 

 

 

.02-...

 

0.2-0.

0.2-0.

 

 

.02-. .

 

.02-...

 

 

: 050......

144

 

 

 

 

 

 

 

000000.
000 00...>..00

 

 

 

 

 

 

 

0.0.0500:.
0.0.0500 ..< 0.0.0500 ..< 0.0.0500 ..< 0.0.0500 ..< 0.0.0500 ..<
0..03 .0...0m. 0500.0...
0003 .0...0m. 00.00.50, ..03 .0...0m. 00.00.50, ..000. 0:03 .0...0m. 0500.50,
..03b0 .0...00 m500... 00000.0
..000. ..000. ..000. ..03bn. ..000. ..000. 00...>..0<
..0330 .0...00 0500.0 =03b0 .0...00 @5000 ..03b0 .0...00 0500.0 ..03b0 .0...00 00.000
..000. ..03bn. ..000. ..03bn. 09.0.0000 0.09m ..000. ..03bn. ..000. ..03bn.
0.2.0 0.2-00 002-0. 0.2-0. .02-:

 

 

 

 

 

 

.. 050......

145

 

 

 

 

 

 

 

000000.
000 00...>..00

 

 

 

 

0.0.0500:.
.0300005 0000 0.0.0500 .0.. 0.0.0500 ..< 0..000. 0..000.
0.05 00... 0..0 0.00. 0.03. 0.05 00... 000 0.00. 0.03.
0..000. 203.05.

0 00000.0
0.53 .0...0m. m:...0.:.... 0..000. :03>.0 0.0.2 0..03 .0...0m. 0500.0... 00...>..0<
0..0: N 05000.00 0..03 .0...0m. 0.0.00.0... ..000. 0:03 .0...0m $0.05... ..000. . . .

..03b0 .0...00 m:.000. ..03b0 .0...00 00.000

 

 

 

0.2-00

 

0.2-00

 

0.2-00

 

0.2-00

 

0.2-00

 

 

0 050......

146

Appendix E

PPC Calculations and Reason Codes

147

 

 

 

 

002...

 

3000 02

 

2.0.: 0. 0...".

 

.00.00>>

 

00.0.08 .0030

 

0000000000 002

 

0.0000. cm

 

3.0000000.

 

 

end 00.0

02

 

 

on:

on

mnd 00.0

oméo

 

 

wwdw
N

cow
N

 

 

3.. on...
00.0. Eoo wxmmh

 

 

582. 8:9

 

.2025 3:2

m

N

 

00002.”. 9.03

 

Dan.

F
Egrm .91...

a"-..

a<0

02-0

Qua

a<-~

.90;

00.2-3”

00.2-9”

 

\I‘

 

 

 

 

 

 

00.0.: 0. 0...“.

 

.00.00>>

 

00.0.03 .0030

 

 

 

 

 

 

de mod

om oow

mmdm

oow

8.8

omd 8.0

 

_. N

F

m

_.

 

 

on
F

on:

c

on

 

N

0

II.

F

 

...... 0%.
080.9000 wxmah

 

.2025 8:9

 

N N

 

 

m

 

m

 

m

.282. 8:9

 

N

 

N

 

N

 

N

 

00000.0 wxmuh

 

 

can.

 

 

ESTwN .mS.-mN .05.-..N ESTMN ESTNN

 

Dan.

 

 

 

52.9 00$73 00.2-: 00.2-9 002-3

 

 

 

_ 05.2.30

 

148

 

 

 

 

 

 

00.0.: 0.. 00....

 

.00.00>>

 

00.0.00... .0030

 

0000000000 002

 

0.0000. cm

 

E0585

 

 

 

omd mnd

 

 

00F

00F

 

09

mm

meow

 

va— oF.o

owbw
N

 

mmdo
N

 

09mm.
N

om
F

 

.0.. 0%.
030. E00 00.00k

 

.282. 8:2

>_0.00>I>_ 00:9

0

m

m

N

00000.0 00.00.F

 

oan.

>020

>053

>053.

v
0.2.0

Gun.

.9200

 

02-00

.0<-mN

02-5

aih

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mud— 00.0

09

o

 

o

02
F

08
F

 

 

wad— om...

09
F

on
F

 

ooF
F

 

09
F

 

2:
N

 

00.0. E00 00.0:

 

582.. 8:9

F

F

5003- 00:2

F

N

F

F

N

00000.0 00.0:

 

Dun.

 

 

02-00

02-00

 

 

.
32:3

 

a<.oIN

 

.
09$.

Ono.

 

 

 

a<.wlF

02-0.

 

a<-$

03-0.

 

03:0.

 

 

 

_ 00:20.0

 

 

149

 

 

 

 

 

 

00.0.: 0.. 000

 

00500.5

 

020.000 000.50

 

0000000000 002

 

0.0000. cm

 

E03800

 

 

 

mvd ood

 

 

00.00

om

 

 

00.00

.0... 000

 

N

o
o

o
o

00.0.9000 00.0:

 

0.0.8.5 8:9

0:82.. 8:9

MN

m

N

N

00000.0 00.03

 

 

000.

030-0

.570

cafinN

:2);

0.2-.0

000

 

i

0.2.00

00.2-00

 

 

.02-00

.02-«N

 

 

 

 

 

 

 

0>000 oz

 

00.0.: 00 000

 

00500.5

 

020.30 000.50

 

0000000000 002

 

0.0000. cm

 

N

N

3.003800

 

 

00.0 00.0

 

 

00F

00F

omd 00.0

mm.mm

ooF

 

 

o
o

o
o

N

o
o

o
o

F

N

02

II.

F

 

 

.0.. 00..
00.0. 0.00 00.00-F

 

 

 

2082. 8:9

N

N

N

N

 

20.82. 8:9

N

m

N

F

00000.0 00.000-

 

 

 

000.

 

 

.02-.0

 

00.2-8-

 

 

00.2-0.

 

00.2-0.

 

00.2-..

 

000

0.2-v.

00.2-0.

 

 

00.2-0.

 

00.2-. .

 

 

 

 

_ 00.2.30

 

150

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

151

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

000

.
. . .
. N .
. .
. .
. . .
.0... 8... 8 8. 00.00 00.00 8 S... 8... o... o 8 om
. N N . N . o N N 0 8.0. 050 9.03
0.0.8.... 8:9 N N 0 0 v 20.83 8:9 N v .- v 0 85.0.0 9.00.
000 a; a<0 .3.-. .30 a<-m 0%. a<-N a<-. 0.2-.0 0.2-8 8.2-0.0
00.0. 500 00.00...
>_x00>> 00:0. >200; 02.0. 00000.0 00.00-r
0%. 02.8 .0:-0N 0.2-.0 0.2-MN 0.2-NN

 

0057B 00.2-8 00.2-: 00.).-olF 00.2-3

 

 

__ 00.0.0.0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00F

00F

 

 

 

02
F

 

 

om

F

.0.. 0%.
00.0.mE00 00.00-F

 

 

N

 

00000.0 00.00-F

 

 

.0.):

>020

 

.
..I-0N

 

aih

 

 

 

 

 

 

 

 

 

00.0.000 00030

 

0000000000 002

 

 

 

 

 

 

 

 

 

00_-

 

 

OD.‘

 

F F
090.00 00<-NN

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

{
{

   

 

8.0. ...8 0.0...

8.00... 0:00.

 

 

152

 

F F .550

 

3a m

 

.2. am

 

25F

 

Bocm oz

 

SE: 9. 9E

 

35925

 

53300 3:30

 

cuanhccoo :oz

 

5.02.: cm

 

33:03.92“.

‘—

 

 

5d 8.0 SF SF on 88 30%.

 

 

 

o
o F N F N tam—mEoo 97.5...

 

582, 8:2 58k, 8:9 F N N m .822“. 93¢

 

 

F
0%. 53. 52” SE 53 >2sz 0%. 32.3 32% 52-3 32-3 32$

 

 

 

 

 

 

 

 

 

Bocw oz

 

F F 23: 2 2n.

 

352:5

 

:o_m_uwo ..oEsO

 

8:25:50 :02

 

5.59: cm

 

3:262".

 

 

 

ooF 2.6 00.0

 

 

Do... 00... 00F
[ m

8 8F 02 02 Son:
>100; .093; n F
hi EEEEEH 53; sing? N N -4!“ {32 58%:

N
__ £23 555% I55: 3...:

 

 

 

 

 

 

 

 

 

O
”('70
‘_
D
O
V—
O
O
‘—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

153

 

Appendix F

Formats for Data Collection and PPC Calculations

154

 

.85 =o=8=oo 53 £8 ”Fm. 2:9”.

 

 

 

 

 

wcommwh

 

was $2265

 

203E005

 

 

 

 

 

 

 

 

umccmi mmE>=o<

 

 

 

 

 

 

 

 

 

5.2-2 ._m_2-mF 52.: 82-9 5.2-9

 

155

 

mcuoo comma”. a cows—5.50 on“. ”Nd 2:9".

 

 

550

 

oomnw

 

.2395

 

aunt.

 

265w oz

 

33552 9:".

 

.5533

 

c2325 ..oEsO

 

mucmctzcoo coz

 

mctmwfimcm

 

£28285 ,

 

 

E on:

 

umumEEoo mxmmh

 

 

 

 

 

umccmE mxmfl.

 

 

 

0%. >282,

 

Dan. U®=9

 

..mSTmF

.22-?

52.:

._m_>_-©F

8.2-3

 

 

156

Appendix G

Production Planning Model

157

 

 

 

 

 

Standard Template
Amie Plan; Sales;
\ Design
Specific Template for

Each community
At Dates Based on Permits;
\ Resource Availability

Specific Schedule for < Should >
Each Home/Unit . .....................

E Centralized Schedule System :
-------------------------- Trade Contractor, Supplier,
Inspector, Superintendent,
PM Involvement

Short Term Lookahead ‘ C i
Plan

AN
Current Weekly
Progress Update

<' WILL > f
Weekly Production f EASURED

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 
 

Constraint Analysis and
Make Work Ready

  
 

 

 

 

 

 

 

 

 

 

 
 

 

     

 

 

 
 

 

 

l M
P an rolled PPC/
Sigma Value
U u
@ Tasks Completed Data AMUS‘S/
Charts/ Pareto

   

Analvsis

 
 
 

DID
ANALYZE

  
 

 

Brainstorming/
Reengineering

 
 

 

 

 
 
   

IMPROVE

 

Figure 1: Production Planning Model

158

 

DI ‘ui. i

 

 

    

II I III IIII III II

1293 02736 036