2.431.... a : .gg-£\:FF‘E33 .En.
. . . an

. ,, . . . . I. a Z...“ 2..“ . 4..
. . . . . , :5.

.. . E

J.
u

_.
,.“._.>..~. . . . , . , .
5...... j . . . . _ . . . _ a...

t. .t
5.. A.
:5

I. V
J .1... I. 1. .
1......

.

 

 

..:.h§.......r.._».ei§a.§c2.... , . 21.5... 253...... , _ , 4 .. : , L

 

T'Hrws

300\

Date AV-‘e
0

 

0-7 639

 

 

, llBRARY
lMichigan State [9
University

 

This is to certify that the

thesis entitled

PRODUCTION AND MATERIAL FLOW PROCESS

MODEL FOR MANUFACTURED HOUSING
presented by

OMAR SENGHORE

has been accepted towards fulfillment
of the requirements for

MS degree in Construction Management

mafia,

Major professor

137 Ram

MS U 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.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIFlC/DateouepBS-nts

 

 

PRO

 

PRODUCTION AND MATERIAL FLOW PROCESS MODEL FOR
MANUFACTURED HOUSING

BY

OMAR SENGHORE

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE
Construction Management Program

2001

l
PF;

 

 

line manul“
innovations

Advanceme
irnlDrovemen
description c
Chapter thre
showed hon
i”Wow-3d ar
all the majo;
fioWCth‘Ung
mapping Sta
deV‘~"<>necl.

USing basic 1
and Normals

DrOCESs base

W1 other 9e

ABSTRACT

PRODUCTION AND MATERIAL FLOW PROCESS MODEL FOR
MANUFACTURED HOUSING

By

Omar Senghore

The manufactured housing industry has been lagging behind in its technological
innovations compared to site-built housing according to PATH (Partnership for
Advancement of Technology in Housing. This report is a contribution to the
improvement of the production process of manufactured housing. A detailed
description of the production process and industry temiinology was made in
Chapter three to help readers not familiar with the industry. The research
showed how the production process of manufactured housing plants can be
improved and resource utilization streamlined. First, a process flow diagram of
all the major activities in each station is developed with the help of ANSI
flowcharting concepts. An extensive data collection phase followed the process
mapping stage. A detailed production process and material flow model was then
developed. This model was then transformed into a computer simulation model
using basic modeling techniques developed by Halpin (1992). Combis, Queues,
and Normals, were used to input the model into a simulation software called
EZSTROBE. The simulation software showed bottlenecks in the production
process based on the time data inputted. The bottlenecks were used together

with other generic what if scenarios to show how the process can be improved.

l l
committe
Syal for ‘
guidance
Tariq Ab
part of tt
Rex LaM
demandi

ln
Partner h
year and

To

helped m

ACKNOWLEDGMENTS

I would like to express my sincere thanks to all members of my thesis
committee, Dr. Matt Syal, Dr. Tariq Abdelhamid, and Dr. Rex LaMore. Dr. Matt
Syal for the opportunity to work on such an exciting project and for his thorough
guidance during my study at Michigan State University. I would like to thank Dr.
Tariq Abdelhamid for his incredible patience in guiding me through the simulation
part of this research. All the comments and help are deeply appreciated. Dr.
Rex LaMore for his willingness to serve on the committee despite a highly
demanding schedule.

In addition, I am very grateful to my classmate, friend, and research
partner Namita Mehrotra. It has been a pleasure working with you for the past
year and a half and thank you for putting up with me and for all your help.

To all the industry personnel I have met during this research, you have

helped me gather valuable ideas to complete this work.

LISTOF'
LISTOFI
CHAPTEF
1.1 OV
1.2 NB
1.3 RE!
1.4 GO.
1.5 ME‘

1.55

1.5.1

1.5.5

1.5.4

1'6 EXPE

TABLE OF CONTENTS

LIST OF TABLE ........................................................................................... x
LIST OF FIGURES ...................................................................................... xi
CHAPTER 1 INTRODUCTION
1.1 OVERVIEW ....................................................................................... 2
1.2 NEED STATEMENT .......................................................................... 3
1.3 RESEARCH SCOPE ......................................................................... 4
1.4 GOALS AND OBJECTIVES .............................................................. 5
1 .5 METHODOLOGY .............................................................................. 6
1.5.1 Developing a Production Flow for Manufactured Housing ..... 6

1.6

1.7

1.8

1.5.2 Perform Time Studies for Each Production Assembly Step 8

1.5.3 Development of Detailed Individual and Integrated Process

Model for Production and Material Flow ................................. 8
1.5.3.1 Process Modeling Concepts ............................. 9
1.5.3.2 The Modeling Procedure .................................. 11

1.5.4 Transformation of Integrated Process Model into a Computer

Simulation Model .................................................................... 12
EXPECTED OUTCOME .................................................................... 16
1.6.1 Streamline Production Process .............................................. 16
1.6.2 Technological Improvement ................................................... 16
1.6.3 Computer Simulation Software for Plant Layout ..................... 17
REPORT ORGANIZATION ............................................................... 17
SUMMARY ........................................................................................ 1 8

CHAPTEF

2.1

2.2

2.3

INT

EXIS‘
2.3.1

CHAPTER TWO LITERATURE REVIEW

2.1

2.2

2.3

INTRODUCTON ................................................................................ 20
DEFINITIONS ................................................................................... 20
2.2.1 Manufactured Home ............................................................... 20
2.2.2 Modular Home ........................................................................ 21
2.2.3 Mobile Homes ........................................................................ 21
2.2.4 Panelized Home ..................................................................... 21
2.2.5 Pre-Cut Home ......................................................................... 22
2.2.6 Manufactured Home Communities ......................................... 22
2.2.7 Single—Section Home .............................................................. 22
2.2.8 Multi-Section Home ................................................................ 22
2.2.9 Site-Built Housing ................................................................... 22
2.2.10Retailer ................................................................................... 23
2.2.11 Dealer ..................................................................................... 23
2.2.1ZDistributor ............................................................................... 23
2.2.13Manufacturer .......................................................................... 23
2.2.14Installer I Servicer .................................................................. 23
2.2.1SConsumer ............................................................................... 24
2.2.16HUD Code .............................................................................. 24
2.2.17BOCA Code ............................................................................ 24
2.2.18lntemational Residential Code ............................................... 24
EXISTING RESEARCH ..................................................................... 25
2.3.1 Manufactured Housing Industry Research ............................ 25

 

24 SU

CHAPTER
31 INT
32 PRC

32.

32;

 

 

 

32.
321
32}
33 00h

HOlr

2.3.2 Construction Simulation Research ......................................... 29
2.3.3 Production Planning Research ............................................... 31

2.4 SUMMARY ........................................................................................ 35

CHAPTER THREE MANUFACTURED HOUSING PRODUCTION PROCESS

3.1 INTRODUCTION ............................................................................... 38
3.2 PRODUCTION PROCESS ................................................................ 38
3.2.1 Floors ..................................................................................... 39
3.2.2 Walls ...................................................................................... 40
3.2.3 Roofing ................................................................................... 42
3.2.4 Exteriors ................................................................................. 44
3.2.5 Finishing ................................................................................. 45

3.3 COMPARISION OF MANUFACTURED WITH SITE BUILT

HOMES ............................................................................................. 46
3.3.1 Code Comparison .................................................................. 46
3.3.2 Production Environment ......................................................... 52
3.3.3 Foundation ............................................................................. 52
3.3.4 Cost Comparisons .................................................................. 54
3.4 SUMMARY ........................................................................................ 57

CHAPTER FOUR PRODUCTION PROCESS & MATERIAL FLOW MODELS
4.1 INTRODUCTION ............................................................................... 59
4.2 DATA COLLECTION ......................................................................... 59

vi

43 PF

4.4 DE

45 SU

CHAPTEF
51 INT
52 VVF
53 TYI
53.
53.

5-4 WH

4.2.1 Process Mapping .................................................................... 60

4.2.2 Process Flow Diagrams .......................................................... 64
4.2.2.1 Standard Flow Chart Symbols .......................... 68

4.2.3 Time Study ............................................................................. 69
4.2.3.1 Cycle Time ........................................................ 70

4.2.3.2 Data Sampling and Population ......................... 73

4.3 PRODUCTION PROCESS FLOW DIAGRAMS BY CLUSTERS ...... 73

4.4 DETAILED PRODUCTION PROCESS AND MATERIAL FLOW
MODEL ............................................................................................. 82

4.5 SUMMARY ........................................................................................ 94

CHAPTER FIVE SIMULATION MODEL

5.1 INTRODUCTION ............................................................................... 96
5.2 WHAT IS SIMULATION .................................................................... 96
5.3 TYPES OF SIMULATION .................................................................. 97
5.3.1 Discrete-Event Simulation ...................................................... 97
5.3.2 Continuous Simulation ........................................................... 97
5.4 WHAT IS SIMULATION USED FOR ................................................. 98
5.4.1 Use of Simulation for Manufactured Housing ......................... 98
5.4.1.1 Identifying Needs and Allocating Resources 99
5.4.1.2 Understand the Interaction Between
Resources ......................................................... 1 01
5.4.1 .3 Trade-Off Alternatives ....................................... 102

vii

5.5 CC
5.6 IN‘I

5.7 ST

5.7
5.7
5.7
5.7
5.8. U8;
WI

5.8.

5.8
5.8.
5.8.
5.8.:

SUI,

5.4.1.4 Determine Durations of Operations and

Operations Cost ................................................ 103

5.5 CONSTRUCTION SIMULATION ...................................................... 104
5.6 INTRODUCTION TO EZSTROBE .................................................... 104
5.7 STEPS IN SIMULATION MODELING ............................................... 105
5.7.1 Define problem ....................................................................... 106

5.7.2 Develop Logic Network .......................................................... 106

5.7.3 Determine or Estimate Operations Time Durations ................ 107

5.7.4 Transfer Logic Network into Computer Simulation Program .. 107
5.7.5 Run the Model ........................................................................ 115
5.7.6 Validate the Model ................................................................. 122

5.8. USE MODEL FOR DECISION MAKING I IMPROVEMENTS]

WHAT IF SCENARIOS .................................................................... 122

5.8.1 SCENARIO ONE .................................................................... 123
5.8.1.1 Sensitivity Analysis ...................................................... 125

5.8.2 SCENARIO TWO ................................................................... 128

5.8.3 SCENARIO THREE ................................................................ 130

5.8.4 SCENARIO FOUR .................................................................. 132

5.8.5 SCENARIO FIVE .................................................................... 134

5.9 SUMMARY ........................................................................................ 135

CHAPTER SIX SUMMARY AND CONCLUSIONS

6.1 OVERALL SUMMARY ....................................................................... 137

viii

6.2 SU

6.2

 

6.2

6.3 CC
64 RE
6.5 AR
APPEND

AF
REFERE

6.2 SUMMARY BY OBJECTIVES ........................................................... 138
6.2.1 To Develop a Production Process Flow Diagram for
Manufactured Housing Production ......................................... 138
6.2.2 To Develop an Integrated Production and Material Flow
Process Model for Manufactured Housing ............................. 139

6.2.3 Transformation of Integrated Process Model into a

Computer Simulation Model ................................................... 139

6.3 CONCLUSIONS ................................................................................ 140
6.4 RESEARCH LIMITATIONS ............................................................... 141
6.5 AREAS OF FUTURE RESEARCH .................................................... 142
APPENDICES .............................................................................................. 144
APPENDIX A - TIME DATA .............................................................. 145
REFERENCES ............................................................................................ 149

 

1.1
1.2

3.1

3.2

3.3

4.1

 

Aft.
80

Co
Sur

Inte

 

Sun
Pro.
UCI
Not

Lat

1.1
1.2

3.1

3.2

3.3
4.1
4.2
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9

LIST OF TABLES
Manufactured Home Shipments ........................................................ 2
Affordability Status of Families by Type of Housing .......................... 4
Summary of Planning Considerations for HUD and BOCA
Codes ................................................................................................ 48

Summary of Fire Safety Considerations for HUD and

International Residential Codes ........................................................ 50
Summary of Cost Comparisons ......................................................... 55
Process Mapping Data Sheet ........................................................... 63
UCIMSU NSF Manufactured Housing Project Data Sheet ................ 71
Notations for Simulation Model ......................................................... 108
Labor Input Parameters .................................................................... 1 16
Material Input Parameters ................................................................. 1 17
Result Formulas ................................................................................ 1 17
Definitions of Queue Attributes ......................................................... 118
Definitions of Activity Attributes ........................................................ 1 18
Labor Utilization ................................................................................ 1 19
Queues Statistical Report ................................................................. 120
Activities Statistical Report ............................................................... 121

5.10 Labor Utilization with Reduced Durations on Roof Setting Activity... 124

5.11 Sensitivity Analysis Data ................................................................... 125

 

 

 

1.1
1.2

1.3

1.4
1.5
3.1
3.2
3.3
3.4
3.5
3.6
4.1
4.2
4.3
4.4
4.53
4.5b
4.5
4.7
4.8

4.9

Ma
Ba
Sc
Ste
80
Sin

Flo

Shi
Inst
TOp
Sing
Exa;
Exai
Estir
Stan
PrOQ
Prom
Prose
PrOCe
prOCe

proce:

1.1
1.2

1.3

1.4
1.5
3.1
3.2
3.3
3.4
3.5
3.6
4.1
4.2
4.3
4.4
4.5a
4.5b
4.6
4.7
4.8

4.9

LIST OF FIGURES
Manufactured Housing Production Process ...................................... 7
Basic Modeling Elements .................................................................. 10

Schematic Relationship of Constituent Activities on a Roofing

Station ............................................................................................... 12
Schematic Representation of Hauling Operation .............................. 13
Simulation Model .............................................................................. 15
Floor System Being Built ................................................................... 39
Wall Feeder Station .......................................................................... 41
Shingles Installation .......................................................................... 43
Installation of Exterior Doors 8. Windows .......................................... 45
Top View of Footing Spacing ............................................................ 53
Single Block Pier ............................................................................... 54
Example Activity Relationship ........................................................... 61
Example Flowchart of Earthmoving Process ..................................... 64
Estimating Process Using ANSI ........................................................ 67
Standard Flowchart Symbols ............................................................ 68
Process Flow Diagram - Floor Cluster ............................................. 74
Process Flow Diagram - Floor Cluster Cont’d .................................. 75
Process Flow Diagram - Walls Cluster ............................................. 76
Process Flow Diagram — Exterior Wall Setting & Fastening ............. 77
Process Flow Diagram — Roofing & Exterior Wall Finishing ............. 78
Process Flow Diagram — Interiors ..................................................... 79

xi

 

 

4.10
4.11
4.12
5.1
5.2
5.3
5.4

5.5

Sc

4.10 Process Flow Diagram — Finishing Cluster ....................................... 81
4.11 Production and Material Flow Model — First Half of Factory ............. 83
4.12 Production and Material Flow Model - Second Half of Factory ........ 90
5.1 Need and Resource Allocation Example ........................................... 101

5.2 Simulation Model of a Section of Manufactured Housing Production110

5.3 Sensitivity Analysis — Duration Vs Utilization .................................... 126
5.4 Sensitivity Analysis - Duration Vs Production Rate .......................... 127
5.5 Schematic Layout of Double Production Line Scenario .................... 131

xii

CHAPTER ONE

INTRODUCTION

 

 

 

1.1 OW

Thl
Developrl
single-far

code man

back in Ihr

 

trailers"
used in ca

easily tran

and their fa
beCause 01
811 to 12ft l
IntTOduced
i"dUSVY wr-
PGOpIe aCrr
Oi life“ (MI “-

manUfactUi

,8’
g

ear
013i

'ngie-

UIII echo”
”'1

tarts Qie-

('5
as.

an»
22*

(593'
66;;
5'0.

.59?

-\

Sal
r

1.1

OVERVIEW

The United States Census Bureau has estimated that Housing and Urban

Development code manufactured homes represented 22.7 percent of all new

single-family housing starts in 1998 (MHI Quick facts 1999-2000). The HUD

code manufactured housing has been evolving since 1976 and goes even further

back in the early part of the 20th century when they were called trailers. These

“trailers” were only 8ft wide and designed to be very mobile. They were mostly

used in campgrounds but slowly changed into mobile homes, which were less

easily transported. Mobile homes served veterans returning from World War II

and their families very well, in the midst of the housing shortages of that period,

because of their mass production. The homes continued to evolve, growing from

8ft to 12ft to 14ft. In the mid 1970’s, doublewide manufactured homes were

introduced, followed by multi-section homes in 1985. Today, it is a thriving

industry with growing increases in sales, production and size. “Over 19 million

people across the country have decided to make a manufactured home their way

of life” (MMHA 2000). The trends, as summarized in Table 1.1 clearly show that

manufactured housing has become a viable option to homebuyers in America.

Table1.1 - Manufactured Home Shipments

(Source: MHI Quick Facts 1999 - 2000)

 

 

 

 

 

 

 

 

 

 

 

 

Year 1992 1993 1994 1995 1996 1997 1998I
Total 210,787 254,276 303,932 339,601 363,411 353,377 372,843
Single-Section 112,117 134,440156,171 173,785173,674148,809 144,328
Multi-section 98,670 119,836147,761 165,816189,737204,568 228,515
% Of Single-Family

Starts 17 18' 20 24 24 24 23
IEsti mated Retail

Sales (Billions) $5.99 $7.76 $10.18 $12.33 $13.96 $14.52 $16.33

 

 

 

 

 

 

 

(PATH), r'
technolog
typical m2
and capa:
flow and L
market in
emerge as
plant fails
shortcomir
Ofextensiv

housing (m

 

differeUCe
prodUCIIOn
hOUSing- Ir
Material Uti I
procuinn I
Hastak 200'
ACCO
married COU

1.2 NEED STATEMENT

According to the Partnership for Advancement of Technology in Housing
(PATH), manufactured housing industry has been lagging behind compared to
technological innovation of the site-built housing industry (PATH-HUD, 1999). A
typical manufactured housing plant is unable to meet the desired production rate
and capacity due to the lack of a streamlined assembly process and the efficient
flow and utilization of material in the production process. Despite the growing
market in this industry, manufactured housing industry has not been able to
emerge as a technologically advanced industry. A typical manufactured housing
plant fails to produce at desired capacity and production rate because of several
shortcomings in the production process. Such shortcomings include (1) the lack
of extensive repetition in the manufactured housing production process. Each
housing unit going through the assembly process is unique in itself. (2) The
difference between site-built and factory-built housing hinders application of
production process innovations in site-built housing research to manufactured
housing. In order to develop an effective and efficient system for production and
material utilization, it is important to identify the deficiencies in the current
production processes used by the manufactured housing industry (Syal and
Hastak 2000).

According to 1995 census bureau data on housing affordability, 42% of all
married couples could not afford a median priced home in their community and

well over half of them (54%) could not afford a new single-family home.

 

Table 1

 

Using Cc
30Year
Type of F
Median P
Modestly
Low-Price

 

 

Mar|
problem ‘
manufactu
this makes
their own.
to the Over
undertakin!

Quarter of a

Table 1.2 — Affordability Status of Families by Type of Housing
(Source: Census 2000)

 

 

 

 

 

 

 

 

 

 

 

 

Using Conventional, Fixed Rate, Cannot Afford Criterion home in US.
30-Year Financing Total [Married-couple
Type of Home INumber Percent INumber IPercent
[Median Price Home 38,075 52.2 23,110 42.2
irredestiy Priced Home 32,396 44.4 18,739 34.2
[Low-Priced Home 29,002 39.8 16,479 30.1
INew Single-Family Home 45,916 63 29,729 54.3
Condominium 37,713 51.7 22,829 41.7
[Price Adjusted Home 37,534 51.5 22,686 41.4

 

 

 

 

 

Manufactured housing can offer a potential solution to the affordability
problem. The link between production, cost, and affordability is that if
manufactured housing production can be improved, and done at a lower cost,
this makes it affordable to most people who otherwise cannot afford a home of
their own. Furthermore, the affordability of manufactured housing will contribute
to the overall housing affordability problem in the US. This is a major
undertaking because manufactured housing as stated earlier, constitute around a

quarter of all housing starts in the US.

1.3 RESEARCH SCOPE

The scope of the proposed research is on how to develop a production
process and material flow model for manufactured housing with focus on multi-
section homes. The production process of two manufactured housing plants will
be studied extensively. The process will be documented in the form of a

production process model. Part of this model will be transformed into a

4

 

simulatio

 

simulatio
modeling
plant dBIE
productio
Before ex,
that the re
NafionalE
State Uni\
MSU is he
Mehrotra a

and three '

Cui.

1.4 GO}
The °VGraIl
manufactu',
ObIGCtives c
1. To dI
mam

2. To 05
menu

simulation model. Using time data that will be collected during the study, the
simulation will be run and resource utilization investigated. “The process
modeling of the assembly process along with the material inventory, storage, in-
plant delivery, and utilization process; promises the possibility of improving the
production efficiency, cost effectiveness, and quality” (Syal and Hastak 2000).
Before explaining the goals and objectives of this research, it is important to note
that the research for this thesis is part of a research project sponsored by the
National Science Foundation. The research is a joint effort between Michigan
State University (MSU) and The University of Cincinnati (UC). The team from
MSU is headed by Dr. Matt Syal and two of his research assistants; Namita
Mehrotra and Omar Senghore. The team from U0 is headed by Dr. Mark Hastak
and three of his research assistants; Ayman Abu Hammad, Varnshi, and Qingbin

Cui.

1.4 GOALS AND OBJECTIVES
The overall goal of this research is to show how the production process of
manufactured housing can be improved and material utilization streamlined. The
objectives of the research are:
1. To develop a production process flow diagram for multi-section
manufactured housing production.
2. To develop a production and material flow process model for

manufactured housing.

 

3.T

CO

1.5 MI

The prOpr

1.5.1 Dr

F ic
assembly
productio
activity wi
day and s
as aetivity
Teeter-y! er

TEIHIIOnSh

beginning

3. To select a section of the production line consisting of 34 stations and
transform the process model into a computer simulation model. The
computer simulation model will help in answering a number of what if

questions to determine how the process can be improved.

1 .5 METHODOLOGY

The proposed research steps consists of:

1.5.1 Developing a Production Flow for Manufactured Housing

Figure 1.1 below is a typical process flow for a manufactured housing
assembly layout developed by the research team. To finalize a generic
production flow model as such, with a little more detail, a process mapping
activity will be engaged in. This process will entail visiting the factory for a whole
day and spend the time studying the factory. While at the factory, certain things
as activity relationship, location of materials, amount of materials stored in the
factory, equipment used, number of labor per activity, and feeder station
relationship to main station, among other things will be observed. This will be the

beginning of the deveI0pment of the production flow diagrams.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

       

——
2‘ c a n
a E - or: 5 6 e 2‘ s '3 2
In I— 8 §= a - o a ‘u v
3 D ID I- - 5
'- 8is “5 °< 553§53
:2 m 2 g 8 3;, u s 3 .7
_ 7 ‘0
3 {will ,.‘- E
— It“ _. KIWI _= L-
.“l‘ In 2 IN 1' >. I“l In 2 — _ s‘v
II‘NI -‘ Q ”In! ‘5 = .. Ta 11“,, if 9 13: g o W
“IIIM‘ V g I; 3 2 8 8 g "Q5. " D I; E 2 E \IIIIIw
(“are j. 3 r: In — 2 \\l
‘4‘“.‘31 8 ”’ 'L w 3 8 5 "' 5 “Int
7
WI” ‘-' '1‘
‘. .i‘rI'i‘ L- t
“I‘m, \‘I‘ g g \‘
”Jun in”: .5 f, I
,n ~ 5 5 g
Irl‘II‘ S u. .
a N .5 3
8 \III“w . g
E \ii“Iw 0 g g
z \IIUI‘U g E
g ”Ni u.
3
2
e. 0“ no 5
i‘lm' g E E
IIIIII“ E a
null '- E
.“ , a as
c . J I a
, . U)
i— — K
\I fl u a. el ,
. ‘ *1 I—" ‘ .\
L /O‘ L e
\\

(Source: Research Team)

Flgure 1.1 - Manufactured Houslng Productlon Process

 

1.5.2 P
T

work at \

 

resource
Table 42'
material 1
activity in
and mate
fUnCtionin
absolute 9
Absolute (

at a Stati01

 

also be he

help to Get

1.5.3 Dev
pror

TWO
Norman In

Of the work:

FIQUre 1'1 \
COIIecmd f”

“8110, the h

1.5.2 Perform Time Studies for Each ProductionIAssemny Step

The time studies will be performed to measure the duration of production
work at various workstations. This will help in identifying bottlenecks and
resource utilization at each station. The format to gather time data as detailed in
Table 4.2 was developed by the research team. Crew size, composition, and
material and equipment usage will also be gathered during this step. Each
activity in the workstation will be measured together with the required labor force
and material. The time data will be used during the simulation stage to depict the
functioning of the production process. While doing the time studies, both the
absolute cycle time and total cycle time for each station will be recorded.
Absolute cycle time is the time when work is actually being done on the unit while
at a station. Total cycle time is the time the unit spends on a station. This could
also be translated into effective time and idle time. These two cycle times will

help to determine the type of delays that occur at each station.

1.5.3 Development of Detailed Individual and Integrated Process Model for

Production and Material Flow.

Two sets of data will be collected from two different factories in the
Northern Indiana area. This area is considered the manufactured housing capital
of the world. Two companies with plant layouts similar to the one shown in
Figure 1.1 will be selected for study. This is to ensure the uniformity of the data
collected from the time studies. To develop the material flow model, a materials

list for the housing modules is needed. This list will be used to determine how

 

material
each wo
together
observat
model wi
transfom
Ti
concepts
the proce

model of

1.5.3.1

TIIGII’ defi
breakdW
reqmred I

material is used on each workstation and the frequency of material usage on
each workstation. With that data in hand, the material flow will be developed
together with the detailed production process flow model and with the help of
observations made during the time studies and data collection. The process
model will eventually be used to show constraints in the assembly line when it is
transformed into a simulation model.

The development of the process model will be based on process mapping
concepts explained in Chapter four. These concepts are also compatible with
the process modeling concepts, which will be used to transformation the process

model of 3 to 4 stations into a computer simulation model.

1.5.3.1 Process Modeling Concepts

“Construction operations contain the basic work processes in construction.
Their definition requires knowledge of the construction technology involved, a
breakdown of the processes into elemental work tasks, the identification of the
required resources, and definition of the work assignment to the labor force

involved” (Halpin and Riggs 1992).

BASIC MODELING ELEMENTS

 

 

 

 

 

 

 

 

 

 

 

>

Modeling Element

NORMAL

COMBI

Q NODE

ARROW

COUNTER

The Normal work task modeling
element, which is unconstrained in its
starting logic and indicates active
processing of or by resource entities.

The constrained work task modeling
element, which is logically constrained
in its starting logic, otherwise similar to
the normal work task modeling element.

The idle state of a resource symbolically
Representing a queuing up or waiting
for use of passive state or resources.

The resource entity directional flow
modeling element.

A node which keeps track of the number
of times a unit passes it.

Name of Element Description of element

Figure 1.2 Basic Modeling Elements
(Source: Halpin and Riggs, 1992)

10

 

1.5.3.2

that ar
NOHTE
resour
units t
(4) Spa

releas

Consnt
tasks I
may or
roof st
0f the I

Consht

1.5.3.2 The Modeling Procedure

“The first step in model formulation is the identification of the flow units
that are relevant to or descriptive of the operation or process to be modeled.
Normally the units selected are physical items such as production units or
resources such as materials, money, and space. The more obvious types of flow
units that are of interest in construction are (1) machines, (2) labor, (3) materials,
(4) space or location, and (5) informational or logical permits (inspection
releases)" (Halpin et al. 1992).

The flow units that Halpin and Riggs (1992), are discussing could include
constituent activities within each station. Constituent activities are actual work
tasks perfomied on the house within the station. To illustrate, the roofing station
may consists of certain constituents activities as roof setting, roof decking, and
roof shingles. These constituent activities and flow units are the building blocks
of the process model. Figure 1.3 below shows a schematic relationship of

constituent activities on a roofing station.

11

 

Ins;

 

 

Set
the

 

Ext
Boar

 

Figure

Afte
reIaIIOnshi

M8I8ria] fiC

 

 

 

 

Roof Roof Cover Shingles

Insulation Decking Decking
i——’ '—‘> with Paper '——>

 

 

 

 

 

 

 

 

 

 

 

T

Setting
the Roof

 

 

 

 

 

 

 

 

Exterior Installation Siding
Boards of Exterior . I
% Doors and
Windows

 

 

 

 

 

 

 

 

 

Figure 1.3 — Schematic Relationship of Constituent Activities on 3 Roofing
Station.
After all the flow units / constituent activities are identified, their
relationship with each other is mapped out. The mapping of the constituent
activities on each station represents the production process flow diagram.

Material flow units will be used to model the material flow.

1.5.4 Transformation of Process and Material Flow Model into a Computer
Simulation Model
In this research, simulation is used to imitate production processes
happening in the factory and to help in making decisions among alternatives. It

is expected that this will help identify bottlenecks in the system and understand

12

 

 

how to t
then the
The obje
interactir
locate hr
and Rigg
transfom
operatiod
lTlOdellng

schemati'

 

USin
the abOVe

mOdEI for II‘

how to better utilize resources and materials. If such constraints are identified,
then the process can be manipulated to answer a number of “what if" questions.
The object of developing the model of a production system is to examine the
interaction between flow units, determine the idleness of productive resources,
locate bottlenecks, and estimate production of the system as constituted (Halpin
and Riggs, 1992). To illustrate how the integrated process model is going to be
transformed into a simulation model, let us follow an example of a hauling
operation (Sawhney and AbouRizk, 1993). This example follows the process
modeling concepts presented in the previous section. Figure 1.4 gives a

schematic representation of the hauling operation.

 

 

 

 

 

 

 

 

 

 
    
   
 

  

 

 

 

 

 

 

 

 

 

 

 

- Empty
Truck In
Queue e TI’UCk
Return
Front- Dirt
end Stockpile - .
loader Dumping
Truck
being , Travel .
loaded

 

 

 

Figure 1.4 - Schematic Representation of Hauling Operation
(Source: Sawhney and AbouRizk, 1993)

Using the basic modeling concepts and elements as shown in Figure 1.2,
the above operation can be converted into a simulation model. The simulation
model for this operation is shown in Figure 1.5. “In Figure 1.5, a COMBI node is

13

used to mod
men all the
for this 099'
queue 3 am
when one u
complete, it
code. Simil
trucks are a
function noc
and to stop

AbouRizk, 1

STROBOSC

 

 

used to model the loading operation of the truck. The truck can be loaded only
when all the required resources are available. The resources that are required
for this operation are an empty truck modeled as queue 1, a front-end loader as
queue 3 and a dirt stockpile as queue 2. The loading operation can only proceed
when one unit is available at Q1, Q2, and Q3. Once the loading operation is
complete, the trucks can start the travel activity, which is modeled as a normal
code. Similarly, the other two work tasks namely dumping and travel of empty
trucks are also modeled as a normal node. In addition to these nodes, there is a
function node that is provided to determine the number of truck cycles completed
and to stop the experiment based on the number of cycles” (Sawhney and
AbouRizk, 1993). The above simulation will be carried out using

STROBOSCOPE simulation software.

14

QI - Emp
N1 z Truc

C1 ‘ Trucl

 

 

 

 

 

 

 

 

 

 

 

 

 

N2
Q1 T
a
/ ._,
C1 N1
—> H—

 

 

 

 

 

 

 

 

Q1 - Empty Trucks in Queue Q2 — Dirt StockPile Q3 - Front-end loader idle
N1 — Truck Travel N2 — Dumping N3 - Empty trucks return
C1 — Truck Loading F1 - Function Counter

Figure 1.5 - Simulation Model
(Source: Sawhney and AbouRizk, 1993)

15

1.6

1.6.1

for the
and la
higher
expect
materia
the sys‘.

streamli

1.6.2 Tr
A 1
highly effi
housing. I
Stigma for
technologir
cost. This 1
could be a 17

Option IO affr

 

 

1.6 EXPECTED OUTCOME
1.6.1 Streamline Producfion Process

The main outcome of this research could be that it can provide guidelines
for the efficient use of the production plant. This can result in savings on material
and labor if the material flow process can be optimized. Tirnesaving and hence,
higher production would result in improvement in the production process. It is
expected that bottlenecks in the system will be found, especially labor and
material bottlenecks. These bottlenecks are expected to be the major barriers to
the system not being streamlined. Correcting these bottlenecks will lead into a

streamline production process.

1.6.2 Technological Improvement

A streamline manufactured housing process delivering at low cost and
highly efficient can offer a potential solution to the affordability problem in
housing. For many consumers, manufactured housing has had a negative
stigma for producing low quality housing. This research can produce a
technologically sophisticated plant to produce better quality housing at reduced
cost. This will lead to more acceptance of the industry in general by the public. It
could be a way of showing the public that manufactured housing is a viable

option to affordable housing.

16

1,6,3 ‘

an idea
specific
material
produce-

different

1.7 REF
This :
Chapter, 5
this resear
methodolo
related wor
Pdeuction
reader undr
description
cornDarison
PIOduction e
background '
01 dBVEIOpjn

Where the ire

 

 

1.6.3 Computer Simulation Software for Plant Layout

The simulation part of this research can help in the future development of
an ideal manufacturing plant layout. The simulation concepts could be tailored
specifically to the plant layout process. This could require inputs such as
materials used, equipment used, desired shape of plant, and type of housing
produced. The software can be interactive and allow the user to try many

different options to suits her needs before producing their ideal plant.

1.7 REPORT ORGANIZATION

This thesis report contains six chapters. Chapter one, the introductory
Chapter, presents an overview of the history of manufactured housing, need for
this research, scope of research, goals and objectives of the research,
methodology, and expected outcome. Chapter two contains literature review of
related work in areas of manufactured housing, construction simulation, and
production planning. In addition, some industry terminology is defined to help the
reader understand the rest of the thesis. Chapter three gives a detailed
description of the manufactured housing production process and presents a
comparison of manufactured housing and site built housing in the areas of code,
production environment, foundations, and cost. Chapter four contains
background work on data collection and presents the concepts and procedures
of developing the production process and material flow model. Chapter five is

where the transformation of the process model to simulation model takes place

17

and anal

report.

of the ma
research
rnethodo
gain frorr

of the chi

and analysis of input data and results. Chapters six presents the summary of the

report.

1.8 SUMMARY

This chapter presented the proposed research with overview of the history
of the manufactured housing industry and future trends, the need for the
research, scope of the research, goals and objectives, and research
methodology. A description of expected outcomes followed as to what could be
gain from this work by the industry in general. Finally, the organization of the rest

of the chapters was presented.

18

CHAPTER TWO
LITERATURE REVIEW

19

2.1 II

1
terrninOi
known 2
terms U!
of avails
inforrnat

simulatir

2.2 D

F r
manufact
definition
Associati
customer.

industry (1

2.1 INTRODUCTION

This chapter is organized into two categories. First, the industry
terminology is presented. The manufactured housing industry is not as well
known as the site-built housing and most people are not familiar with some of the
terms used in the industry. The second category on this chapter is the summary
of available research information in areas related to proposed research. This
information is presented in three categories; manufactured housing, construction

simulation, and production planning.

2.2 DEFINITIONS

For the reader who is not familiar with manufactured housing and
manufactured housing terminology, it may be helpful to introduce basic
definitions before writing any further. The Michigan Manufactured Housing
Association introduced some of the following definitions to help potential
customers to the industry, have a better understanding of the tenns used in the

industry (MMHA Fact Sheets, 2000).

2.2.1 Manufactured Home

A home built in a controlled, factory environment on a permanent chassis that is
designed to be used with or without a permanent foundation when connected to
the required utilities. Manufactured homes are built to the federal Manufactured
Home Construction Safety Standards enforced by the Department of Housing

and Urban Development (HUD) in Washington, DC. Manufactured homes are

20

2.2.:

Thee

engir
comp:
licensi
Michig

maybe

22.3 Ii
The fedé
making ti
corrply w

durability,

22.4 pan
Panelized |
Whole wall 1
WOW and .

"’99! the star

 

single story and are delivered to the home site in one, two, or occasionally, three
sections; they may be placed on private property or in a manufactured home

community.

2.2.2 Modular Home

These are Factory-built homes that begin as components and are designed,
engineered, and assembled in a controlled factory environment. These
components come together at the building site and the home is completed by a
licensed builder under standards enforced by state and local agencies. In
Michigan, modular homes are regulated by the BOCA Code. Modular homes

may be one-or-two-story dwellings and are placed only on private property.

2.2.3 Mobile Home

The federal government mandated the term “manufactured housing” in 1976,
making the term “mobile home” obsolete. All of the homes built after 1976 must
comply with federal standards. These standards regulate the things such as

durability, materials, systems, wind safety, fire safety and energy efficiency.

2.2.4 Panelized Home

Panelized homes are defined as homes where panels (flat units that represent a
whole wall with windows, doors, wiring and outside siding) are constructed in the
factory and then transported to the site and assembled. Panelized homes must

meet the state or local building codes of where they are sited.

21

2.2.5 P
Pre-Cl-lt
of home
and then

hOITBS 81

2.2.6 M2
Private Ia
most sites

referred tc

2.2.7 Sing
A manufac

average sq

2.2.8 Multi
A manufacti

I
Square feet.

219 Site.

HOUS’W Con

 

2.2.5 Pro-Cut Home

Pro-Cut homes are another type of factory-built housing. Materials for this type
of home are factory-cut (pre-cut) to design specifications, transported to the site
and then assembled. Pre-cut homes include kit, log and dome homes. These

homes are built to meet either local or state building requirement codes.

2.2.6 Manufactured Home Communities
Private land developed as homes sites for manufactured homes. In Michigan,
most sites are leased to the homeowner for a monthly fee. They are sometimes

referred to as a land-lease community.

2.2.7 Single-Section Home
A manufactured home delivered to the home site in one intact section; the

average square footage is 1,130 square feet.

2.2.8 Multi-Section Home
A manufactured home delivered to the home site in two or three sections. The
average square footage is 1,640 square feet, but may be as large as 2,400

square feet. It may have a (site-built) garage attached after the home is installed.

2.2.9 Site-Built Housing

Housing constructed mostly at the home site.

22

2.2.1
A refs
assISf.

prepal

2.2.1 ‘I I
Any per
homes it

home for

2.2.12 Dr's
Any perso

resale.

2.2.13 Mani

ANY person

InCIUdII’Ig an'

2.2. 141mm! 1
Any Person ii

IEpairS home.

 

2.2.10 Retailer
A retailer is a licensed, professional seller of manufactured homes. A retailer
assists in arrangement for financing, and has home installed on home site and

prepared for move-in.

2.2.1 1 Dealer
Any person engaged in the sale, leasing, or distribution of new manufactured
homes primarily to persons who in good faith purchase or lease a manufactured

home for purposes other that resale.

2.2.12 Distributor
Any person engaged in the sale and distribution of manufactured homes for

resale.

2.2.13 Manufacturer
Any person engaged in manufacturing or assembling manufactured homes,

including any person engaged in importing manufactured homes for resale.

2.2.14 Installer I Servicer

Any person including a mobile home dealer, who for compensation installs or

repairs homes.

23

2215C

Any Pf?rs

coach fol

2.2.16 HL
mesa:
Developm
strength a:
quality con
plumbing, z
preempts Ic

manufacturr

2.2.17 BOC,
Building Ofifi
three ‘model

Under this

United State

1118 Imam
Code prOVISip
OOI‘IStrum'On’ .

eqUiDment, U

 

2.2.1 5 Consumer
Any person who purchases a manufactured home, mobile home, or commercial

coach for her own use and who does not intend to resell it.

2.2.16 HUD Code

This is a code developed by the Department of Housing and Urban
Development. A HUD code regulates the home’s design and construction,
strength and durability, transportation, fire resistance, energy efficiency, and
quality control. It also sets stringent performance standards for the heating,
plumbing, air-conditioning, and electrical systems. The HUD Code specifically
preempts local building codes as they relate to construction codes for

manufactured homes.

2.2.17 BOCA Code

Building Officials and Code Administrators lntemational, Inc. It is the oldest of
three “model" codes used in the United States. Site-built and modular homes fall
under this code. It is used by states in the Midwestern and Eastern parts of the

United States.

2.2.18 lntemational Rosidontiat Code
Code provisions for one- and two-family dwellings which apply to the
construction, alteration, movement, enlargement, replacement, repair,

equipment, use and occupancy, location, removal and demolition of detached

24

one— 3
not W

30085

2.3

subto;
indust:
proces

not fou.

2.3.1 A
T.
literature
housing.
In '
of Michiga
quality, ma
manufactur
housing anc

Ownership a

one- and two-family dwellings and multiple single family dwellings (townhouses)
not more than three stories in height with a separate means of egress and their

accessory structures (International Residential Code).

2.3 EXISTING RESEARCH

The literature review will be divided into three different subtopics. The first
subtopic will deal with research that has been done in the manufactured housing
industry. The second subtopic will cover research on simulation of construction
processes. The third topic is about production planning. A single research was

not found that dealt with production aspects of manufactured housing.

2.3.1 Manufactured Housing Industry Research

The manufactured housing industry has several research projects and
literature, many of which argue that it is a viable option to provide affordable
housing.

In 1993, the College of Architecture and Urban Planning at the University
of Michigan, published a study in six reports addressing manufactured housing
quality, manufactured housing costs and finance, manufactured housing values,
manufactured housing impacts on adjacent property values, manufactured
housing and the senior population, and manufactured housing as an alternative
ownership and innovative uses (Warner and Johnson, 1993).

The first area the researchers tackled was the area of quality and how

manufactured housing compares to site-built housing. “... The researchers

25

cautr'o

concluc
that mar
consume
Tt

answer w
concludec
1987 and ‘
can be shc
residential .
in their neig
The L
manutacture
researchers

bEen Shown

 

residential p:
The fi'

”Beds 0f Sen.

 

concluded that manufactured housing has demonstrated reasonable
performance in areas related to structural durability, maintenance, wind safety,
fire safety and thermal efficiency” (Warner and Johnson, 1993). They also
cautioned that there is room for improvement.

The second issue the team looked at is cost and finance. The research
concluded that the cost is not only comparable to that of site-built houses, but
that manufactured housing is the most affordable option available to most
consumers.

The third focus was on housing values. The question they wanted to
answer was, whether manufactured houses appreciate. “The researchers
concluded that a sizable proportion of the manufactured homes sold between
1987 and 1990 in Michigan, most of which were located in rental communities,
can be shown to have appreciated” (Warner and Johnson, 1993). Most site-built
residential communities do not want to allow manufactured housing communities
in their neighborhood for fear of loss of property values.

The University of Michigan’s forth study was about the impact
manufactured housing communities have on adjacent property values. “The
researchers concluded that rental manufactured home communities, have not
been shown to have a significant effect, positive or negative, on adjacent
‘ residential property values” (Warner and Johnson, 1993).

The fifth issue examined is how manufactured housing can meet the

needs of senior citizens. The study found potential for the manufactured housing

26

comuni
realized

Fi
residenti
promisin

A
program
State Ur
Foundat
housing
the Mich
divided i
industry’
industry,
based or
the indu:
installer:
Officials
for the p'
plasente
reQuite-m

0VErall k

TECOm’he

community within the senior population and this potential are being slowly
realized but cautioned on the rising cost of rent.

Finally, the study closed with an overview of how to broaden consumers’
residential choices. The study suggested that manufactured housing has
promising use in areas of affordability and redevelopment.

A multidisciplinary faculty group from the construction management
program and the department of human environment and design at Michigan
State University, undertook a research project in 1995 to help “Harvest
Foundation” expand the manufactured housing industry’s position in Michigan’s
housing market. The Harvest Foundation is the educational outreach branch of
the Michigan Manufactured Housing Association (MMHA). The project was
divided into two phases. Phase one’s goal was to do an assessment of the
industry's current position with the help of key groups and their perception of the
industry. Phase two was to develop a ten-year educational plan for the industry
based on the findings from phase one. In the first phase, nine groups related to
the industry were identified: manufacturers, retailers, community managers,
installers, inspectors, site/modular builders, consumers, and local government
officials. Geographically diverse focus groups were utilized to gather infomiation
for the project. Findings and recommendations from each focus group were then
presented. Installers for example would like to see the state institute standard
requirements for the installation and site preparation of manufactured homes.
Overall key issues where then identified by the researchers and

recommendations made. For example in the case of sales persons, the

27

researChers

 

behavior af‘
Therefore. -
industry in
Baya
HUD codes
house from '
code in mos
were cases |
A doc

changing ma

 

salesperson
Ways to impr
Nader (1934
homes has ll
Developmer
advances he
housing N
attramed to

provide MN.
8 better mar

Paper for th.

researchers found that since salespeople interact with the general public, their
behavior affects customer perceptions positively or negatively about the industry.
Therefore, sales people must be more knowledgeable about the product and the
industry in order to present if correctly (Burkhardt et al, 1996).

Bayazit (1997) did an extensive and detailed comparison of the BOCA and
HUD codes and analysis from planning considerations to transportation of the
house from factory to the site. The HUD code was up to standard to the BOCA
code in most aspects and even more stringent in other aspects although there
were cases in which it was below standard to the BOCA code.

A doctoral dissertation by Nader Taghavi (1984), investigated the
changing market of manufactured housing by seeking the opinions of
salespersons and consumers in the industry. The study was also looking for
ways to improve the market segment for manufactured housing in the future.
Nader (1984) concluded that salespersons believe the quality of manufactured
homes has improved since 1976, when the department of Housing and Urban
Development were initially involved in the industry. He also concluded that
advances have been made in easing zoning restrictions for manufactured
housing. Nader’s (1984) results also showed that younger people are less
attracted to manufactured housing communities because they believe it does not
provide privacy. He suggested that subdivisions and private lands would provide
a better market for the industry and the possibility of new kinds of buyers.

Diane R. Suchman formerly of the Urban Land Institute (ULI) wrote a

paper for the ULI in March 1995 on manufactured housing as an affordable

28

altemath
as 35990
segments
opportun
attractive
also encr
housing
manufac
afiordabl
Fi
William i

housing .

2.3.2 Cl

At

Process 1

alternative to housing problems. She identified cost savings among other things,
as aspects that make manufactured housing attractive in different market
segments. “The cost savings, design flexibility, and expanding financing
opportunities possible with manufactured housing today makes this housing type
attractive in many markets and for many specific uses” (Suchman, 1995). She
also encouraged public officials to be more proactive in creating more affordable
housing options using this industry by stating that“... by encouraging wise use of
manufactured homes, public officials can help create more opportunities for
affordable homeownership within their communities” (Suchman, 1995).

Finally, “Factory Constructed Housing Developments”, is a book by
William F. Albem (1997), which looks at the planning, design, and construction of

housing developments and housing units in the factory.

2.3.2 Construction Simulation Research

Abourizk and Mather (2000) discussed an approach to simplify the
process of building and manipulation of computer simulation models of heavy
earthrnoving construction operations. Their approach is based on integrating 3D
CAD with simulation. “The simulation models are automatically generated from
high-level descriptions in CAD by using add-on tools to capture key information"
(Abourizk and Mather, 2000). Infomiation to be captured includes product
information, resources used, construction methods, and general product data.
Using a case study, they developed a prototype earthrnoving simulation system

by extracting the information they need form 30 CAD. Their conclusions were

29

that the 5
efficient.
on hierar

Tomn
construct
object—0r
models c
compone
modeling
using Cll

represen

that the simulation system used was too general and the integration was not very
efficient. They suggest improvements by developing a simulation system based
on hierarchy.

Tomrnelein, Carr, and Odeh (1994), presented a way of modeling
construction processes, using designs, plans and construction methods. “An
object-oriented and interactive cormuter system is presented that realistically
models construction processes by matching resource properties with design
component properties and operation durations” (T ornrnelein et al, 1994). The
modeling procedure was done by creating a discrete-event simulation model
using ClPROS, which includes a seven consecutive step knowledge-based
representation and simulation primitives. The seven steps are:

1. Define project design and specifications
Create activity-level plan and relate activities to the design components
Choose construction method for each activity in the plan
Initialize product components
Identify construction resources

Custorn-tailor simulation network

NP’S-"PP’N

Simulate and interpret results

Their conclusions were that “ClPROS can realistically model construction
processes by integrating data from design drawings and specifications, and
construction plans, in combination with a library of construction methods"

(T ommelein et al, 1994).

30

 

 

A.

Sawhney and AbouRizk (1993) gave a very easy to understand overview
of construction simulation and its applications. In their paper, they gave a brief
history of construction simulation, described the tools that have been developed
over the years and their application, and outline the procedure for doing
construction simulation. The paper had a target audience of students in the early
stages of Ieaming construction simulation. It discussed the CYCLONE software;
its early and recent developments, other software like INSIGHT, RESQUE,
COOPS, MicroCYCLONE, SLAMSYSTEM, SIMSCRIPT, GPSS, and SLAM II.
The authors closed by arguing for construction simulation as a great tool in the
areas of construction planning and control and analysis of construction

operations.

2.3.3 Production Planning Research

Effective production planning can give a company a strategic advantage
over its competitors in a very competitive market. A production process that can
be simulated to help in the allocation of company resources is a vital tool in the
production planning, project controls or procurement stages of the process.
Production planning is a very important part of any production process.

Warszawski’s (1996) paper, presents a methodological procedure for
strategic planning in a construction company. The procedure consists of five
stages. Stage number one examines the company’s mission. “The mission of a
company, as defined here, involves the scope of its business activities,

objectives, and special codes of behavior” (Warszawski, 1996). The second

31

stage is for the company to analyze its main resources. Warszawski divided
main resources into construction, marketing and procurement resources.
Construction resources for example must have the capability to support the
physical construction of a project from start to finish. ‘Construction, as viewed
here, refers to the capability of a company to perform a variety of direct activities
for the physical realization of construction projects, i.e., construction planning,
coordination, monitoring, directing, and control” (Warszawski, 1996). The third
stage is the development of a competitive strategy. “The development of a
strategy is based on “mapping” of the relative attractiveness of the various
possible activity areas” (Warszawski, 1996). The fourth and fifth stages consist
of developing alternative strategies and selecting the preferred one.

Shaked and Warszawski (1995) discussed a knowledge-based system
called HISCHED for construction planning of high-rise buildings. “The system
uses an object-oriented representation of the building and production rules,
routines, and functions to manipulate objects and to generate the construction
plan” (Shaked and Warszawski, 1995). The paper first described the
organization of the knowledge base.

“ The knowledge about the system involves an object-oriented
representation of the building and production rules, routines, and
functions for the manipulation of object attributes and generation of
the construction plan. The representation of the building includes
three types of objects: ( 1) Zones, which define the topology and
nature of the building; (2) functional systems, which define the
components of the building and their construction technology, and

(3) works, which define the construction activity" (Shaked and
Warszawski, 1995).

32

 

The P3
genera
Genert

autorne

The ne
plannir
HlSCl‘
be fart
graphi-
view 0'

Warsz

li‘anag
Ploduc
”Ethoi
pr Ogre:

30(lulsi

The paper then went on to explain how project construction activities are
generated and the determination of the dependencies between them.
Generation of the tasks can be either interactive with the user or done
automatically and it consists of four stages:
(1)Zoning — user defines all the floors of the building, their
sequence, area and designation, and assigns the repetitive floors
into multimodules. (2) Identification of functional systems —
Functional system for each zone can be selected by the user, or
can be automatically defined by HISCHED on the basis of the
designation of the zone (for such typical designations as office
floor, lobby, mechanical floor, and others). (3) Definition of Works —
Each functional system is defined in HISCHED by works necessary
for its completion (their nature and quantity). (4) Generation of
Tasks — Tasks are generated by the assignment of works to zones.
(Shaked and Warszawski, 1995).
The next part of the paper is to show how different managerial objectives affect
planning considerations. Finally the paper closed with an example of the
HISCHED application. The authors concluded by stating that“ the system can
be farther extended to consider additional managerial strategies to integrate
graphical representation of the building and to reason about selection of works in
view of different economic and non economic considerations” (Shaked and
Warszawski, 1995).

In 1999, Andrew Seidel submitted a masters thesis to the construction
management program at Michigan State University in which he developed a
production planning methodology for large homebuilding firms. “The
methodology includes planning systems; estimating, scheduling, cost and

progress control, financial and accounting, subcontracting, and material

acquisition” (Seidel, 1999). After developing the methodology, Seidel (1999) did

33

an industry analysis to evaluate the applicability and essentiality of the
methodology. He concludes that the production planning methodology provides
a detailed sequential process for production planning system development for
large homebuilding firms.

Robertson and Ulrich (1998) argued that a good product planning platform
was necessary to gain and keep market share. In their paper, they define what a
platform is as it relates to production planning, benefits and challenges of
platform planning, presented a method for planning a new platform of products,
and provide recommendations for managing the platform planning process.
They define platform as the collection of assets a set of products share such as
the manufacturing processes, design knowledge, parts and components, and the
manufacturers (Robertson and Ulrich, 1998). Benefits to platform planning
include reduced development cost, time, and lower investment required for new
products. The authors advocated the development of a product plan,
differentiation plan, and commonality plan as a structured platform planning
process. Product plan included the specification of distinct market offerings over
time. Differentiation plan encompasses specifying each product’s differentiating
attribute. Finally, commonality plan describes the extent to which the products
share physical elements. Recommendation for managing platform planning
process include getting top management involved since the products developed
from the process may determine what the company introduces to the market for

five to ten years. One person should be in charge of each of the plans (product,

differentiation, and commonality) and a different person overseeing the whole

process.

2.4 SUMMARY

This chapter presented the reader with important terminology needed to
have a basic understanding of the manufactured housing industry and research
that has been done in three different areas pertaining to the thesis topic were
also discussed.

More acceptance of manufactured homes into the mainstream housing
industry is required since it can provide affordability for low income families as
argued by Suchman (1995). The study performed at the University of Michigan
in 1993 demonstrated that manufactured housing is of good quality when
compared to site-built housing at comparable prices. The study showed
manufactured housing appreciates in value in contrast to the conventional
wisdom, which says manufactured housing depreciates. The study performed at
Michigan State University in partnership with the Harvest Foundation of MMHA,
gave insightful recommendations to help the industry advance in the state of
Michigan coupled with an analysis of the state of the manufactured housing
industry in the state of Michigan.

Three papers on construction simulation research were discussed. The
paper by AbouRizk and Mather (2000) attempted to simplify the process of
building simulation models. Their approach was to automatically generate

simulation models from 3D CAD drawings by integrating them. Although there

35

demonstr:
The 5800'
new way ‘
proposing
introduce
(l993) ga

Nu
and menu
strategic r
company
Shaked (1
utilizing a
homebuilc
discussed
Ulnch‘s (1
design. (it

and Cost.

demonstration was not successful, the approach showed some promising signs.
The second paper was very similar in a way since it also wanted to develop a
new way of developing simulation models except in this case, the authors were
proposing using designs plans and construction methods. For anyone being
introduced into the field of construction simulation, Sawhney and AbouRizk
(1993) gave a through introduction to the subject

Numerous research has been performed both in the field of construction
and manufacturing. Warszawski (1996) presented a five-stage procedure for
strategic planning in construction companies. A major part of his analysis of
company resources stage deals with construction resources. Warszawski and
Shaked (1995) looked at a knowledge—based approach to construction planning
utilizing a system called HISCHED. A production planning methodology for large
homebuilding firms developed by Seidel (1999) as part of a masters thesis was
discussed. Finally, in the area of production planning in general, Robertson and
Ulrich's (1998) paper on platform planning, discusses how consolidating product,
design, development, and manufacturing can help lower risk and reduce time

and cost.

36

CHAPTER THREE
MANUFACTURED HOUSING PRODUCTION PROCESS

37

3.1 INTRODUCTION

The construction process of manufactured housing is both similar and
different to that of site-built housing in various ways. Some of the similarities
they share are in materials used and differences occur in areas such as the
environment in which the house is built. This chapter explains the process of
building a HUD code manufactured home using the practices of two companies
in illustrations. An attempt is also made to compare and contrast the process
with that of a site-built home in areas of codes, production environment, and

foundations. Finally, a cost comparison is presented.

3.2 PRODUCTION PROCESS

The names of the two companies utilized in the research process will not
be mentioned for competitive reasons. Their practices will be illustrated in this
Chapter and the entire report, but it does not mean that the way they do things is
a complete representation of the industry’s practices. The description will be
made as generic as possible. The activities and stations in Figure 1.1,
represents the typical manufactured plant as the title implies. It will be used to
walk through the description of the process.

A manufactured housing plant can generally be divided into five areas.
These are floors, walls, roof, exterior, and finishes. On Figure 1.1, floors consist
of stations 1 and 2, walls consist of stations 3 and 4, roofing is station 5, Exteriors

is station 6 and 7, and finishes consist of stations 8, 9, and 10. A plant can have

38

up to 16 stations depending on the complexity of the homes they build, the

number of activities in various stations or the size of the plant.

3.2.1 Floors

Manufactured houses are built on a steel base frame called a chassis.
The process starts with the floor frame constructed on top of the chassis. The
floor joists are put in place with the right spacing and according to plan. After the
floor frame is finished, all the mechanical work that needs to go underneath the
house is installed at this point. Mainly HVAC and water lines are installed during
this stage. The floor is then properly insulated. All the above activities take
place in station 1. Figure 3.1 shows a floor being built with insulation and floor

joists already installed.

 

Figure 3.1 — Floor System Being Built

39

Then it is moved to station two for floor decking. The movement of the units itself
can take place in many different forms depending on the plant. One method
observed in one of the plants is the use of inflatable base plates that make it
possible to move the house with only four or five workers. On station 2, the floor
joists, attached to the chassis with mechanical lines running through it and
insulated, is covered by adding decking. Up to this point, the house is separated
in two units as shown in Figure 1.1. All the activities described above are
performed on both units in parallel. Next, the process moves on to the wall

stations.

3.2.2 Walls
The walls are generally built at a feeder station and then hoisted in place
using an overhead crane that run on tracks attached to the ceiling of the plant

building. Figure 3.2 shows a picture of a wall being built at a feeder station.

40

mum-vi. IIIII In? """ E

"‘5

9;." IT ‘l‘i'.

 

Figure 3.2 — Wall Feeder Station

Although it is not shown on the diagram in Figure 1.1, certain activities from
feeder stations have to be performed before the walls are in place. These
activities are installation of kitchen cabinets, some bathroom fixtures such as
tubs, and installation of water heater and furnace. The items are installed
because of the ease without the walls in place, and then the partition/interior
walls are placed. After the partition walls are placed, the house moves to station
four for the marriage and exterior walls. All the walls are completely built with
insulation at the feeder stations. At station four rough electrical and mechanical

works are performed.

41

 

In

3.2.3 Roofing

Roof assembly usually takes more than one station. First, the feeder
stations for roof stations will be described. Fabrication of the individual roof
trusses is generally contracted out to independent contractors. They are then
stored in the plant and used as needed just like any other material to assemble
the whole roof and ceiling structure.

First, the sheets of gypsum ceiling boards are laid on a flat surface, and
then the individual trusses are placed at the correct spacing. From there, the
trusses are nailed together with 2 X 4’5 between them. A special type of glue is
sprayed to the trusses at its point of contact with the ceiling board to hold them to
each other. Insulation is placed on the down side of the roof slope and then it is
ready to be moved to the adjacent feeder station for paint to be sprayed on the
ceiling. Before the paint is sprayed, the joint between the boards are covered
with tape and mud for a smooth connection. After the paint is applied, the roof is
ready to be hoisted up for connection to the house. This process just described
happens in a feeder station. When the whole roof and ceiling structure is ready,
it is sent to the main line for assembly. Figure 4.3 shows very good pictures of a
roof structure at various stages of its assemble process. The picture on the left
shows the trusses being placed on the gypsum board at the correct spacing.
The picture on the right shows a completed roof still at the feeder station with
insulation being sprayed on top. The bottom picture illustrates the installation of

shingles after the roof has been secured on top the house.

42

 

Figure 3.3 - Shingles Installation

The first activity that takes place when the roof is moved to the line is it
has to be set on top of the house in which it belongs to. The structure is properly
secured to the two units, which have been joined together at this point. The two
units will be separated again when it gets to station 9 as shown in Figure 1.1. In
fact, the main reason why the two units are attached at this point is to make sure
the roof structure properly fits on the house. The activity that follows is roof
insulation. This activity is accomplished by having one worker standing on top of

the house and using a powered hose to spray the insulation between the trusses
43

and on top of the ceiling. This activity can also happen when the roof structure is
still at the feeder station. After insulation is sufficiently packed on top of the
structure, the roof has to be covered with decking. This activity usually takes a
crew of three workers about ten to fifteen minutes to finish one unit. It consists of
laying down sheets of plywood on the roof and nailing them to the trusses. The
lower edge of the roof is then covered with paper for proper drainage measures
before the shingles are placed. Placement of shingles is shown in the bottom

picture of Figure 3.3.

3.2.4 Exteriors

Figure 1.1 shows stations five and six running in sequence but essentially,
these two stations can run in parallel. As work is progressing with the roof on
top, workers are simultaneously installing doors and windows, exterior boards,
siding, and so on.

Exteriors and exterior finishes start after the walls are installed. The first
activity is the exterior walls have to be covered with boards. Doors and window
openings are then carefully cutout. Then the jams are installed before installing
the exterior doors and windows. The last activity is siding. Figure 3.4 shows the
crew installing exteriors and windows after the exterior boards have been placed.

These activities can be done while the units are either attached or separate.

 

Figure 3.4 — Installation of Exterior Doors and Windows

3.2.5 Finishing

Electrical, mechanical, and HVAC work are the main activities in station 7.
Certain appliances such as the fireplace are installed at this station. The breaker
box is also installed at this stage. A significant amount of testing and inspection
is also performed at this stage. Some of them include, gas test, plumbing tests,
check for proper receptacle spacing, check for interior switches and their
installation to manufactures specifications and so on.

At station 8, the extent of the interior finishes depends on whether the

house needs painting or whether it is going to have wallpaper. First, the axles

45

and the wheels are installed for transportation. Installation of axles and tires can
happen on station one depending on how the houses are moved at a particular
plant. Assuming the house has wallpaper instead of painting, interior finishing
can go into full gear at station 9. Window curtains are installed; appliances such
as dishwasher, toilets, washing machine, and dryer are fitted into place. In
addition, any repair work that needs to be done also takes place at this station. If
the house needs paint, then a whole day is spent on drywall finish. This is not
because of the activity taking too long, but they have to wait overnight for the

walls to dry up.

3.3 COMPARISION OF MANUFACTURED VS SITE BUILT HOUSING
Manufactured and Site-built housing have their similarities and
differences. One major difference is the absences of a regular foundation in the
manufactured houses. This section compares and contrasts the differences
between manufactured housing and site-built housing in the areas of codes,

environments in which they are built, foundations, and cost.

3.3.1 Code Comparison

The manufactured housing industry uses the HUD (US Department of
Housing and Urban Development) code, which was developed by the housing
and urban development branch of the United States government. The code was
developed in 1976 mainly to raise the standards of the manufactured housing

industry and increase safety. The housing industry professionals have been

46

debating whether the HUD code is up to the same standards as that of the major
site-built housing codes such as BOCA (Building Officials Conference of
America), IRC (lntemational Residential Code), and U30 (Uniform building
code). Two studies will be cited that have been performed in the past to
compare the HUD code and BOCA code. Another study by Mehrotra, Syal, and
Senghore (2001) will be cited for comparison of the HUD code with the
lntemational Residential Code. The BOCA code is the most widely used
residential building code in the United States. The first study on the BOCA and
HUD codes comparison was done at the University of Michigan ‘8 College of
Architecture and Urban Planning and the second one is a Plan B paper written by
Hasan Bayazit for a Master's degree at Michigan State University’s department
of construction management.

The Bayazit (1997) study, which is more detailed, looked at all the
subparts (Planning considerations, fire safety, body and frame construction
requirements, testing, thermal protection, plumbing systems, heating, cooling and
fuel burning systems, electrical systems, transportation and systems) of the HUD
code and compared them to the corresponding values on the BOCA code. Some
of the standards were higher, some lower and some the same. For example, the
planning considerations for HUD and BOCA codes were summarized as shown

in Table 3.1.

47

Table 3.1 - Summary of Planning Considerations for HUD and BOCA Codes

(Source: Bayazit - 1997)

 

 

 

PLANNING HUD Code/1994 BOCA Codol1993
CONSIDERATIONS
Light and Ventilation
8% of floor area 8% of floor area
Glazed Area
4% of floor area 4% of floor area
(Minimum)
Required in bathrooms, Required in all rooms
Open able Area (Min) kitchen, laundry and
utility rooms, and toilet
Artificial Light

Mechanical Ventilation
Ventilation in

Bathroom
and Kitchen

Room Dimensions
(Minimum)

Living Area

Bedrooms

Kitchen

Minimum Room
Dimension In any
Direction

Hallways Width (Min)

 

Required
1.41 m3 [minute

exhausting or 0.14 m2
operable area for

bathroom and 2.82 m3
[minute for kitchen

13.9 m2

4.6 m2 a 6.5 m2 for 2
persons
Not available

1.52 m

71.1 cm

2.13m

 

Required

0.14 m2 operable area or
air change in every 12
minutes for bathroom,
toilet compartment, and
similar areas

13.9 m2

4.6 m

4.6 m2 for kitchen

2.13 m except kitchen

91.4 cm

2.28 m

 

48

 

 

a

Exit

lnteri
Toile

same.
one-p.
height
distanl
the cor
subpar

lhe BO:

 

 

Ceiling Heights (Min)
Habitable Rooms
Hallways

Exit Facilities
Egress Doors
Travel Distance

Exit Door Size

Egress Window

Interior Privacy and
Toilet Compartment

Privacy Lock

Minimum Size

 

1.98 m

2 required
10.6m

71.1cm by 188cm for
swinging doors and
71.1cm by 183 cm for
sliding doors

Required in all
bedrooms unless there
is an exit door. 91.4
cm from the floor

Required

76.2 cm in width and
53.3 cm clearance in
front of each toilet

 

2.13m

1 required
22.8 m

91.4 cm by 203 cm

Required in every
bedroom unless there is
an exit door. 111 cm
from the floor with a net

opening of 0.52 m2.

Privacy Required

76.2 cm in width and 53.3
cm clearance in front of
each toilet from the tub
and the wall

 

The minimum glazed and open able areas for light and ventilation are the

same, the minimum room dimensions for living area are the same and that of

one-person bedrooms are the same. There is little difference in the ceiling

heights for habitable rooms and hallways among other things. Minimum travel

distance to exit facilities is shorter in the HUD code, which shows the attention

the code shows when it comes to fire safety. Similar comparisons of all the

subparts are made and it consistently shows the HUD with similar standards to

the BOCA code and even better in some cases.

49

 

 

tied
Tab

iii?
con:

Flani
Rafin
Iteni
waHs.
pant.

Celr
imsh

Vials
endo
range

Kflcm
doors
backs
eXpos
endgn
bathtt
‘ ShoWE

\

The study performed at the University of Michigan was also very similar.
The difference is, one table of summary was produced showing major subparts
but not all. Thirty-two (32) elements were looked at. Of all the elements, sixty-
three percent (63%) had the same requirements. Twenty-two percent (22%) of
the elements looked at were more restrictive in the HUD code and thirteen (13%)
of the elements were more restrictive in the BOCA code.

A paper on comparison of the HUD and the international building code is
cited for Table 3.2.

Table 3.2 - Summary of Fire Safety Considerations for HUD and lntemational
Residential Code. (Source: Mehrotra and Syal, 2001)

 

 

 

FIRE SAFETY HUD CODE/1994 INTERNATION
CONSIDERATIONS RESIDENTIALCODE I 2000
Flame Spread

Rating Requirement
Interior finishes of Should not exceed 200. Should not exceed 200.

walls, columns, and [HUD 3280.203] [IRC R3191]

partitions

Ceiling interior Should not exceed 75. Should not exceed 200.
finishes [HUD 3280.203] [IRC R3191]

Walls adjacent or Should not exceed 25. Should not exceed 200.
enclosing cooking [HUD 3280.203] [IRC R3191]

range

Kitchen cabinets, Should not exceed 200. Should not exceed 200.
doors, countertops, [HUD 3280.203] [IRC R3191]
backsplashes,

exposed bottom and
end panels, plastic
bathtubs, and
shower units

 

 

 

 

50

 

’riRi
cor

Can

Fire

Foa
lnsr

Flar

Ma)
Fire

Sml
Shc

par
COU
sho
200
men

Wall:

Table 3.2 —- cont’d

 

 

Foam plastic thermal
Insulating material

Flame spread

Maximum smoke index
Fire detection equipment

Smoke detectors/alarms
Should be placed

 

 

prevent spreading or fire.
[HUD 3280.206]

Not exceeding 75.
[HUD 3280.207]
450 [HUD 3280.207]

One in each bedroom
[HUD 3280.304]

 

FIRE SAFETY HUD CODE/1994 IRCI2000
CONSIDERATIONS
Carpeting Should not be used in (Not Specified
Compartments for furnace
and water heaters.
[HUD 3280.205]
Fire stopping At least 1” lumber or 5l16” At least 2” lumber or 72”
thick gypsum board should thick gypsum board
be provided between should be provided
concealed spaces to between concealed

spaces to prevent
spreading of fire.
[IRC R6028]

Not exceeding 75

450 [IRC R3191]

One in each bedroom
area, also outside, in its
vicinity. [IRC R3171]

 

The flame spread rating requirements for interior finishes of walls, columns, and

partitions are the same in both the HUD and IRC code. Kitchen cabinets, doors,

countertops backsplashes, exposed bottom and end panels, plastic bathtub, and

shower units, all have the same flame-spreading requirement of not more than

51

200 in both the HUD and IRC code. The flame spread rating requirement is
more stringent in the HUD code when it comes to ceiling interior finishes and

W3. '8 adjacent or enclosing cooking ranges. The ratings are 75 and 25

 

[9509C
Dew

iorfiar

3112

ranul
aebL
isone
modu
gerro
[Sin
mean:
Super

worke

1313

manu
Insal
Under

bGIOW

respectively compared to 200 for both in the IRC. Requirements for firestopping
between concealed spaces are more stringent in the IRC code. The requirement

for flame spread and smoke index are the same in both codes.

3.3.2 Production Environment

With the exception of the installation process on foundations,
manufactured housing are strictly factory built houses. This means the houses
are built in a controlled environment with no adverse effects from whether. That
is one major advantage it has over site-built housing. A manufactured housing
production schedule is usually not delayed because of weather if the workers can
get to the factory. A second advantage it has over site-built housing is that, since
it is in a very closed and controlled environment, supervision is very easy. This
means, a highly skilled labor force is not necessary for production of the houses.
Superintendents, forernen, and production managers can hire fairly skilled

workers and still obtain desired production with close supervision.

3.3.3 Foundation

Foundations are also fundamentally different between site built and
manufactured houses. Here are two definitions from the Manufactured Home
Installer/servicer Course offered by the MMHA that can help the reader
understand the process of installing a manufactured home foundation discussed

below.

52

 

dug

are

con
incr
wig
The

afore

FOUNDATION FOOTING: Part of the support system located at or below ground
level. Piers are placed on foundation footings, which are made from concrete or
treated lumber.

PIER: The portion of the support system between the foundation footing and the
manufactured home, exclusive of caps, plates and shims.

In a typical site built home with a basement, the footing is continuously
dug around the whole perimeter of the house. Then basement/foundation walls
are placed on the footing, and the house sets on the foundation walls. In a
manufactured housing foundation, the footings are not continuous. Instead of one
continuous footing, there are several small ones with an area of 144 square
inches. The footings are placed on site and spaced based on dimensions and
weight of home, the roof load zone, size and type of construction, of the footings.
The figure below shows how the footings could be spaced based on the

aforementioned criteria.

 

O

O O
O O
O O

O
O

 

 

 

Figure 3.5 — Top View of Footing Spacing.

53

 

0V6

COIT

The footings, just like in site built houses, must be extended below the frost line
or protected from frost effects. Piers are then placed on top of the footings, on
which the beams under the house are placed. Below is a diagram of a single

block pier sitting on a foundation footing.

--
u.
in
..-'

 

 

.
.
.
‘-

 

.u"
,n—"'

Figure 3.6 — Single Block Pier
(Source: MH Installer Course, 2000)

It is sometimes required to put piers around the perimeter of the house.

Otherwise, it may lead to sagging floors, walls and roofs.

3.3.4 Cost Comparisons

In 1998, the US. Department of Housing and Urban Development
published an extensive report titled “Factory and Site-Built Housing; A
Comparison for the 21" Century”. The comparisons include the regulatory
process, code requirements, and cost. Table 3.3 shows the different contributors
to the total cost that were analyzed such as construction costs, land costs,
overhead/administration costs and financing costs. The original table included

comparisons with modular housing but since this thesis is focusing on

54

IEIEEIQ‘IE’IS’ S’l

‘_
C.)
O

li’o’lil.

L
J

Th.

ho:

CO

Sm

manufactured housing, it is vital that the scope of the cost comparisons is also

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

focused.
Table 3.3 - Summary of Cost Comparisons
(Source: HUD Report on Cost — 1998)
Site-Built Manufactured Homes

Description: Two—Story Double-Section Double-Section Double-Section Single—Section

Square Feet: 1,990 1680 1680 1680 1215

Foundation Blocks Permanent Permanent Blocks

Land lease
Private Private Private Community
Individual Lot Subdivision
Construction
Costs 3 76,752.00 5 36,150.00 5 37,650.00 5 37,650.00 3 21,650.00
Structure $ 70,765.00 $ 34,650.00 $ 34,650.00 $ 34,650.00 $ 20,850.00
Foundation 3 5,987.00 $ 1,500.00 $ 3,000.00 $ 3,000.00 $ 800.00
Costper squareft $ 38.57 $ 21.52 $ 22.41 $ 22.41 $ 17.82
Land Costs 5 35,136.00 5 34,425.00 3 34,881.00 5 1,167.00 5 711.00
Lot density 4 per acre 2 per acre 46 per acre 4-6 per acre 68 per acre
lm roved lot $ 33,941.00 $ 33,714.00 $ 33,714.00
Site preparation $ 1,195.00 $ 711.00 $ 1,167.00 $ 1,167.00 $ 711.00
Monthly land rent $ 250.00 $ 200.00
OverheadlAdmin 5 29,232.00 5 11,448.00 5 20,179.00 5 12,088.00 S 7,035.00
Overhead & Ge.
[exp $ 8,325.00 3 1,908.00 $ 3,363.00 $ 2,015.00 $ 1,172.00
Marketrlg $ 3,024.00 $ 954.00 $ 1,682.00 $ 1,007.00 $ 586.00
Sales commission $ 4,752.00 $ 1,431.00 $ 2,522.00 3 1,511.00 $ 879.00
Profit $ 13,104.00 $ 7,155.00 $ 12,612.00 $ 7,555.00 $ 4,397.00
FInancirLg Costs 5 2,880.00 5 477.00 5 841.00 5 504.00 3 293.00
Construction
finanfig $ 2,880.00
Inventory financing $ 477.00 $ 841.00 $ 504.00 $ 293.00
OTAL SALES

PRICE $144,000.00 5 82,500.00 5 93,551.00 5 51,409.00 5 29,689.00

 

 

 

 

 

 

 

The table starts with a comparison of typical homes with different traits. Site-built

homes are compared to manufactured homes.

CONSTRUCTION COST: The authors included labor and material for both the

structure and foundations as part of construction cost. Included as part of the

structural cost are additional construction of site finishes such as siding and
55

roofing completion, interior finish of marriage walls and flooring, and site-built

garages (HUD 1998).

LAND COST: Land cost consists of developed land costs, such as, cost of

subdivision infrastructure (roads, sidewalks, and utilities) (HUD 1998).

OVERHEAD [ADMINISTRATION COST: These items, which also affect the
cost of a home, include office space, office cost, administrative and management
staff, and company vehicles. Overhead cost as a percentage of sale price is
lower for manufactured housing (25%) than it is for site-built housing (22.3%)

(HUD 1998).

FINANCING COST: The length of time it takes to build site-built homes is
usually three months or more. On the other hand manufactured homes usually
takes about five days to build after an order is received from a retailer. The
financing cost of the inventory is minimal for the producer, and much of the cost
of inventory is shifted to the retailer. This allows for cheaper financing cost for
manufactured housing over site-built housing (HUD 1998).

All the four different categories of cost looked at, the report consistently
shows manufactured housing as much cheaper compared to site built housing
even if the manufactured house is on a permanent foundation, private lot, or

located on a subdivision.

56

3.4 SUMMARY

This chapter dealt with two main themes. The first is the production
process of manufactured housing. An effort was made to educate the reader
who has never been in such a factory to have a basic understanding of how
these houses come together. Pictures were used in the illustrations instead of
diagrams to help the reader visualize the process. In case the reader is familiar
with site built housing, a comparison was made with manufactured housing to
present the similarities and differences and further help the reader understand
the manufactured housing process. Cost comparisons were made to re—enforce
the point made in the introductory chapter that manufactured housing is a viable

option for affordable housing.

57

CHAPTER FOUR

PRODUCTION PROCESS AND MATERIAL FLOW MODELS

58

4.1 INTRODUCTION

This research project consisted of a great deal of legwork. The data collection
process to the Northern Indiana area consisted of approximately five thousand
miles of traveling including one trip to Cincinnati, Ohio. Data collection was
conducted through plant visits with other research team members. This chapter
deals with two focus areas. The first one is the data collection process and the
second is the development of the production process and material flow model. In
the data collection part, first, the less rigorous part of process mapping and how
that was used to develop the process flow diagrams and eventually the
production process and material flow model, is presented. Then, the data
collection part (i.e) time data collection is discussed. Finally, the product of the

data, which is the production process and material flow models, are presented.

4.2 DATA COLLECTION

The data collection part of the project was done in two phases for two
factories in Northern Indiana. The first phase was a familiarization phase and the
second was the actual data collection. Since time is the main data collected
throughout the process, a portion of this chapter is dedicated to the concept of
time study and what it means to this project, but first the author would like to

discuss the concept of process mapping and how is was used in this project.

59

4.2.1 Process Mapping
Process mapping is a concept used in developing the process model. “The

fundamental concepts of process mapping are based on the idea of structured

analysis, which has produced significant payoffs in diverse business enterprise

applications such as banking, insurance, manufacturing (auto and aerospace

industries), pharmaceuticals, and service enterprises” (Hunt, p. 14, 1996).

Another basic concept of process mapping as stated by Hunt (1996) is to:
“Understand a process or system by creating a “process map” that
graphically shows things (objects or information) and activities
(performed by men or machines). The process map is designed to
property relate both things and activities”.

The data collection process started with the process of familiarization with
the factory floor from which data will be collected. This was the process-mapping
phase. The factory floor is divided into stations and certain activities are
performed in each station. First, all the stations were clearly identified. Each
station is then visited and studied carefully. The purpose is to identify the
activities taking place in each station and their relationship with one another.
Feeder stations are also identified, the process that takes place in the feeder
station, material used, and activities are all observed and recorded. To illustrate,
after observing and mapping stations 13, 14, and 15 in one of the factories, a list
of major activities was generated as follows:

- Setting the roof

- Roof Insulation

- Roof Decking

- Cover Decking with paper

60

- Shingles with paper

- Exterior boards

- Installation of Exterior Doors and Windows

- Siding

The activity relationship is as shown below in Figure 4.1. This activity

relationship is simply in the form of a process flow diagram.

 

Roof
Insulation

 

 

 

l

 

Setting
the Roof

 

 

 

I

 

Exterior
Boards

 

 

Roof

 

Decking

 

 

 

 

Cover
Decking

with Paper '———>

 

 

 

 

 

 

 

Installation
of Exterior
Doors and
Windows

 

 

 

 

Shingles

 

 

Siding

 

 

 

Figure 4.1 — Example Activity Relationship

61

 

an

sis

flo

Similar sketches were developed for all the stations throughout the factory and
their activity relationships shown. Another product of the process-mapping
phase is shown in Table 4.1 that lists the activities in each station, the crew size,
and feeder names. If a feeder station accompanies a main station, then that
station is also visited and mapped. With the process mapping data, the process

flow diagrams can now be developed.

62

Table 4.1 — Process Mapping Data Sheet

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Station # Activities Crew Feeder Comment
Size Name
8 Roof Setting 3 Roof
Feeder
8 Finish Ceiling Wiring 1 Activities sometimes
overlap into successive
stations
9 & 10 Roof Decking 3
11 Cover Roof Deck WI 3
Paper
12 Install Shingles 2 Shingles
Feeder
8 Exterior Boards 2
9 & 10 Exterior Doors and 2
Windows
1 1 Siding 2 Siding
Feeder
Feeder Name Activities Crew Main Comment
Size Station
Roof Feeder Trusses are
assembled
Shingles Shingles package is 2 12
Feeder unwrapped and
Placed on moving
bucket to installers
Siding Feeder Cut siding material 1 11

 

to spectified lengths

 

 

 

 

63

 

4.2.2 Process Flow Diagrams

Process flow diagrams are also called simply, flowcharts. To fully explain
what a flowchart is, let us begin with a simple analogy. A globe is a scaled
representation of the earth. A flow chart is a representation of a process. The
process can be a business process, a construction process, or any process that
includes work tasks and activities. “Flowcharting is defined as a method of
graphically describing an existing process or a proposed new process by using
simple symbols, lines and words to display pictorially the activities and sequence
in the process” (Harrington, p. 86, 1992). Let us consider a flowchart for the
process of a simple earth moving operation. Figure 4.2, shows a simple

flowchart of an excavation operation.

Start

 

 

 

Excavate

I

Load

l

Haul

l

Dump

l

Return

Figure 4.2 - Example Flowchart of Earthmoving Process

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

64

 

 

The process begins when the soil is first excavated and ends when the truck has
returned to its original site for another load. This flow chart is typical of most flow
charts in its ability to simplify a very complex operation. To be able to represent
a complex process clearly is very big advantage of flow charts. Another
advantage is the ease of identifying bottlenecks in the process by using
flowcharts. It clearly shows how activities/work tasks are done, whether they are
supposed to be done that way and help in determining how it could be done
better. The author has encountered four different kinds of flow charts during the
literature search phase of the project. These four different kinds of flowcharts are
discussed below. Only the flowcharting techniques of the first two will be utilized
in this project. Illustrations for functional and geographic flow charts can be very
complex. Since they will not be used in this project, examples of them will not be

shown only definitions will be provided.

BLOCK DIAGRAMS: “A block diagram, also known as a block flow diagram, is
the simplest and most prevalent type of flowchart” (Harrington, 1991). Figure 4.2

above is classified as a block diagram.

ANSI STANDARD FLOWCHARTS: ANSI is an acronym for American National
Standards Institute. ANSI flowchart is a more detailed form of a block diagram
and deals with more detail interrelationships among activities. Figure 4.3 is a
detail flowchart of the estimating process of a small homebuilding company. The

estimating process mainly takes place in two different phases of a project. One

65

is at the onset of a project when the front end needs to know the starting price for
the house. The other time a thorough estimate is done, is during the project
when prices and contractors change. An estimate is needed at this point to
adjust the budget. A major difference between the first and second estimate is
that in the second one, actual numbers supplied by the contractor is used to
come up with the budget. Minor value engineering estimates are also done
throughout a project. The two paths in Figure 4.3 represents the two different
times estimates are performed. Only one of two paths is followed after the
decision is made as to whether a new estimate is being performed or an update

in being performed.

FUNCTIONAL F LOWCHART: “A functional flow chart pictures the movement
between different work units, an additional dimension that is particularly valuable

when total cycle time is a problem” (Harrington, 1991).

GEOGRAPHIC FLOW CHART: “A geographic or physical flowchart analyzes the

physical flow of activities. It helps to minimize the time wasted while work output

and/or resources are moved between activities” (Harrington, 1991).

66

 

Beginning
of project

I

Access
already
built
templates

1

 

 

 

 

 

 

 

 

 

Input set
numbers

 

 

 

l

 

Input
activities
set up by
square feet

 

 

 

l

 

 

Use historical
data for

activities such
as basements

 

 

Estimating
Process

Beginning
of project

or during

project?

I

   
 

 

l

 
 

 

 

 

 

~>

 

Do individual
take-offs for
the rest of the
activities

—W

 

 

During

 

 
   
 

 

project

 

 

l

 

Access
already
built
templates

 

 

 

l

 

Input
numbers
given by
vendor

 

 

 

 

V

 

rFinish I

summarize
estimate

 

 

 

k

Figure 4.3 — Estimating Process Using ANSI

67

4.2.2.1 Standard Flowchart Symbols

Several symbols are used in the development of a flowchart. The symbols

shown below are just a few that will be used to develop flowcharts for the

process models on this project.

 

 

 

 

 

 

d

Operation: Rectangle. Use this symbol
whenever a change in an item occurs. The
change may result from the expenditure of
labor, a machine activity, or a combination of
both.

Boundaries: Elongated circle. Use an
elongated circle to show the beginning and
end of the process. Normally, the word start,
stop, or end is included within the symbol.

Connector: Small circle. Use a small circle
with a letter inside it at the end of a flowchart to
indicate that the output from that part of the
flowchart will serve as the input to another
flowchart.

Decision point: Diamond. Put a diamond at the
point in the process at which a decision must
be made.

Direction of flow: Arrow. Use an arrow to
denote the direction and order of process
steps. An arrow is used for movement from
one symbol to another

Figure 4.4 — Standard Flowchart Symbols
(Source: Harrington, 1991)

68

4.2.3 Time Study

The primary reason for doing time study is to measure job performance; that
is productivity as stated by Lawrence S. Aft in his book Productivity Measurement
and Improvement (1992). In addition, time study according to Carroll (1972) is a
way of gathering facts. The data collection for this thesis project is also
concerned with gathering facts; such as the fact of how long it takes to perform a
certain activity. Different companies and industries have different needs for
performing time study according to Aft (1992). He mentions that use of time
study to the critical operation of an organization could include;

. Developing production schedule

0 Determining wage payment plans

0 Estimating manufacturing costs

0 Providing a base for estimating productivity increases

0 Identifying employee training needs

. Appraising employee performance

0 Justifying the addition of production capacity

The time collected in this thesis project, will be used to eventually identify
bottlenecks and resource utilization for the manufacturing process being studied.
To have a timing device is the first requirement for formulating a time study. The
timing device could be a stopwatch, a wristwatch or any form of electronic device
that measures time. For this project, both a stopwatch and a wristwatch were

utilized. After establishing the timing devices used, it is important to understand

69

the process that is to be studied. This was done by doing process mapping as

explained above.

4.2.3.1 Cycle Time

“Cycle time is the total length of time required to complete the entire
process“ (Harrington, 1991). This time includes the time it takes to perform the
work and the time spent waiting for something, time spent chatting, or time spent
waiting for resources to be available. The time that it takes to complete the
process if there are no interruptions or stoppage is the actual cycle time; that is to
say the time it will take to do the job in a perfect situation. Total cycle time can
be thought of as actual cycle time plus idle time. Both total cycle time and actual
cycle time are collected for this research project. This applies to both individual
activities and combination of activities in station(s). The data sheet used for this
process is shown in Table 4.2. This datasheet was used to collect data on the
following parameters:
Chassis Number: Every section going through the line has a number. This
number is the chassis number and it is usually written on the steel chassis. The
number is the identification number for the product.
Station Number: This number is entered depending on the station that is being
observed. Multiple station numbers can be used based on the activity

relationship.

70

Table 4.2 - UCIMSU - NSF Manufactured Housing Project Data Sheet

 

Station # Station Description:

 

Chassis Sub Sub Sub Sub Sub Sub
Number 1 2 3 4 5 6 Comment

 

Duratio

 

 

Time

 

No. of labor

 

Time I

 

No. of labor

 

Time I

 

No. of labor

 

Time |

 

No. of labor

 

TimeT

 

 

No. of labor

 

Activity Description

 

Activity related Material

 

Activity Precedesl
Follows

 

 

 

 

 

 

 

Supporting Stations Data

 

Number Activity Duration Labor Material Sub Sub Sub
Description Force Activity Activity Activity

 

 

 

 

 

 

 

 

 

 

 

 

 

Activfiy’ Related Material: Material that is being used to accomplish the task.
Activy PrecedeleolIows: This is the predecessor and successor of the activity
being observed.

71

 

Duration: The column under duration is used to enter the total cycle time for all
the activities. This is from the start time of the first activity in the cycle and the
finish time of the last activity.

Sugmrting Stations data: Supporting stations data is used for entering data on
feeder stations.

A very important aspect of collecting data was to make sure idle time and
actual work time are both captured in any duration of the activities. This further
illustrates the point made above about the difference between total cycle time
and actual cycle time. The problem was diffused by entering the actual start time
and finish time of the activity cycle using a wrist watch and also using a
stopwatch to record simply durations for each activity. This means that stop time
for the last activity on the station could be subtracted from the start time of the
first activity on the station and subcontracting lunch and break time, that will give
us the total cycle time. This means durations/absolute cycle times can be
gathered, and total cycle time gathered, including idle time.

For practical reasons, it was necessary to merge activities from certain
consecutive stations while collecting time data. The reason for this was that
activities sometimes overlap to successive stations. As a result, certain activities
can occur in multiple stations especially when there is a delay of some sort. The
activity is identified with the station it normally starts in, but that same activity

could overlap to the next station.

72

4.2.3.2 Data Sampling and Population

It is usually impossible to collect a large amount of data in a study to
confidently say that one has data on every single scenario imaginable. More
data can always be collected because there is no way of knowing if one has
exhausted all possible sets of data. To get a complete data set for this research
would be to collect data on every single size and design of manufactured houses
produced in both factories. Doing that would qualify as a complete set of data on
the whole population of houses produced. Since time and resources does not
allow this, the statistical concept of sampling is used to collect the data. The
concept of sampling says that the set of data collected should be representative
enough to make inferences about the whole population. In this research, the
eleven data points on roofing stations and exterior finishing activities, nine data
points on interior finishing activities, and five data points on axle and tires were
from diverse sizes and design of houses to draw inferences on the whole
population of houses produced at both plants. Majority of data was collected on
the most popular size in both factories but also included are the small and large

extremes.

4.3 PRODUCTION PROCESS FLOW DIAGRAMS BY CLUSTERS
This section shows the production flow diagrams of the five main
manufacturing clusters of the plant. These areas are floors, walls, roofing and

exteriors, interiors, and finishing.

73

FLOOR CLUSTER

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1-a Transfer Chassis Vapor
Chassis to 1-b Framing Barrier
Pads & “'——’ Floor
Hoses building
Plumbing Electrical Heat Duct R.W.
Roughing Wires and Fix Insulation
+——" Roughing “—" ‘
Cold & 1-c Glue floor Decking
Hot water Move to sheets
network "———’ floor
decking
If A l
\ Transfer Drill Holes
to station fPrwate'
2 lrnes

 

 

 

 

IfB

 

 

Figure 4.5a - Process Flow Diagram - Floor Cluster
(Source: Modified from UC Research Team)

Floor cluster is divided into three different parts. Station 1-a is the chassis
preparation, 1-b is the floor building part and station 1-c deals with floor decking.
Figure 4.53 shows activities such as chassis framing, insulation, and rough
electrical and plumbing work. The floor is then moved to station 1-c were the
deck is laid. The activity called holes consists of drilling holes on the floor for

74

openings for water lines. At this point, A and B are moving in parallel as shown

 

 

 

 

 

in Figure 4.5b.
FLOOR CLUSTER CONT’D
Sanding
If A
’

 

 

 

Sweeping

 

 

/\

 

 

Air
pumping

 

 

 

 

 

Water
Seal

APP'Y
Glue

 

 

 

\/

 

 

 

Floor Tile
Fix

 

 

l

 

 

Transfer
to station
3-a

 

 

Figure 4.5b — Process Flow Diagram - Floor Cluster Cont’d

 

(Source: Modified from UC Research Team)

Figure 4.5b is a continuation of the floor cluster stations with tiling being the

dominant activity. This activity is mainly performed on the section that contains

the bathroos and is labeled as section A. Section B is mostly idle at this point

because it needs carpeting instead of tiles and the carpeting activity is performed

on the interior finishing station.

75

WALLS CLUSTER

  
 
 

  
 

Is unit
Section
A or B?

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Bathtub Plumbing Transfer
transfer to station
3-b
Cabinets
transfer
v
Transfer Electric Interior
to exterior Install wall
' wall F— setting
setting

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 4.6 — Process Flow Diagram — Walls Cluster
(Source: Modified from UC Research Team)

The first symbol on the wall cluster diagram represents a decision point.

The sections are usually divided with the bathrooms occupying one section and

the kitchen the other section. The flowchart in Figure 4.6 represents section A as

the one were the bathrooms are located and hence the installation of the bathtub

and plumbing activities. Section B contains the kitchen and hence the cabinets

transfer. Cabinet and bathtub transfer means that they are transferred from the

feeder stations and installed. Not all bathroom fixtures are installed at this

station. For example, both factories visited install toilets in the finishing stations.

After these fixtures are installed, then the interior walls can be set.

WALL CLUSTER CONT’D — EXTERIOR WALL SETTING

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

End wall Sidewall Marriage
setting ,_, setting ,_N wall
setting ‘
Framed
Floor wl
interior Gas-line
walls Installation
Move to
[station 4-b ]—>
Electrical Furnace &
Outlets Electrical
’ box
Installation

 

 

 

 

 

 

 

Figure 4.7 - Process Flow Diagram -- Exterior Wall Setting & Fastening
(Source: Modified from UC Research Team)

WALL CLUSTER CONT’D — EXTERIOR WALL FASTENING

 

Face

board \ Face
setting board ‘
Section seam

available .
from 4_ a Sidewall 1
Fastening

 

 

 

 

 

 

 

 

 

 

 

Ready to
move to
station 5

 

 

 

Perimeter
Rail I
Fastening

 

 

 

 

Figure 4.7 cont’d

77

Section 4 mainly consists of wall setting. The walls are built on a feeder
station and moved via an overhead crane to the main production line. First, the
end walls are set, then the sidewalls, and finally the marriage walls. Electrical

work including wiring for the fixtures are perfon'ned whiles the walls are going up.

ROOFING AND EXTERIORS CLUSTER

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Roof Roof Cover Shingles
Insulation Decking Decking ‘
“—9 ‘T——’ with Paper .L__, F—F
Setting Exterior Installation Siding
the Roof Boards of Exterior
“—" '—’ Doors and —’ ' ’
Windows

 

 

 

 

 

 

 

 

 

 

 

Figure 4.8 - Process Flow Diagram - Roofing & Exterior Wall Finishing

Roofing cluster starts when the whole roof structure built from the feeder
station is transferred to the main line to be set on the section. From this point on
two sets of activities are taking place. The roof is being worked on and in parallel
so are the exterior walls. Work is also being performed inside the house but the
activities are considered the major activities. After the roof is set and secured to
the walls of the house, insulation is sprayed on the roof. This gives way for the
trusses to be covered with roof boards. In other words, the deck on top of the
roof is placed. For drainage purposes, the deck on top of the roof is covered with

78

paper before shingles are placed. At the same time the above—described
activities are taking place, the exterior walls are also being finished. First, the
walls are covered with exterior boards. Door and window openings are cut so
that the doors and windows can be installed. Then the house is covered with
siding. These are the major activities happening in the above-described stations.

After the roofing and exterior finishes have been performed, interior finishes can

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

be started.

INTERIORS CLUSTER

Install

Urethane

Foam \ Install Installing Repairs
Carpet ’ Vents ’ Nicessary

Molding /

 

 

 

Figure 4.9 — Process Flow Diagram - Interiors

Urethane foam is the material used to pad the floor before the carpet is
installed. The flow diagram shows that at the same time the urethane foam is
being installed, moldings is being places on doors, windows and floor and wall
linings. This does not always happen as the author has observed the two
activities happening in sequence. Carpeting is then installed; the vent openings
are cut and the vents coverings placed on the openings. This clears the way for
the house to be moved to station six depending on how much repairs are

79

needed. Repairs from this point on are ongoing activities but they do not affect
the progress of the station unless it is something major such as a wrong
bathroom fixture or a whole section of drywall to be replaced. In such cases then

the work may be delayed while repairs are in progress.

80

 

 

 

 

85
new o_x<

 

 

 

\ a... 50.0

3301

86358

 

FINISHINGS CLUSTER

 

2:301 \

 

 

69.8

832.32. 5\ :98.

85:2?

 

 

 

/

3-56.0 w
cozm=flm£
8:532

 

 

 

 

 

 

 

Began. T1

 

 

 

 

is“.
.325

 

 

Econ
cause:
:39:

 

 

 

 

was...
as 22

 

 

 

 

 

26am”.

 

No

 

 

Yes

 

   
    

3:353
new:
canon
moon.

Figure 4.10 — Process Flow Diagram — Finishing Cluster

81

When the house moves to the finishing stations, it can go in two different
directions. Some of the houses coming down the line already have wallpaper on
them and others need to be painted. Whether the house needs painting or not,
axles and tires have to be installed. In cases where the house needs painting,
the drywall process has to be completed first. This is usually a one-day activity
since the drywall has to be left overnight to dry. Painting is also a one-day
process for the same reason. In cases where the house does not need painting,
axle and tires are installed simultaneously with kitchen and remaining bathroom

appliances. The house is then vacuumed, dusted and completely cleaned.

4.4 DETAILED PRODUCTION PROCESS AND MATERIAL FLOW MODEL

Over a period of several weeks of recording time data among other things,
the familiarity with the process grew, coupled with the process mapping session
and the process flow diagrams developed at that stage of the project. It is now
appropriate to develop a detailed production process and material flow model for
manufactured housing. The model will be developed in two parts. The first part
starts with the chassis framing and ends at the wall stations, based on the flow
diagrams above, and the second part begins from roofing and exterior up to the
end. This process model is transformed into a simulation model in the next

Chapter.

82

-——9

 

Floor
Transfer

 

 

 

Floor
Ready

 

 

 

 

 

 

 

 

 

 

 

'0
CO)
war (05
as ——————b c»-
wE .53
“to .:.c
«CL.
Cu. 33
f N
t
E
(D
L /

 

Figure 4.11 — Production and Material Flow Model: First Half of Factory

83

comma
ucmw

 

 

mcficmm

 

 

 

 

N conga
9 535:.

 

 

 

 

0:280
50E

 

 

mo=>>

Satan

 

 

9.52,

 

 

 

 

383
8:0 “mm:

 

 

 

£3sz
uoEmE

Figure 4.11 cont’d

    

em .0 <
cozoom
as a.

 

 

 

Row Lm~m>>

 

 

 

   

 

x_..._
6E. 52“.

 

 

 

 

mcEEzn.
._<

 

 

@53on

 

 

 

 

 

8.0 >32

 

 

800.5

d

Figure 4.11 cont’

85

 

 

 

 

 

Lon—acme
£9.58

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4
Al. 9.53 All 65cm:
:9, 8:95 35:55 nafimm
£3.95:
9: E: a
2.25 nacho
8:25 3253

a a8:

 

   

8 a <
888
as a.

 

Figure 4.11 cont'd

86

 

 

3 Sean
2 90.2

 

 

 
     

$83...
8:82.“.

 

 

  

 

 

 

 

 

 

 

 

:25
35:52

 

 

 

.3385

 

 

 

 

 

 

 

 

 

 

8:588...

xon 8:588...
8030mm £0.50
w 885:“. .moEomm.
8:588...
05.880

. Alt
95.8 =95 ft 85$ 953
03:85. =m>>o5w .83 new

 

 

  

583

8:25
:5 80K
uoEmE

 

 

 

 

  

 

:22.
28mm

 

 

Figure 4.11 cont’d

87

 

 

m :03ma
2 o>o_2

 

@5886“.

:3".
8858a.

 

 

 

 

 

 

@5889.
.8335

 

 

 

 

 

3:8 Emma
:88 comm

 

 

 

958m
:83
com...

 

 

 

mé :8:
29588
8:08

 

Figure 4.11 cont’d

88

PART 1

The first part of the model shows the chassis wheeled in with nothing on it.
the floor is built with insulation, wiring and duct work all in place. After the floor
has been covered, it is sanded and tiled as applicable. The floor is then installed
with cabinet and bathroom fixtures. More plumbing work is performed. The wall
sections are then placed accordingly, wired, and fastened appropriately. At this

point, a framed house exists.

PART 2

With the frame ready, the roof structure can be moved from the feeder
station to be set on top on the house, one section at a time. Insulation is sprayed
on the roof, the roof covered and shingles installed at the same time doors and
windows are being installed followed by siding. The major activities now consist
of interior finishes such as molding around the doors and windows, pad the floor
with urethane foam before carpeting the floors provided the house does not need
painting but already has wall paper instead. The last finishing stations consist of
installing axles and tires for transportation, installing the last of the kitchen
(refrigerator, oven) and bathroom (toilet) appliances. The house is then

vacuumed thoroughly and ready to be shipped out.

89

 

 

 

96:53 8

 

8:8
=85

 

 

 

asses.
8:8

 

   

0.53.03
.2 2028
Sofia

    
  

 

98 2000
:22...

 

 

 

$855
w 2000

8:252

8.8.3.5

 

 

 

053.080 A1
woo”.

 

 

 

 

.8800

 

8:2xw T
:99:

 

 

 

2.82.
8.85

22250

Foom.

 

 

 

seem
28.35

 

 

 

 

 

8.86»
.828: m
5.; 88

 

 

 

seem «ow

 

 

 

Figure 4.12 — Production and Material Flow Model: Second Half of Factory

90

 

 

Sauna.
80E

 

 

 

 

 

 

8.05.8
=2m5

 

 

 

..

 

8.8.8

:80...

2852:

asses.

 

 

asses.
asses.

 

88a.
80m.

 

 

Lemma

5.; x80
89 880

 

 

Figure 4.12 cont’d

91

     
   

asses.
Ewen.
0852:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Ewen.
. ,, 8:80 9 28520 F
.885 .885
82....
:8
5?.
82F .w
8:2. .885
_ 8.5“.
:85:
288m

 

 

 

 

 

 

        

«@588:
com:
08:0...
8000

 

 

 

885:0
.885

 

 

Figure 4.12 cont’d

92

 

 

K

8:0...
88:58

82....
w 8:2.

  
      

  

88:28.2

8.5“. w 8
3.8.82

.

 

a:

-820 28 A
8.8.885
38:84.

 

 

 

 

82....

Ba 92 /

 

 

 

 

288m
.28
8.8.885
88:84.

 

8:80
.885

 

 

 

 

 

 

8:80

 

 

8.5a:

 

 

 

8.5“.
.8380

 

 

Figure 4.12 cont’d

93

4.5 SUMMARY
This chapter described the process-modeling phase of the thesis. The

chapter started out by describing the concepts of process mapping and how it was
utilized. Different types of process flow diagrams were discussed with examples,
which emphasize block diagrams and the American National Standards Institute
format. Time study background and how time data was collected followed the
process mapping discussion. Process flow diagrams were then developed for
different clusters such as floors, walls, roofing, exteriors, interiors, and finishing,
showing activity relationships. These process flow diagrams lead to the detailed
production process and material flow model. The model was divided into two parts

to make the transition to the next stage, (the simulation model) much easier.

CHAPTER FIVE
SIMULATION MODEL

95

5.1 INTRODUCTION

In this chapter, a simulation model is developed using cycle time data
collected from the manufactured housing plants and the developed process
model from Chapter four. First, computer simulation will be discussed including
general use and specific utilization to the manufactured housing industry.
Construction simulation and its evolution will be briefly discussed. Then,
simulation software used for this research will be introduced. The next step is to
describe the steps used in transforming the detailed production and material flow
model developed in the previous chapter into a simulation model. Finally, results
obtained from running the simulation model are presented and the discussion is
extended to potential uses of the simulation model to study alternatives to

improving overall production efficiency.

5.2 WHAT IS SIMULATION

Construction simulation or any other kind of simulation is a way of
mimicking reality. Simulation attempts to represent on paper what is happening
in real life as close as possible. Several different definitions of simulation were
encountered during the literature review phase of this project. The definition that
suits this work best states: “The object of developing the model of a production
system is to examine the interaction between flow units, determine the idleness
of productive resources, locate bottlenecks and estimate production of the

system as constituted” (Halpin and Riggs, 1992). In Chapter four, flow units have

96

already been developed and in this chapter, idleness of productive resources will

be studied along with the location of bottlenecks.

5.3 TYPES OF SIMULATION
There are two different types of simulation. These are discrete-event
simulation and continuous simulation, both of which are briefly defined in the next

two subsections. Discrete- event simulation is used for this thesis.

5.3.1 Discrete-event Simulation

In discrete-event simulation, the system is reproduced using activities,
which start and end according to specified durations and number of attributes.
According to Pritsker (1986), “the state of a system is defined in tenns of the
numeric values assigned to the attributes of the entities”. The system is
characterized by its state at any given point in time. In discrete-event simulation,
the assumption is that the state of a system changes instantaneously at specific
times marked by events (Martinez, 1996). An example of discrete-event
simulation include cars arriving at a traffic signal involving a finite number of units

(cars) or activities (arrivals), which start and end according to specified durations.

5.3.2 Continuous Simulation
In the case of a continuous simulation model, the state of the system is
represented by dependent variables that change continuously over time

(Pritsker, 1986). Examples of continuous simulation include nuclear reactions

97

and population growth whose states are changing continuously. Continuous
simulation requires differential equations to define the continuously changing

state of the system.

5.4 WHAT IS SIMULATION USED FOR
Simulation has been in the academic and business management fields for
the last several years before it came to construction. Universities have used
simulation to perform research on Industrial operations. Furthermore, simulation
is used because of the cost and time it takes to build a real system and run
experiments on it, whether it is experiments on aerodynamics, manufacturing, or
business process. Unlike manufacturing, in construction real-size full-scale
models/prototypes are almost non-existent. The following are further benefits of
simulation: (BCM 817 Class Notes, Spring 2001).
- Identifying needs and allocating resources
— Understand the interaction between resources
— Trade-off alternatives
— Determine the duration of operations taking into account the non-deterministic
nature of things

— Determine operation costs

5.4.1 Use of Simulation for Manufactured Housing
The benefits of simulation to the manufactured housing industry will be

discussed in the context of the benefits described above. Manufactured housing

98

utilizes assembly-line production with activities and processes, which require the
use of resource. Proper use and allocation of these resources is essential which
requires a determination of operation durations and cost, and trade-off
alternatives. In simulation, these issues of concern are usually studied. This
section will deal with how some of these issues can be applied and used in the

manufactured housing industry.

5.4.1.1 Identifying Needs and Allocating Resources

The needs of all plants are unique and so are the type and amount of
resources needed. Simulating the process over a certain amount of time, one
can expect to find durations over which the activities in the system consume
resources and deliver the product. Simulation will provide an estimate for
resources needed based on what the user wants to produce or the needs of the
user.

To illustrate, consider a new company in the planning stages of setting up
their first plant. This particular company may have a unique product or a unique
style of manufacturing, which are not well documented, making staffing decisions
quite haphazard. In other words, they want to know what their needs are and
how to allocate the appropriate resources. In this case, the production system
could be modeled using simulation \M'th estimated numbers on labor use. After
running the system, the results could be used to find out what the labor utilization
is and then changes made accordingly to the best possible scenario. Although

simulation may not be able to optimize the system, it will be able to give results

99

based on how well resources are being utilized. The plant can be staffed
according to such information. Let us consider Figure 5.1. An activity requires
one unit of material, one forklift, and one laborer to start, but there is a crew of
three. The first laborer is pulled and after activity “A is finished, the laborer
moves with the process to activity “B”. The second laborer is pulled as soon as
the forklift is available so activity “A” can start again. When the first laborer is
finished with activity “B” it is pulled back to the queue so she could be used
again. The question now is how many laborers are needed in other to maximize
the utilization of both the forklift and the laborers. One does not want to have the
forklift waiting for availability of laborers or a very large wait-time for the laborers
either. How does one allocate the ideal number of resources (laborers) in this
case to maximize utilization? Simulation will be able to help answer such a

question.

100

 

 

Material ' - Activity A —-—H Activity B

100
fi

 

 

 

 

 

Figure 5.1 — Need and Resource Allocation Example

5.4.1.2 Understand the Interaction Between Resources

Similar to real life, simulation system entities, which include all kinds of
resources, come together to make an activity happen. The resources discussed
in this section are those that are shared by multiple activities. The timely start or
completion of activities, depend on the availability of all the applicable resources.
Consider there is an existing crane that moves walls from the feeder stations to
the main assembly line. When the crane is not moving walls, it is sitting ideal.
Similarly, a separate crane exists for moving roofs from the roof feeder station to

the main production line. When the roof crane is not transporting roof structures,

101

it is sitting ideal. It is feasible that one of these cranes can be utilized for both
activities. In both factories visited during this project, the walls and roofing
stations are close enough to each other that there should not be any problems
with the movement of the crane from station to station. The effect this will have
on production can be answered by simulation. Simulation can show whether the
crane is constantly moving, is it keeping up between the two stations, and if not,

will installing a faster crane meet the criteria.

5.4.1 .3 Trade-off Alternatives

The process of trading-off alternatives involves trying to balance the costs
and benefits of different options. A company may want to invest in new
equipments to help improve the production on the floor. They may be
considering two cranes with different capacities and speeds, or they may be
considering pumps for delivering roof insulation with different speeds. Analysis
of overall production rate on each option can be perfomled. In cases where cost
of equipment is not known, certain simulation software can provide cost data.
The above example is for comparing two new alternatives. The argument is also
valid when the two alternatives are being compared and one already exists. It
illustrates as to how trading one alternative for the other, affect production. Let
us say a company is considering installing new overhead crane to move walls
from the feeder station to the main line. The new crane has a larger capacity and
faster speed. The questions the company wants answered could include: (1)

how would the process be affected by installing a faster crane? (2) can we

102

produce more houses? (3) how much more houses can we produce in a given
amount of time? (4) how much will our profitability increase by increasing
production? Specifically, on the interaction between resources, the company
might want to know: (1) how much more wall insulation will be needed if we are
to produce so much more units? (2) how will staffing be affected by changing
production rate as a result of the new crane? (3) Will the ordering cycle for other
materials be affected?

Simulation can help answer these questions by studying the plant with the
current crane and develop a model. Based on the speed of the new crane under
consideration, the time input can be changed and simulation results will show

impact on the process. Answers to other questions will follow.

5.4.1.4 Determine Durations of Operations and Operations Cost

This is a very important benefit to manufactured housing. Since we live in
an uncertain world, durations are almost never deterministic. Approximately,
how long will it take to produce so many units? Measuring duration for a whole
operation can be a very tedious task. Furthermore, the non-deterministic nature
of things suggests that a single measurement is not enough to determine the
right duration. Simulation will allow the user to input few durations and then
repeat it for a large number of times. This will allow averaging out durations,
which will then be close to the actual duration. The same concept can be applied
to manufactured housing since building a single house can take multiple days

with many activity durations and different cost variable for slightly different

103

options on the houses. So simulation will not only help answer the question, how

long it will take, but also how much is it going to cost?

5.5 CONSTRUCTION SIMULATION

Construction simulation was developed over time to meet certain
demands of the industry especially the highway/transportation industry.
CYCLONE, developed by Daniel Halpin in 1977 has become the basis from
which most simulation software has been developed (Halpin, 1977). Several
attempts though have been made before that to develop a credible system of
network modeling from which simulation can be perfon'ned but none of which had
the impact of CYCLONE. Two of these attempts were that of Teicholz’s “link
node” model in 1963, and Halpin’s CONSTRUCTO in 1976. STROBOSCOPE
(State and Resource Based Simulation of Construction Processes) is the most
recent simulation language in academia. To make STROBOSCOPE available to
the less rigorous users of construction simulation a version of STROBOSCOPE
called EZSTROBE, which uses a graphical user interface, was developed

(Martinez, 1998). EZSTROBE is what will be used for this research.

5.6 INTRODUCTION TO EZSTROBE
The simulation software used for this project is called EZSTROBE.

EZSTROBE is a simulation system with the following features (Martinez, 1998):

— Uses MS VISIO as the front end of the simulation system with which the user

104

interacts and constructs an Activity Cycle Diagram of the construction

operations being modeled

— Uses STROBOSCOPE Simulation Software as the simulation engine where

the actual simulation is performed and results are generated

— Allows the use of variables to parameterize the modeled construction

operation

— Allows the use of standard and customized results to be generated.

EZSTROBE follows the basic modeling elements developed by Halpin for

CYCLONE and discussed in detail in Chapter one, section 1.5.3.1. It uses

normal, combi, and queue nodes for developing Activity Cycle Diagrams.

5.7

STEPS IN SIMULATION MODELING

The following steps are generally used in developing simulation models

(Paulson, 1995, Sawhney and AbouRizk, 1993).

1.
2.

Define Problem
Develop logic network
Determine or estimate:

— Operations time durations
— Production quantities
— Cost information

Transfer logic network to computer simulation program
Run model
Validate model

Use model for decision making

105

These steps will now be defined as each applies to this research. Some of these
steps have already been accomplished in previous chapters and will not be

reproduced.

5.7.1 Define Problem

“In the problem definition stage the goals of the simulation experiment and
the boundaries of the problem are defined” (Sawhney and AbouRizk, 1993). The
problem in this research is to show how the production process of manufactured
housing can be improved. A process model has been developed based on the
layout of two different factories. From that model, 3-4 stations will be
transforrmd into a computer simulation model. The computer simulation model

is the focus of this chapter.

5.7.2 Develop Logic Network

In this stage, the complete description of the construction operation which
has to be modeled is developed (Sawhney and AbouRizk, 1993). Logic network
is the same as activity cycle diagram or the process model. As detailed in
Chapter four, developing the process model included steps such as the process
mapping, use of process flow diagram techniques like block diagrams and ANSI
format. The process flow diagrams were further developed during the data
collection stage when the understanding of the activity relationship was greatly
improved. This finally led to the production process and material flow model

shown in Figures 4.11 and 4.12.

106

5.7.3 Determine or Estimate Operations Time Durations

In this phase, the input data required for the simulation experiment is
collected (Sawhney and AbouRizk, 1993). Depending on a particular goal of a
project, this stage could involve: Operations time durations, production quantities,
and cost infomiation depending on what the problem is. In this research,
operations durations were the main data collected. Actually, this stage translates

into the data collection stage. Data collection is fully detailed in section 4.2.

5.7.4 Transfer Logic Network into Computer Simulation Program

In this stage, the preliminary model is converted into a form that can be
easily understood by the computer (Sawhney and AbouRizk, 1993). This step is
the main purpose of this chapter. To transfer the Activity Cycle Diagram into a
simulation model, we need more than notations used for building process flow
diagrams. At this point, instead of using block diagram notations or the ANSI
format, the basic modeling elements from Figure 1.2 will be utilized. The use of
Normals, Queues, and Combis is what transforms our logic network into a
simulation model. Figure 5.2 shows the simulation model transfomied from the
production process model. The simulation model came about as such: one of
the goals of this project is to develop a simulation model for the manufactured
housing process. An extensive process model (logic network) was developed
with the assistance of the University of Cincinnati team and the team from
Michigan State University. These process models are shown in Figures 4.11 and

4.12. Time durations were collected over a period. A part of that process model

107

is now being transformed into a simulation model as one of the authors major
contribution to the research. In addition, “what if“ scenarios for possible process
improvements to the system will be presented in this Chapter.

At this point, the reader will be reminded that only four consecutive
stations were transformed into a simulation model. The process being simulated
starts from where the wall frames are installed to the station were axles and tires
are installed. Before the model is explained in detail, Table 5.1 is presented to
show alphabetically all the queues and activity notations used in the model. The

model is then presented in Figure 5.2. A detailed explanation of the model

follows.

Table 5.1 — Notations for Simulation Model

ACTIVITY NOTATIONS IEXPLANATIONS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[CoverRDWPaper Cover Roof Deck With Paper

IGetCarpet Get Carpet

InstallCarpet Install Carpet

lnstallDW Install Doors & Windows

InstallExtBoard Install Exterior Boards

lnstallShingles Install Shingles

InstallSidigtL 'lnstall Siding

InstallTires Install Tires

lnstalIUreFoam 'Install Urethane Foam

lnsulateRoof Insulate Roof

Molding Install Molding for Interior Doors & Windows
RoofDecking Build Deck on top of Roof Trusses
SectionControll [Dummy Activity to Control Process

SetTruss [Roof Truss Structure placed on Framed House

 

 

 

108

Table 5.1 - Cont’d

QUEUE NOTATIONS EXPLANATIONS

 

 

 

 

 

 

CarpetReady Carpet Readyfor Installation
CompletedHouse |House production complete
Control [Dummy queue to control model
|DoorFeeder [Material - Door Feeder

 

[ExteriorBoardsFeeder [Material - Exterior Boards Feeder
[ExteriorBoardslnstal [Install Exterior Boards

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[FramedHouse [Framed House
[HousecoveredWSi [House Covered With Siding
[lnsulatedRoof Insulated Roof
finsulationMater Insulation Material
[LaborDWWt [Labor for Door & Window Installation
[LaborExtBoarth [Labor for Exterior Boards
[LaborlnsulatWt [Labor for Roof Insulation
[LaborMoldinth [Labor for Interior Door & Window Molding_
[LaborRDWt [Labor for Roof Decking
[LaborRPWt [Labor for Roof Paper
[LaborShinglesWt [Labor for sniggles Installation
[LaborSidinth [Labor Siding
[LaborTiresWt [Labor for Tire Installation
[LaborUFWt [Labor for Urethane Foam
[Molds [Material - Interior Doors 8. Window Mold
[RoofBoards [Material - Roof Boards
fictbecked [Roof Decking Completed
RaofFeeder Activity Feeder Station for Building Roof
WofPaper [Material - Roof Paper
[RoofRForShingles [Roof Ready for Shingles
[RoofReady Work on Roof Completed
IRoolsofl=oam [Material Urethane Foam
SettersWt [Labor Roof Setters
ShinglesFeeder [Material Feeder - Shingles
SidingReady [Finished Installing Siding
Tires Material - Tires

iresReady [fires Ready to be Installed
SideShigle Material For Sicmg
Sidinglnstalled Siding Installed
Windoereder [Material Feeder - Windows

 

 

109

      
      
  

RoofFeeder InsulationM

RF

l

T-* lnsulatedR

     
 

SetTruss
oof
Normal[43.5,10]

 
   

[ I

«a A
:"Sidin Read '-_
lnstallDW I». a 5

Normal[44.2,9.8]

InstallExtBoard

  

 

 

Normal[29.7,18.5]

 

DoorFeede
r
DF

 

Figure 5.2 - Simulation Model of a Section of Manufactured Housing Production

110

  

:"Root'R Fors "z
,, .

Rooflaaper . .
hingle

RP

    
      
 

     

RoofD ecke
d

RoofBoards

  
   
 

CoverRDWPaper
Normal[18.12.8]

  
 

RootDecklng

 
   
 

   

 

  

Normall18.7,6]

  
   
 

InsulatedR

 

 

00!
RB
SectionControll
SidingRead
sideshingle .....
Y

 

g'lHouseCovsH

  
 
 

 

InstellSiding

  

 

Normal[41.6.1l.2]

  

Molding
Normal[27.9.8.6]

 

Figure 5.2 cont’d

111

 
     
   
 

ShinglesFe

eder

     
        

RoofRForS
hingle

    
   
 
   
 
   

lnstallShingles TlresReady

InstallTires ~———-———

Normal]43.3,1.7] Normal[46.8.6.5]

 

 

LaborTires

Completed

H so
RoolsofF as ou

 

 

 

 

  
    
 

 

 

 

 

 

 

 

LaborUFWt ~ m
nLUF 'oF
InstallCarpet [T'
CarpetRea
—- ___._ lnstallUreFoam 0-. GetCarpet d Nomall33 4 14 2]
y ' ' '
Normall23,10] 1
HouseCove
aborCarpe
redWSl

 

Figure 5.2 cont'd

112

The process goes as follows: there are two paths shown in the model.
Activities on both paths are being performed on each section that passes
through. The top path will first be explained followed by the bottom path.

TOP Path: a framed house enters the first station applicable; the crew
starts setting the truss from the feeder station as shown in the first cornbi. The
activity time duration has a normal duration of 87 minutes with standard deviation
of 30.9 minutes. The activity needs resources from two queues to start. These
are the setters and the roof from the feeder station. The activity that follows the
roof setting is root insulation. This activity takes place with one worker on top of
the trusses and spraying insulation on the roof using a hose. This activity is
supported by two queues, one representing the worker spraying the insulation
and the other representing the insulation material. After insulation, the roof
decking activity follows. Roof decking consists of laying the sheets of roof
boards on top of the roof structure. The roof decking combi shows three queues
feeding into it. One is for the labor, the second one is for the roof boards and the
last queue represents the fusion queue from the previous page. The acitivity that
follows roof decking is to cover the roof with paper. This is done for drainage
purpose and to prevent freezing. Again, the activity needs labor and material.
Installing shingles is the last roofing activity. From there, the tires are attached
and the house is ready for interior finishes.

BOTTOM PATH: the first activity on the bottom path is the installation of
exterior boards. “InstallExtBoards” represents the material feeder needed for this

activity and the other queue is for labor. Installation of doors and windows

113

follows. This activity has two separate material feeder queues one for doors the
other for windows. The top queue represents labor. After the doors and
windows are installed, siding is placed on the walls. After this activity, interior
finishing activities are represented on the bottom path such as molding the
interior doors and windows, installing the padding for the carpet represented on
the model as Urethane Foam. After the carpet in laid, the boundary of the model
of this research stop there.

There is a dummy combi and queue called “SectionControl” and “Control”
respectively on the model. These are not activities happening on the factory floor
but they were inserted on the model to serve a very important purpose. The
activities on the lower half of the model, progress faster than those on the upper
half. This is caused by the large duration of the roof setting activity. Install
exterior boards, doors and windows, and siding crews sometimes have to wait for
the assembly line to move because of space restrictions on the factory floor, they
can progress only so much. The dummy combi “SectionControl” forces the lower
activities to stop and wait for the upper ones since they are working on the same
section. “SectionControl” needs 0.25 of a resource from both the top and bottom
paths for the process to continue which in reality is the same 0.25. When the
process is running and the lower path reaches “SectionControl”, the 0.25
resource delivered is kept until the top portion releases the same 0.25 resource
before 0.5 is released to the queue “Control”. Unless both 0.25 are in the combi,

the section will not be released.

114

Referring back to the model, quarter (0.25) of a house is removed from
queue “framed house” each way which is received by the activity “SetTruss” and
“InstallExtBoard”. The activities on the top and bottom of the sheet are actually
being performed on the section simultaneously. The quarter (0.25) of a framed
house resource represents a section being sent down the line. Since quarter
(0.25) plus quarter (0.25) represents half a house or one section, the model is
releasing one section at a time down the production line by saying if the content
of the queue is greater than zero, pull a quarter (>0, 0.25). The process
continues as such down the line. The cornbis show the statistical distribution of
the data in the form of Nomial [87, 30.9] as in the case of “Set Truss” combi.
This shows a normal distribution with mean 87 minutes and standard deviation of
30.9 minutes. The factory floor is limited in space. The notation (==0,0) is a
constrain, which says, if the content of the successor queue is not equal to zero,
then do not start the predecessor activity (combi). This is to ensure that the
model does not show the houses stacked on top of each other because that is
not what happens in the factory. The dotted queues are EZSTROBE’s way of

fusing two pages and facilitating continuity of one model.

5.7.5 Run the Model

Running the model is the process of getting results. This also serves as
debugging and testing to find out if it behaves as expected. Tabular and
graphical reports are checked to make sure the results conform to what was

observed on the field.

115

Simulation models in EZSTROBE can be parameterized to give results in
many formats. One is cost and another is resource utilization. For the purposes
of this research, resource utilization will be analyzed with focus on labor
resources. The labor parameterization is shown on Table 5.2. All the material
and labor queues are parameterized for more efficient trial and error runs. Table
5.2 and 5.3 show a list of parameterized name and value of labor and material
queues. The value on material is from the estimates of how much material it will
take to build 624 houses based on the cornpany’s materials listing. The result
drawing element is used to tabulate formulas for all the queues under
investigation. The mathematical formula used to calculate utilization is:

Utilization = 1 - (QueueWait.AveCount) I number of resources.

The formulas for all the labor resources are tabulated in Table 5.4.

Table 5.2 - Labor Input Parameters

 

 

 

 

 

 

 

 

 

 

 

 

Name Description Value
nS Labor Setters 4
nLEB Labor Exterior Boards 3
nLDW Labor Doors 8 Windows 3
nLl Labor Insulation 1
nLRD Labor Roof Decking 3
nLSI Labor SidinL 3
nLRP Labor Roof Paper 3
nLM Labor Molding 2
nLSH Labor Shingles 3
nLUF Labor Urethane Foam 1
nLT Labor Tires 2
nLC Labor Carpet 3

 

 

 

 

 

116

Table 5.3 — Material Input Parameters

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Name Description Value

FH Framed House 624

RF Roof Feeder 624

EBF Exterior Boards Feeder 678912

IM Insulation Material 21840

WF Window Feeder quantity 3120

DF Door Feeder 1248

RB Roof Boards 917280

SS Sidirg shirgles 474240

RP Roof Paper 499200

M Molds 62400

SF Shingles Feeder 870480

RoF Rools of Foam 61152

T Tires 3744

C Carpets 61152

Table 5.4 - Result Formulas

SettersUt Setters Utilization - Roofing 1-(SettersWt.AveCount)/ns
LabExtBoardsUt Labor Utilization - Exterior Boards 1-(LaborExtBoarth.AveCount)/nLEB
LaborDWUt Labor Utilization - Doors 8 Windows 1-(LaborDWWt.AveCount)lnLDW
LaborlnsulationUt Labor Utilization - Insulation 1-(LaborlnsulatWt.AveCount)/nLl
LaborRDUt Labor Utilization - Roof Decking 1-(LaborRDWt.AveCount)/nLRD
LaborSidingUt Labor Utilization - Siding 1-(LaborSidinth.AveCount)/nLSI
LaborRPUt Labor Utilization - Roof Paper 1-LLaborRPWt.AveCount)/nLRP

 

LaborMoldingUt

Labor Utilization - Molding

1-(LaborMoldinth.AveCount)/nLM

 

LaborShinglesUt

Labor Utilization - Shingles

1-(LaborShinglesWt.AveCount)/nLSH

 

 

 

 

 

 

LaborUrethaneFoam Labor Utilization - Urethane Foam 1-(LaborUFWt.AveCount)/nLUF
LaborTiresUt Labor Utilization - Tires 1-(LaborTiresWt.AveCount)/nLT
LaborCarpetUt Labor Utilization - Carpets 1il.aborCarpetsWt.AveCount)/nLC

 

117

 

With the above formulas entered into the model, it can now be run to find
out the labor utilization. Table 5.5 shows definitions of queues measured

attributes. Table 5.6 shows definitions of activities measured attributes.

Table 5.5 - Definitions of Queue Attributes

 

 

 

 

Cur Content at the Time of the Report
Tot Total Amount of Resource to Ever Enter
AvWait Average Waiting Time
AvCont Time-Weighted Average Content

 

SDCont Time-Weighted Standard Deviation of Content
MinCont Minimum Content
MaxCont Maximum Content

 

 

 

 

 

 

Table 5.6 - Definitions of Activity Attributes

 

 

 

 

 

 

 

 

 

 

 

 

 

Cur Current it of Times the Activity is Being Performed at Time of Report
Tot Total Number of Times it has Started
1stSt Time at which the first instance started
LstSt Time at which the last instance started
AvDur Average Duration
SDDur Standard Deviation of Duration
MinD Minimum Duration
MaxD Maximum Duration
Avlnt Average Time Between Successive Starts
SDlnt Standard Deviation of Time Between Successive Starts
Minl Minimum Time Between Successive Starts
Maxl Maximum Time Between Successive Starts

 

 

118

 

Table 5.7 — Labor Utilization

 

Setters Utilization - Roofing: 0.9808

 

Labor Utilization - Exterior Boards: 0.3462

 

Labor Utilization - Doors & Windows: 0.5035

 

 

 

 

 

 

 

 

 

 

Labor Utilization - Insulation: 0.4503
Labor Utilization - Roof Decking: 0.2130
Labor Utilization - Siding: 0.4689
Labor Utilization - Roof Paper: 0.1997
Labor Utilization - Molding: 0.3195
Labor Utilization - Shingles: 0.4919
Labor Utilization - Urethane Foam: 0.2609
Labor Utilization - Tires: 0.5299
Labor Utilization - Carpets: 0.3801

 

The Future Events List is empty at simulation time = 109989
Total Number of Named Objects: 159

Total Number of Variables: 249

Total Number of Statements: 43
Table 5.7 to 5.9 shows the results of running the model with 624 houses, which is
assumed to be roughly a years worth of production. Table 5.8 shows the content
of each queue after the simulation run on the first column. The amount to
resources passing through each queue during the entire process is depicted in
the second column. Average waiting time of each queue is the third column
followed by the time weighted average content, standard deviation, minimum,
and maximum content of queues during the operation. The activity results Table
is a detail report of durations after the operation is done. Average duration of

activity is shown as well as the average interval of successive starts. Also shown

is the first and last time each activity was initiated.

119

 

Table 5.8 — Queues Statistical Report

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Queue Cur Tot AvWait AvCont SDCont MinCont MaxCont
CarpetReady 0 312 1.13 0 0.03 0 0.25
Carpets 0 61152 55458 30578 17629 0 61152
CompletedHouse 624 624 55458 311.73 179.89 0 624
Control 0 624 55406 0 0.01 0 0.5
DoorFeeder 0 1248 0.08 621.45 359.78 0 1248
[ExtBoardsFeeder 0 678912 55228 337526 195716 0 678912
[ExtBoardslnstal 0 312 55139 0.16 0.12 0 0.25
EarnedHouse 0 624 57.67 310.55 179.89 0 624
[HouseCoveredWSi 0 312 0.15 0 0.01 0 0.25
InsulatedRoof 0 312 0.18 0 0.01 0 0.25
[lnsulationMater 0 21840 55344 10898 6296.1 0 21840
[LaborCagietsWt 3 3747 55.44 1.87 1.45 0 3
[LaborDWWt 3 3747 45.27 1.53 1.5 0 3
[LaborExtBoarth 3 3747 57.83 1.95 1.43 0 3
[LaborlnsulatWt 1 1249 49.25 0.55 0.5 0 1
[LaborMoldinth 2 2498 60.61 1.37 0.93 0 2
[LaborRDWt 3 3747 70.12 2.37 1.22 0 3
EborRPWt 3 3747 70.53 2.38 1.21 0 3
LaborShinglesWt 3 3747 45.52 1.54 1.5 0 3
[LaborSidinth 3 3747 47.02 1.59 1.5 0 3
[LaborTiresWt 2 2498 42.03 0.95 1 0 2
LaborUFWt 1 1249 65.7 0.74 0.44 0 1
Molds 0 62400 55404 31172 17989 0 62400
[RoofBoards 0 91728 55384 458052 264440 0 917280
fictbecked 0 312 1.81 0.01 0.04 0 0.25
[RoofFeeder 0 624 55255 310.88 179.89 0 624
[RoofPaper 0 499200 55404 249373 143912 0 499200
[RoofRForShingle 0 312 1.38 0 0.03 0 0.25
[RoofReady 0 312 1.45 0 0.03 0 0.25
RoolsofFoam 0 61 152 55433 30564 17629 0 61 152
SettersWt 4 4996 1 .6 0.07 0.53 0 4
ShinglesFeeder 0 870480 55424 434998 250953 0 870480
SidingReady 0 312 44.65 0.13 0.12 0 0.25
ires 0 3744 55468 1872.4 1079.4 0 3744
TiresReady 0 312 0.67 0 0.02 0 0.25
SideShiggles 0 474240 55316 236528 136715 0 474240
Sidinglnstalled 0 312 46.04 0.13 0.12 0 0.25
Windoereder 0 3120 55228 1553.6 899.44 0 3120

 

 

 

 

 

 

 

 

 

120

 

 

$30.?qu

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8.8: 2.2 3.8 8.8 2.8: 8.: 8.: 8:: 3.28: 8.: 2.2 :

8.2: :5: 2:: ::.8 8.: 8.: 8.: 8.: :.8:: 3.22 2.2 : 5888.89:
8.2: 8.2 :3: 3.8 8.8 8.: 8.: 3.2 3.3:: 8:: 2.2 : 8.3.898:
8.2: 8:: 2:: ::.8 2:: 8.: 2.: 8.8 :.8:: 3:2 82 : I856...
8.8: 2.2 3.8 ::.8 8:2 8.: 8.2 8.8 3.32: 2.8: 2.2 : 08:88.85.
38: 8.2 3.8 8.8 2.8 8.: 3.2 :18 3.8:: 2.3: 82 : 58:22.83.—
::.::: 8.8 :8 ::.8 2.8 2:: 3.: 8.8 3.82: 8.2: 2.2 : 8.5.82..—
::.2: 3.8 8.8 8.8 8.8 8.: 8.: 8.3 8.82: 8.8 82 : 852252:
8.2: 3:: 3.8 ::.8 2:: 8.8 :5 8.8 8.3:: 8.2: 82 : 8.8.8.82:
8.2: 8.8 8.8 8.8 8.8 8.: 3.2 8:: 2.38: 8.: 82 : 2:85:83.—
::.2: 8.8 3.8 8.8 3.2 2.: 8.: 2.3 8.88: 8.8 82 : >888...—
ehms: 8.: 8.8 3.8 8.3 8.: 2.: 8.8 8.22: 8.2: 2.2 : 89888....
3.8: 8.: 8.8 3.8 8.: 8.. 8.: 8.: 8.22: 8:: 82 : 898.30
8.2: :5: 8.8 ::.8 8.8 8.: :2 8.2 :.8:: 3:2 82 : 8.8.8528
:62 .52 5:: 2.2 96.2 9...... 58: 5:2 83.. 8.2 :8. 50 3.28....

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.68: .8888 8:28.: - :.: 8.8:

121

5.7.6 Validate the Model

Validation of the model requires that the data that was collected and
inputted, to give the same results as to what was observed on the factory. It is a
process of comparing model results with reality. In this research, since the
statues quo from the factory is modeled, validation is done strictly by comparison.
At the time of recording time data from the first factory, the production rate of the
factory as stated by management was 2.5 houses (5 sections) a day. This was
also observed while collecting data at the factory. When the model was run with
the eleven points (5.5 houses) collected from the factory, it showed a production
rate of 4.9 sections a day. This production rate was improved to 5.4 sections a
day when the model was run with 624 houses. Six hundred and twenty-four
(624) is an assumed number of houses the factory can produce based on the
production rate of 2.5 houses a day, under ideal conditions. Six hundred and
twenty-four (624) is the product of 2.5 houses a day by 5 days a week by 50
weeks a year (2.5 * 5 * 50). The exact result is 625 but for practical purposes, an
even number was needed so 624 is used. The fact that the observed production
rate (5 sections a day) is close to the simulated production rate (4.9 sections a

day) validates the model.

5.8 USE MODEL FOR DECISION MAKING I IMPROVEMENTS I WHAT IF

SCENARIOS

In this sub-section, five possible scenarios will be introduced based on

122

how they can improve production. Of these five, two will be inputted into the

model and the results analyzed. All five scenarios were shown to industry

representatives for their input and possible pitfalls. The scenarios are, what if:

1.

2.

Roof setting activity durations was significantly reduced.

Insulate roof activity was moved to the feeder station.

Two separate production lines running in parallel with one set of
feeder station serving both lines were laid on the same factory floor.
A quality control team was established to do repairs during finishing
stages.

One crane could be used for movement of wall sections and roof

structures from feeder stations to main production line.

5.8.1 SCENARIO ONE

Table 5.7 shows low utilization of all the labor except the setters. This is

consistent with what was observed on the factory floor. The roof setting activity

takes so long that the other activities were always severely delayed. The future

event list empty at simulation time 109989, means that it took 109989 minutes to

finish the operation. The calculation for the production rate is shown below.

624 houses _, 109989 minutes = 1833.15 hrs

Production Rate = (624 house I 1833.15 hrs) = 2.72 houses per day

2.72 houses per day approximates to 5.45 sections a day. The observed

production rate at the factory was 5.5 sections a day.

123

Referring back to the issue of low labor utilization, the benefits of
simulation can now be shown because what if scenarios can be use to find a way
to remedy the situation. For example, how will reducing the duration for the roof
setting activity by half, affect the production? This can be done by changing the
quality of roof setters by hiring more experience workers or by increasing the
crew size. Input time reduction by half into the model shows different numbers
on labor utilization as shown in Table 5.10.

Table 5.10 — Labor Utilization with Reduced Duration on Roof Setting Activity

 

 

 

 

 

 

 

 

 

 

 

 

 

Setters Utilization - Roofing 0.82801
Labor Utilization - Exterior Boards 0.59054
Labor Utilization - Doors & Windows 0.84240
Labor Utilization - Insulation 0.74489
Labor Utilization - Roof Decking 0.35723
Labor Utilization — Siding 0.78782
Labor Utilization - Roof Paper 0.35519
Labor Utilization — Molding 0.53800
Labor Utilization - Shingles 0.82771
Labor Utilization - Urethane Foam 0.43692
Labor Utilization — Tires 0.89867
Labor Utilization — Carpets 0.63806

 

 

Statistics report at simulation time = 65274.6
The labor utilization for the setters decreased by sixteen percentage points but
that of all the other labor crews increased. The labor utilization for tires
increased by thirty-seven percent (37%). More importantly, the average time
between successive starts decreased from eighty-eight (88) minutes to fifty-one
(51) minutes. The total operation time dropped to 65274.6 minutes. What is now

the production rate with these new numbers?

124

624 houses _, 65274.6 minutes = 1087.91 hrs

Production Rate = (624 housesl 1087.91 hrs) = 4.59 houses per day

There is a forty percent (40%) increase in production just by solving the delay

problem with the setters.

5.8.1.1 Sensitivity Analysis

In this section, a sensitivity analysis will be performed to study what effect
further reduction of roof setting activity duration has on both labor utilization of

the activity and overall production. Figures 5.3 and 5.4 show the results of the

analysis.

Table 5.11 — Sensitivity Analysis Data

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IDuration(Minutes) Utilization(%) Production Rate
10 34.32 4.59
20 47.1 4.51
30 59.61 4.44
40 73.53 4.26
50 83.96 3.97
60 88.91 3.6
70 94.95 3.21
80 96.57 2.93
87 98.08 2.72
90 98.63 2.59
100 99.31 2.35
1 10 99.6 2.17
120 99.65 2.02
130 99.79 1.84
140 99.87 1.71

 

 

 

 

125

 

o:

o9.

CE.

0:.

53355 m> confine - m_w>_mc< b_>:_mcmw m6 959”.

$552 :_ 8:930

oo_. om um om on om om

ov

om

ON

9.

 

 

 

 

 

 

 

 

 

 

 

T

i

I

 

 

 

 

co=m~___5 IT

 

 

ON

CV

00

cm

00?

ON_.

(%) uonezunn

126

cozoauoi m> cozmso - m_m>_mc< b_>=_w:om - v6 059“.

$552 E 225.50

mm ow on

om

om

0v

om

ON

0..

 

 

 

T T T
n. .— LO. 0
‘- O

 

 

 

c2839.”. IT

 

 

 

1T T T T T T
L0 L0. <1- ‘0. co to. N
<- co N

Aep Jed sasnoH - atea uononpOJd

127

Figure 5.3 shows that as the activity duration increases, the utilization also
increases, but the rate of increases becomes smaller as shown in the graph
turning into a horizontal line. The production rate on Figure 5.4 keeps

decreasing as the activity duration increases.

PITFALLS FROM INDUSTRY

There is no pitfall with this scenario according to the industry personnel. This is
no longer a problem in the specific plant that it arose from according to the
production manager. That claim though cannot be justified by the author for two
reasons. The first is, the scope and timing of the research did not allow for
another time study to justify the claim. The second reason is the whole plant has
gone through tremendous overhaul in terms of its production rate and labor force
since the initial data was collected. They currently have half as many people in
the plant and produce sixty-percent fewer houses because of economic

downturn.

5.8.2 SCENARIO TWO

The model on Figure 5.2 shows the activity roof insulation right after roof
setting. Insulating the roof on the main assembly line is actually practice in only
one of the factories. After the roof structure has been assembled on the feeder
station, it is moved to an adjacent station on the feeder station and the ceiling
boards painted. While it is being painted, roof insulation could be an activity

happening simultaneously. This may require some quality control in order to

128

avoid the paint mixing up with the insulation. The result of implementing these
changes in the model gave a simulation time of 107361 minutes. The production
rate is:

624 houses ———> 107361 minutes = 1789.35 hrs

Production rate = (624 houses I 1789.35 hrs) = 2.79 houses per day.
2.79 houses per day is a negligible improvement from the 2.72 houses per day
with the insulation on the main line. This further emphasizes the point that the
set truss activity is taking so long that it affects the entire top path. The author
believes that for improvements to materialize anywhere on the top path, the
duration for set truss activity has to improve. To show this, the insulate roof
activity is moved to the feeder station with the roof set activity duration reduced
by half and the model ran again. The simulation time is 64089.62 minutes. This

translates into a production rate of 4.67 houses per day.

PITFALLS FROM INDUSTRY

According to industry personnel, the above scenario would cause
numerous problems. When the roof is being set, the setters need to know the
location of the interior walls and mating walls looking from the top of the roof.
This is so that they can screw the roof to the walls. If the roof were insulated
before setting it, the setters would have a difficult time moving insulation out of
the way to screw the roof to the walls. The same thing applies when the
electrician is doing the ceiling wirings. Insulation again will have to be moved to

get the wires on top of the ceiling panels. Another problem is that of cutting

129

holes on the ceiling for skylights. The person doing the cutting will have the

problem of insulation dropping from the ceiling on top of them.

5.8.3 SCENARIO THREE

Figure 1.1 shows a typical manufactured housing production assembly
line. There are feeder stations on both sides of the line supporting the production
process. What if two production lines running in parallel are laid out on a factory
floor large enough and have all the feeder stations between the two lines serving
both production lines. A schematic layout is shovm in Figure 5.5. The way this
system will work is that both lines will share all material and activity feeder
stations. The advantage could be the savings from using one common feeder
instead of two feeders for both production lines. This will require precise

scheduling, and highly efficient activity and material feeder stations.

130

>

/\ /\ /\/\

/\ Feeder Station Feeder Station Feeder Station

/\ Feeder Station Feeder Static

Feeder Station Feeder Station

Feeder Statw Neda Sta

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

> >

 

 

 

 

 

 

 

 

 

 

>

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

>

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 5.5 - Schematic Layout of Double Production Line Scenario

PITFALLS FROM INDUSTRY

There were two kinds of pitfalls on this scenario raised by members of the
industry during the conversations with them. The first has to deal with the actual
laying out of the building itself. The building obviously has to be large enough
which means the upfront building cost will be relatively large. Then one has to
decide what kind of capacity is needed. The normal practice in the industry is

that when more capacity is needed, a new plant is simply built on the side to

131

accommodate that extra demand. It is easier to shut down a plant when the
demand gets lower than trying to curb production in a massively large plant.
Another issue is that the activity feeder stations have to be larger than the normal
space required for two plants if they are to serve two lines simultaneously
according to the industry personnel. For example, if a plant’s roof feeder station
has three substations each 20 * 80 feet, a plant with parallel production lines may
need eight instead of six substations. Other kinds of issues raised deal with the
process such as the moving of materials between lines during production, since
the material feeder stations will be in between the two lines. This could be
solved by moving all materials for a days production before or after work hours.
Another process pitfall raised was the larger amount of space the outside line
enjoys over the inside line. Surge stations may have to be utilized to solve this
problem. The surge stations may also serve another purpose of feeding

materials to the feeder stations.

5.8.4 SCENARIO FOUR

The fourth “what if’ scenario is to implement a quality control team
towards the end of the production line. Their primarily task is to do repairs. This
team will consists of a dry wall person, a painter, a carpenter, electrician,
plumber, and any other skilled trade applicable. Their job will be to strictly do
repairs to failed inspections and damages sustained on the production line. It
was observed in both factories that it takes considerable amount of time to get

the skilled trades from their respective stations to the finishing stations to do

132

repairs when needed. For example, when there is a problem with a drywall
section that needs to be replaced, the drywall person(s) do not want to leave
their stations for fear of causing delays on the line. It is even harder when the
problem was not their fault initially. The quality control crew should be able to

eliminate this problem.

PITFALLS FROM INDUSTRY

Members in the industry suggested that this will cause sluggishness. If the
skilled trades who are doing the job initially know that a repair crew is available,
they will not do the job right in the first place. This was a unanimous concern
among all the people I talked to. Another issue that came up is that of the labor
cost. The industry usually on keeping the labor cost around twelve percent
(12%) of the dollar value of the produced houses. Adding more labor may make
it hard to keep the cost at the desired level. A third pitfall has something to do
with the way workers are compensated. lf compensation depends on the
number of houses or sections built per day, a quality control team will cause the
hasty processing of units in order to make the maximum amount of money in a
short period. Finally, most of these pitfalls can be addressed with good
management skills. If a production manger has good enough skills to make
everybody do their job right and not have to depend on someone else to cover
their mistakes, then the issue of quality control is a brilliant idea according to

some industry members. Others countered that those management skills could

133

be use to avoid mistakes in the first place and totally avoid the issue of quality

control.

5.8.5 SCENARIO FIVE

A fifth scenario that could possibly save a plant large amount of money is that of
consolidating cranes/hoists. There are usually two pairs of cranes on a factory
floor. One is for moving walls from feeder stations to the main production line
and the other for moving roofs. The cranes run on tracks anchored to the plant
ceiling and their movements are usually not very flexible. These cranes spend a
large amount of time waiting to move structures to the main line. The scenario
proposed here is to make the tracks flexible enough that one crane or one pair
can be used for both the walls and roof stations. This will reduce wait time for

the cranes and initial cost of setting up a plant.

PITFALLS FROM INDUSTRY

Pitfall number one is the cost of making the tracks flexible may offset the savings
on the cranes. Although no hard numbers are available, one-industry personnel
does not like the idea because the initial cost benefit analysis does not favor
fewer cranes. In terms of production, the issue of scheduling the production so
that the wall and roof station does not need the crane at the same time is another
pitfall from the industry. If this is not accomplished, the initial idea of reducing
wait time is invalid. A third pitfall is the speed of the cranes. The cranes can pick

up heavy material but they are very slow. The movement of the cranes over long

134

distances may slow down production even with accompanied with a good
scheduling system. This may be very costly if it causes the scheduled delivery of

a house.

5.9 SUMMARY

This chapter dealt with the simulation part of this project. The chapter
started with definition of simulation and the two different types of simulation.
These are discrete-event simulation and continuous simulation. The use of
simulation and its possible application to manufactured housing is discussed
thoroughly. Construction simulation specifically is then introduced together with
EZSTROBE. Then the steps to follow in transforming the process model from
the previous Chapter to the simulation model are discussed. A detailed
description of transferring the logic network developed in Chapter four was a
major part of this section and consequently the Chapter. Finally, a section on
what if scenarios closed the chapter. The what if scenario consists of five
options which could possible increase production or save the plant money were
discussed and with the help of people who work in the manufactured housing

industry, the pitfalls of the options were presented.

135

CHAPTER SIX
SUMMARY AND CONCLUSIONS

136

6.1 OVERALL SUMMARY

This research presented a tool for improving the manufactured housing
production process. The literature review Chapter was divided into three
different subjects. Literature on manufactured housing in general was presented,
followed by construction simulation literature and finally literature on both
construction production planning and general business planning. In addition, an
extensive list of industry terminology was presented in the literature review
Chapter. In Chapter three a detailed description of the production process was
presented. This was done to help the readers who are not familiar with the
process of building houses in a factory. To further help those readers who are
familiar with site-built housing, a comparison on codes, costs, foundations, and
production environment was also presented in Chapter three.

A detailed study of the production process on the factory floor was carried
out. The process was mapped out using concepts such as process mapping to
obtain a process model. This process model was partially transformed into a
computer simulation model. The simulation part of this project has shown how
the production process can be improved. Applications of computer simulation to
manufactured housing are presented and the “what if" scenarios presented a
procedure for trying possible improvement ideas.

This thesis is part of a larger research project between two universities
and many students have contributed both directly and indirectly to both the thesis
and the larger project. The author’s strict contributions to this thesis include but

not limited to identifying major activities during the process mapping stage on

137

several stations including those presented in the simulation model. Further
contributions include collecting time data on major activities from the roofing
stations to the start of the finishing stations. The process flow diagrams of Figure
4.8 to 4.10 were developed by the author. The production process and material
flow model was developed by the author using all the process flow diagrams.
The simulation model is also the work of the author with the help of Dr. Tariq
Abdelhamid. Use of simulation as applied to the manufactured housing plant is
the exclusive work of the author. The “what if" scenarios including the pitfalls
from industry personnel, the interview process to gain those pitfalls, were also

carried out by the author.

6.2 SUMMARY BY OBJECTIVES
In this section, each of the objectives presented earlier will be revisited. A

summary of each one of them will be presented and how they were met.

6.2.1 To Develop a Production Process Flow Diagram for Multi-Section
Manufactured Housing Production
The first step in this objective is the process-mapping phase. A factory is
visited on the first day to gain the basic understanding of activity relationship
among other things. Table 4.1 is an example of a data sheet used in this
process. Data on feeder stations, activities, crew size, and number of stations
are collected. Only major activities are identified. After this preliminary

information is collected, ANSI flow charting concepts are used to represent the

138

process observed on paper. The result is the process flow diagram of Figures

4.53 to 4.10.

6.2.2 To Develop a Production and Material Flow Process Model for

Manufactured Housing

The data collection process continues with the time study phase. The
objective in this phase is to measure activity durations of the major activities
identified in the process-mapping phase. A stopwatch and wristwatch were used
together with Table 4.2 to collect time data. After six months of data collection,
the familiarity of the process grew greatly. With the help of the process flow
diagrams developed earlier, the material listing of the housing modules, and the
observed relationship of the activities and materials, a production and material

flow model was developed as shown in Figure 4.11.

6.2.3 Transformation of Production and Material Flow Process Model into

a Computer Simulation Model

Three to four stations were picked to be transformed into a simulation
model. The stations start from the framed house to the beginning of the interior
finishing stations. The reason for picking these stations is simply based on
familiarity. The process modeling procedure discussed in Chapter one was
utilized to develop a computer simulation model. The simulation software used
for this project is EZSTROBE. The model was validated by comparing the

observed production rate on the factory floor and the production rate obtained

139

from running the simulation, which were comparable. The simulation showed
bottlenecks in the production process. The bottlenecks were included in the
“what if’ scenarios section. These scenarios are possible ways of improving the
production process. There were a total of five what if scenarios. Two of the
scenarios were specific to the model and were inputted to find out the effect on
the production process. Three of them were generic scenarios. All the scenarios
were presented to personnel working in the manufactured housing industry
including production managers and their input and pitfalls on the scenarios
gathered. That section closed the scope of the project and provided a framework
for the overall goal of showing how the production process and material flow of

manufactured housing can be improved and streamlined.

6.3 CONCLUSIONS

After analyzing the production process in two different manufacturing
plants, the author agrees with many other researchers that the manufactured
housing industry is still primitive in terms of its equipment use and use of modem
technology compared to other manufacturing industries. The manufacturing
process is very labor intensive and the industry as a whole can benefit from use
of more modern equipments such as faster cranes and a less labor intensive way
of moving materials from feeder to main stations.

Based on the “what if’ scenarios discussed in section 5.8, the author
concludes that the manufactured housing industry can gain valuable insight by

studying the process diligently. The scenarios presented in section 5.8 showed

140

how using simulation, different options can be tested and studied for possible
future implementation. The author feels that the manufactured housing industry
should also adopt manufacturing technologies from other sectors. This can be
achieved by studying and doing a systematic analysis of the production process
in order to understand what systems will work well from other industries.

Finally, developing a production process model as was done in Chapter
four should be taken to a further step of making the workers understand the
process. Each worker should not only understand their responsibilities, but they
should understand how their work affects the whole production process. This

strategy will help alleviate errors.

6.4 RESEARCH LIMITATIONS

The first limitation to this research is the selection of major activities during
the process-mapping phase. The major activities that were selected are not the
only activities taking place in the stations. More importantly, they are not the only
activities with durations taking place. Major activities are the ones that stand out
the most and easily observable to the researchers. A different research team
may select different major activities to represent the production process.

The second limitation deals with the changing nature of labor during the
production process. When a crew is delayed, they may spread out to other
stations and give assistance. This means an activity that started out with three
laborers and end up with four by the time they are finished and consequently

affecting the activity duration. During the simulation phase, this was not modeled

141

because of limitations in the simulation software used. EZSTROBE cannot
model changing labor force but its parent software STROBOSCOPE can model
changing number of laborers.

The third limitation of this research is the variability in size of the houses
observed during data collection and their effect on activity durations. The length
of the houses can range from fifty to seventy-two feet long. In few activities such
as siding, this difference in length was not adjusted for when considering the
average and standard deviation of the activity durations.

It is the understanding of the researchers from the literature review and
sampling methods that eleven data points provide sufficient data to construct a

simulation model.

6.5 AREAS OF FUTURE RESEARCH

The first obvious point for further research is to transform the whole
factory process into a simulation model. This will be more detailed and more
rigorous, consequently more bottlenecks will show up.

Another area that deserves investigation is that of inputting cost into the
simulation model. The cost of labor, material, and equipment can be
parameterized in simulation. An investigation in this area can be very helpful to
the industry especially in the area of cost benefit analysis.

The scope of this research during the simulation phase, assumed that
labor was constant throughout the process and there is always enough material

on the feeder stations available for use. This is obviously not what always

142

happens on the factory floor. It has already been mentioned on the research
limitations section that the number of laborers may change for any crew
depending on the idleness of workers. Accounting for the changing amount of
material will facilitate studying material utilization in addition to labor utilization.
Finally, another important area that can be looked at is the relationship
between the skills and experience of workers and compared to the efficiency of
the production process. It is of the author’s opinion that the production process
does not need highly skilled workers because the controlled environment makes

supervision easier. This premise is worth investigating.

143

APPENDICES

144

APPENDIX A

Roofing Station: 11 Data Points

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Activity
Chassis # Total Cycle Time
Cycle Timel Sub 1 Sub 2 Sub 3 Sub 4 Sub 5
15728 Time 232 106 21 22 24 44
No. of Labor 4 1 3 2 2
1566A Time 264 121 24 26 24 41
No. of Labor 4 1 3 3 2
15663 Time 268 1 15 26 25 18 43
No. of Labor 4 1 3 4 2
1567A Time 240 102 16 24 21 42
No. of Labor 3 1 3 3 2
15678 Time 279 123 36 23 31 45
No. of Labor 4 1 4 3 2
Time 159 67 32 10 9 41
No. of Labor 3 1 3 3 2
Time 193 51 66 21 12 43
No. of Labor 3 1 3 4 2
Time 233 102 60 10 15 46
No. of Labor 3 1 3 3 4
Time 198 55 72 16 13 42
No. of Labor 3 1 3 3 4
Time 192 62 56 14 16 44
No. of Labor 3 1 3 3 4
Time 192 58 59 15 15 45
No. of Labor 3 1 3 3 3
Sub Activity Description Set Insulation Roof Roof Shingles
Deckin
Truss 9 Paper

 

145

 

Exterior Finishes: 11 Data Points

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Chassis # Total Activity Cycle Time
Cycle Time Sub 1 Sub 2 Sub 3
1567B Time 77 17 60 55
lNo. of Labor 2 3 3
1 567A Time 96 1 3 56 85
No. of Labor 2 3 3
1658A Fl’ime 125 75 40 33
No. of Labor 2 3 3
16588 Time 135 60 50 25
No. of Labor 2 3 3
1657A Time 60 25 35 55
lNo. of Labor 2 2 2
16578 Time 100 35 55 45
INo. of Labor 32 2 2
1646A Time 1 30 2 45 50
No. of Labor 3 3 2
16468 ime 210 58 53 50
lNo. of Labor 2 2 2
1521A Time 71 13 30 19
No. of Labor 3 3 3
15218 Time 82 1 1 34 26
INo. of Labor 3 3 4
1522A Time 72 14 12 35
[No of Labor 3 3 4
Sub Activity Descriptionl Exterior Wall Door & Window Siding
Boards Installation

 

 

 

146

 

Interior Finishes: 9 Data Points

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Chassis # Total Activity C cle Time
Cycle Time Sub 1 Sub 2 Sub 3

15378 Time 118 29 32 39
No. of Labor 2 1 3

1538 Time 130 25 30 35
No. of Labor 2 1 3

1539A Time 130 33 30 35
No. of Labor 2 1 3

16528 Time 1 10 25 60 35
No. of Labor 2 3 3

1657A Time 103 45 65 50
No. of Labor 1 3 3

16578 Time 210 36 67 45
No. of Labor 1 3 3

16458 Time 126 20 13 10
No. of Labor 2 1 3

1645A Time 65 20 15 42
No. of Labor 2 1 3

1646A Time 70 18 10 10
No. of Labor 2 1 3

Sub Activity Description Molding Install Carpeting
Urethane
Foam

 

 

 

 

 

147

 

Axles & Tires: 5 Data Points

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Chassis # Activity Cycle Time
Sub 1
16438 Time 52
No. Labor 3
16568 Time 55
No. Labor 3
1652A Time 42
No. Labor 3
1657A Time 45
No. Labor 3
16578 Time 40
[No. Labor 3
Sub Activity Description Axle & Tires

 

148

REFERENCES

149

AbouRizk, S. M., (1990). Input Modeling for Construction Simulation.
Doctoral Dissertation. School of Civil Engineering, Purdue University,
West Lafayette, IN.

AbouRizk, 8., and Mather, K., (2000). “Simplifying Simulation Modeling through
lnteg ration with 3D CAD.” Journal of Construction Engineering and
Management, 126(6), 475-83.

Aft, L. S., (1992), Productivity Measurement and Improvement, Prentice Hall,
Englewood Cliffs, NJ.

Aguilar, R. J ., (1973), Systems Analysis and Design in Engineering, Architecture,
Construction, and Planning, Prentice Hall, Englewood Cliffs, NJ.

Artiba, A., and Elmaghraby, SE, (1997). The Planning and Scheduling of
Production Systems: Methodologies and Applications, Chapman and
Hall, London, UK.

Ballard, 6., and Howell, G., (1998) “Shielding Production: Essential Step in
Production Control”. Joumal of Construction Engineering and
Management, 124(1), 11-17.

Bayazit, H. C., (1997). Manufactured Housing Construction Process 8: A
Comparison of Manufactured Homes Construction and Safety Standards
with the BOCA National Building Code. Plan 8 Paper. Department of
Construction Management, Michigan State University, East Lansing, MI.

Bernhardt, A. D., (1980). Building Tomorrow: The Mobile/Manufactured Housing
Industry, MIT Press, Cambridge, MA.

Brately, P., Bennett, L. F ., and Linus, E. S., (1987), A Guide to Simulation,
Springer-Vedag, New York, NY.

Building Officials and Code Administrators lntemational, Inc. (1993). The BOCA
National Building Code. Building Officials and Code Administrators
lntemational, Inc. Country Club Hills, IL.

Burkhardt T.H., Mireley, S.L., and Syal, M. (1996). “Manufactured Housing
Research Project - Final Report,“ Michigan State University, East
Lansing, MI.

Carroll, P., (1972), Time Study Fundamentals For Foremen, McGraw-Hill Book
Company, New York, NY.

150

Chopra, 8., and Meindl, P., (2001), Supply Chain Management: Strategy,
Planning, and Operation. Prentice Hall, Inc., Upper Saddle River, New
Jersey.

Halpin, D. W., and Riggs, L. S., (1992), Planning and Analysis of Construction
Operations, John Wiley and Sons, Inc., New York, NY.

Handa, V.K., and Barcia, R. M.,(1986) “Construction Production Planning.”
Journal of Construction Engineering and Management, 112(2), 163-177.

Harrington, H. J., (1991), Business Process Improvement, McGraw-Hill, Inc.,
New York, NY.

Hunt, V. D., (1996), Process Mapping: How to Reengineer your Business
Process, John Wiley and Sons, Inc., New York, NY.

lntemational Code Council, (2000). lntemational Residential Code for One- and
Two-Family Dwellings.

Manufactured Home Installer/Servicer Course Book.

Martinez, J.C., (1996). STROBOSCOPE - State and Resource Based
Simulation of Construction Processes. Doctoral Dissertation. Department
of Civil and Environmental Engineering, University of Michigan, Ann Arbor,
MI.

Martinez, J.C., (1998). Introduction to EZSTROBE - Tutorial

Mehrotra, N., Syal, M., Senghore, O., (2001). “Manufactured Housing Trends
and Regulations In Michigan” — Project Report, Michigan State University,
Construction Management Program, East Lansing, MI.

MHI - Manufactured Housing Institute (2000). “Quick facts 1999 - 2000,” MHI,
Arlington, VA.

MMHA - Michigan Manufactured Housing Association fact Sheets (2000),
Okemos, Ml

Nutt-Powel, Thomas E. (1982). Manufactured Homes: Making sense of a
Housing Opportunity, Auburn House Publishing Company, Boston, MA.

PATH-HUD (1999). “Innovations and Cutting Edge: New Ideas in Manufactured

Housing,” US. Department of Housing and Urban Development,
Partnership for Advancing Technology in Housing, Washington, DC

Paulson, Jr., BC, (1995) Computer Applications in Construction, McGraw-Hill,
Inc., New York, NY.

151

Pen ne, R. L., (1986). Manufactured Housing: An Industry Reconnaissance,
National League of Cities, Washington, DC.

Pritsker, A., (1986), Introduction to Simulation and SLAM II, A Halsted Press
Book, John Wiley 8. Sons, New York, NY.

Robertson, 0., and Ulrich, K., (1998), “Planning for Product Platforms” Sloan
Management Review, 39(4), 19-31.

Sawhney, A., and AbouRizk, SM, (1993), “An overview of construction
simulation and its applications” Journal of Construction management,
National Institute of Construction Management and Research, India,
Volume Vlll, Number ll, July 1993, pp. 93-100.

Seidel, A. W., (1999). Production Planning Methodology for Large Homebuilding
Firms. Masters Thesis. Construction Management Program, Michigan
State University, East Lansing, MI.

Shaked, O., and Warzawski, A., (1995) “Knowledge-Based System for
Construction Planning of High-Rise Building”. Journal of Construction
Engineering and Management, 121(2), 172-182.

Suchman, Diane R. (1995). “Manufactured Housing: An Affordable Alternative”,
The Urban Land Institute: Paper 640. March 1995.

Syal, M., Smith, G., and Willenbrock J.H., (1990). “Simulation Modeling of
Modular Housing Production,” Proceedings of the Annual Convention,
Canadian Society for Civil Engineers, Hamilton, Ontario, Canada.

Syal, M., and Hastak, M. (2000). “Modeling of Manufactured Housing Production
and Material.Utilization". Research Proposal to the National Science
Foundation. Michigan State University.

Taghavi, N. (1984). Manufactured Homes: The Mobile Home Industry and a
Changing Market: Doctoral Dissertation. Department of Business
Administration, Marketing United States lntemational University, San
Diego, CA.

Tommelein, l. D., Carr, R. I., and Odeh, A. M., (1994), “Assembly of Simulation
Networks Using Designs, Plans, and Methods.” Journal of Construction
Engineering and Management, 120(4), 796-815.

US. Department of Housing and Urban Development, Factory and Site-Built

Housing A Comparison for the 21" Century, 451 Seventh Street, S.W.,
Washington, DC, October, 1998.

152

US. Department of Housing and Urban Development, Manufactured Home
Construction and Safety Standards, Part 280: Interpretative Bulletins to
the Standards. HUD, Washington, DC.

Warner, K., and Johnson, R. (1993). Manufactured Housing Research Project,
University of Michigan, College of Architecture & Urban Planning. Ann

Arbor, MI.

Warszawski, A., (1996), “Strategic Planning in Construction Companies”.
Joumal of Construction Engineering and Management, 122(2), 133-140.

www. census. gov/hhesMww/housing/hsgafl'rd/afford95/af95tp2. htrnl

153

   

IIIIIIIIIIIIIIIIIIIIIIIII