Production System Design and Implementation in the Automotive
Components Industry
by
Guillermo Oropeza
B.S. Mechanical Engineering
Massachusetts Institute of Technology, 1999
Submitted to the Department of Mechanical Engineering
in Partial Fulfillment of the Requirements for the Degree of
Master of Science in Mechanical Engineering
at the
Massachusetts Institute of Technology
February 2001
MASSACHUSETTS
1N§TITT
OF TECHNOLOGY
JUL 16 2001
D 2001 Massachusetts Institute of Technology
All rights reserved
LIBRARIES
Signature of author.....................................................
.....................
Department of Mechanical Engineering
Januay/ 1 z 0l
C ertified by .......................................................
David S. Cochran
Assistant Professor of Mechanical Engineering
.gpervisor
Th
.. ................
Ain A. Sonin
Chairman, Department Committee on Graduate Students
Accepted by...................................................
......
f
Production System Design and Implementation in the Automotive
Components Industry
by
Guillermo Oropeza
Submitted to the Department of Mechanical Engineering
on January, 2001 in Partial Fulfillment of the Requirements for the Degree of
Master of Science in Mechanical Engineering
ABSTRACT
As manufacturing systems evolve and competition increases and globalizes, it has
become increasingly important for the survival of any manufacturing company to take
full advantage of their capabilities. However, the design of a manufacturing system is a
highly complex process due to the interaction between the many elements that comprise
it. The Manufacturing System Design Decomposition (MSDD) presented here was
developed by the Production System Design Laboratory at MIT as a tool to aid the
designer of manufacturing systems to make low-level decisions while appreciating the
impact on each other as well as on high-level requirements. The methodology promotes
overall system performance optimization over local sub-optimization. It helps the
designer to separate means from objectives. The MSDD can be used to track progress and
to communicate information across the organization.
In this thesis, the MSDD is used to frame the work undertaken in the automobile
components industry. The author focuses in two Visteon plants: Indianapolis and North
Penn. The projects in which the author was involved vary greatly in terms of product
manufactured, stage of development, and analysis approach. However, they encompass
typical mass production plants evolving into lean manufacturers and transforming their
production processes. Therefore, the MSDD is very useful in analyzing, evaluating and
improving these production systems. Within the Indianapolis plant, the author was
involved in two major projects both related to the assembly of rack and pinion steering
gear assembly systems. One of them includes a financial analysis to prove the feasibility
and advantage of a cellular approach. Also, the proposed layout and conceptual station
designs are presented. The other project includes the improvement of the effectiveness of
a recently launched cell. Within the North Penn plant, a comparison of a traditional mass
production method is contrasted to a cellular approach at the production of automobile
electronic engine controllers. The comparison is performed using traditional performance
metrics and the MSDD.
Thesis Supervisor: David S. Cochran
Title: Assistant Professor of Mechanical Engineering
3
Acknowledgments
Acknowledgments
As time goes by I have realized that, just as in manufacturing, where the resulting
system is the result of the interaction of all of its elements, we are the result of our
interaction with our own environment. Every stage in our lives imprints part of the
character that forms us over time. Now that I have ended an important stage in my life I
think that it is not only necessary to recognize this but also a privilege to be able to thank
all of you who have shaped a part of my life.
I would like to start by thanking the professional front, the one that has given me
the possibility to become a better-prepared individual. First of all Prof. David S. Cochran,
my advisor, who trusted me from the beginning, and who has always had a great sense of
humor making the journey through this Master's more enjoyable. Also, I'd like to thank
the folks at Visteon who opened the doors for my work and allowed me to gain industry
experience: Paul Cope, Bill Ramirez, Jeff Clark, Stuart Anderson, Steve Watkins, Ted
Davenport and Tim Grbavac.
On the personal front, I must start by thanking my family. The ever-lasting,
unconditional love that you have transmitted to me has given me strength to endure the
various rocks I've hit throughout the past years. My parents Gabriel and Estela, who have
always been an example of success in my life, you've made me set high goals for my life
as well. Your opportune word of advice and open ears have surely strengthened the bond
that join us. My brother and best friend Gabriel, with whom I've shared so many good
times, I'd like to thank you for being there always for me. My "little" sister Estelaris,
who has always cared for me, I'd like to thank you for all the good times and your great
enthusiasm in life. That New Year's eve dress looks better on you.
My Latin friends in Boston deserve a special thank too. The awesome time we
spent together will always remain in my memory - Carlos, Pablo, Danny, Dalia, Deny,
Cesar, Ferran, Luis Mario, Rodrigo, Jonathan and Jose. Also, I would like to thank all my
friends from the lab who have created a very pleasant working environment - Kola,
Abhinav, Salim, Dan, Jongyoon, Yong-Suk, Jochen, Keith, Quinton, Jey, Zhenwei, Jorge,
Brandon, Alex, Jim, Charlie, and Ania.
5
Production System Design and Implementation in the Automotive Components Industry
Table of Contents
ACKNOWLEDGMENTS .......................................................................................................
5
TABLE OF CONTENTS ......................................................................................................
6
LIST OF FIGURES ..................................................................................................................
8
LIST OF TABLES .................................................................................................................
10
CHAPTER 1: INTRODUCTION ............................................................................................
11
1.1.- Thesis objective........................................................................................
11
1.2.- Thesis outline.............................................................................................
11
CHAPTER
2: MANUFACTURING: THE INDUSTRY AND THE SCIENCE................................
14
2.1.- The Impact of the M anufacturing Industry................................................
14
2.2.- M anufacturing as a Science......................................................................
15
CHAPTER
3: THE PRODUCTION SYSTEM DESIGN FRAMEWORK.......................................
3.1.- Lean M anufacturing .................................................................................
17
17
3.1.1.- Background..................................................................................
17
3.1.2.- Principles ...................................................................................
18
3.1.3.- Implementation...........................................................................
23
3.2.- Axiomatic Design......................................................................................
25
3.3.- The Production System Design Framework .................................................
28
3.3.1.- The Manufacturing System Design Decomposition.......... 29
CHAPTER 4: VISTEON INDIANAPOLIS STEERING GEAR ASSEMBLY .................................
4.1.- U222 Project .............................................................................................
32
32
4.1.1.- Background................................................................................
32
4.1.2.- Net present value analysis ..........................................................
33
4.1.3.- Analysis ......................................................................................
43
4.1.4.- Recommendation........................................................................
45
4.1.5.- Proposed Layout ........................................................................
46
4.1.6.- Lessons learned from the DEW98 Cell ...................
49
4.1.7.- Conceptual Station Designs........................................................
52
4.2.- U204 Project .............................................................................................
6
60
Table of Contents
CHAPTER
4.2.1.- Introduction .................................................................................
60
4.2.2.- Launch state...............................................................................
60
4.2.3.- Short-term approach: Improving labor efficiency ...........
64
4.2.4.- Long-term approach ......................................................................
68
4.2.5.- Analysis with the MSDD...........................................................
73
4.2.6.- Conclusion.................................................................................
76
5: VISTEON NORTH PENN ELECTRONIC ENGINE CONTROLLER MANUFACTURING
...............................................................................................................................
77
5.1.- Introduction ..............................................................................................
77
5.2.- Material and Information Flow..................................................................
78
5.2.1.- Material Flow ............................................................................
78
5.2.2.- Information Flow........................................................................
85
5.3.- Lamination Analysis..................................................................................
86
5.3.1.- Observed performance at lamination: transfer line and cell..... 86
5.3.2.- Analysis of Lamination Processes using the MSDD..................
88
5.3.3.- Recommendations for cellular implementation derived from the
M SD D .................................................................................
90
5.4.- Equipment Design ......................................................................................
92
5.4.1.- Equipment comparison based on the MSDD .............................
92
5.5.- C onclusions ...............................................................................................
96
C O N CLU SIO N ......................................................................................................................
97
REFEREN CE S ......................................................................................................................
98
APPENDIX A: MANUFACTURING SYSTEM DESIGN DECOMPOSITION v5.1........................
100
APPENDIX B: RECOMMENDED ACTION FOR SHORT-TERM EFFICIENCY .............................
107
C: RECOMMENDED ACTION FOR LONG-TERM EFFICIENCY...............................
109
APPENDIX
APPENDIX D: EQUIPMENT EVALUATION TOOL ................................................................
7
112
Production System Design and Implementation in the Automotive Components Industry
List of Figures
Figure 1: Hierarchy of manufacturing objectives [Hopp and Spearman, 1996]............ 16
Figure 2: Automotive "Push" Production and Scheduling System [Adapted from
C ochran, 1999]..............................................................................................
19
Figure 3: Automotive "Pull" Production and Scheduling System [Adapted from Cochran,
19 9 9 ] ................................................................................................................
19
Figure 4: The Toyota Production System Design Model [Cochran, 1999]..................
21
Figure 5: Definition of a Production System [Cochran, 1999]......................................
24
Figure 6: Mapping between Domains [Suh, 1990]........................................................
26
Figure 7: "Zig-Zagging" between FRs and DPs ..........................................................
26
Figure 8: Implementation Relation between FRs and DPs..........................................
27
Figure 9: The Production System Design and Deployment Framework ......................
29
Figure 10: First Levels of the MSDD and Schematic Overview ...................................
31
Figure 11: Trade-offs and Ideal Cycle Time for Capacity Selection in Cells [Cochran]. 37
Figure 12: Overcapacity using Cellular and High-Speed Systems...............................
38
Figure 13: High-Speed Approach to Satisfy Vehicle Assembly Demand....................
39
Figure 14: Cellular Approach to Satisfy Vehicle Assembly Demand ...........................
40
Figure 15: Cash Flow for U222 Project under Mass and Lean Approaches ................
41
Figure 16: Sensitivity of the NPV of the U222 Program to the Introduction of the F150
Program using $2.5M cells ..............................................................................
44
Figure 17: Sensitivity of the NPV of the U222 Program to the Introduction of the F150
Program using $4M cells .................................................................................
Figure 18: Proposed U222 Steering Gear Assembly Cell Layout ................................
44
47
Figure 19: Standard Work Combination Chart for the U222 Assembly Cell................ 48
Figure 20: Work Loops for U222 Assembly Cell........................................................
48
Figure 21: FR/DP Pairs Related to Equipment Design [Arinez, 2000]........................
53
Figure 22: Proposed Conceptual Station Designs for the U222 Project.......................
54
Figure 23: Proposed Conceptual Station Designs for the U222 Project.......................
55
Figure 24: Proposed Conceptual Station Designs for the U222 Project .......................
56
Figure 25: Proposed Conceptual Station Designs for the U222 Project .......................
57
8
List of Figures
Figure 26: Proposed Conceptual Station Designs for the U222 Project.......................
58
Figure 27: Proposed Conceptual Station Designs for the U222 Project ........................
59
Figure 28: Work Content per Station at Launch..........................................................
61
Figure 29: U204 Work Pattern at Launch: 14 Operators...............................................
62
Figure 30: Distribution of Work at Launch with 14 Workers .......................................
63
Figure 31: Overall Worker Utilization at Launch with 14 Workers .............................
64
Figure 32: Short-term Recommended Work Pattern: 10 Operators ...............
66
Figure 33: Distribution of Work per Cycle with 10 Operators......................................
67
Figure 34: Distribution of Work per Cycle with 10 Operators......................................
67
Figure 35: Overall Distribution of Work Time per Cycle with 10 Operators ...............
68
Figure 36: Distribution of Work per Cycle to Achieve Target Production ..................
69
Figure 37: Work Pattern to Achieve Target Output with 14 Operators.........................
70
Figure 38: Overall Distribution of Work per Cycle to Achieve Target Production ......... 71
Figure 39: Overall Distribution of Work Time per Cycle to Achieve Target Production 71
Figure 40: Average Takt Time per Worker at Launch and with Proposed Improvements72
Figure 41: Overview of Unsatisfactory FRs at the Launch of the U204 Cell.............. 73
Figure 42: EEC production steps ...................................................................................
79
Figure 43: SM D Process Sequence...............................................................................
80
Figure 44: Relative Size Comparison between Cell and Transfer Line.........................
82
Figure 45: Lam ination transfer line layout ...................................................................
82
Figure 46: Lamination "Lean" Cell Layout ...................................................................
84
Figure 47: Value Stream Map of the EEC Production .....................................................
86
Figure 48: High-Speed Line Evaluation Using the MSDD ..........................................
88
Figure 49: Lean Cell Evaluation Using the MSDD ......................................................
89
Figure 50: Proposed V alue Stream M ap ...........................................................................
92
Figure 51: PCB-Casting screw -down ............................................................................
95
Figure 52: Solder application at the cell........................................................................
95
Figure 53: Loading conform al coater ............................................................................
95
9
Production System Design and Implementation in the Automotive Components Industry
List of Tables
Table 1: Summary of assumptions and requirements for production alternatives........ 36
Table 2: Unsatisfactory FR-DP pairs at the launch of the U204 line ...........................
73
Table 3: Unsatisfactory FRs at Launch and with Proposed Long-Term Approach.......... 75
Table 4: Transfer Line Process Steps.............................................................................
83
Table 5: Lam ination Cell Process Steps .......................................................................
83
Table 6: Observed performance at the lamination transfer line and cell......................
87
Table 7: Achievement of MSDD leaf FRs at Lamination ............................................
90
Table 8: Low Performing FR/DPs at the Lamination Cell............................................
90
Table 9: Evaluating of processes at both lines using the EET ......................................
93
10
Introduction
Chapter 1: Introduction
1.1.- Thesis objective
The objective of this thesis is twofold: to present a structured theoretical
framework for the design of manufacturing systems and to show its applicability in the
automotive components industry.
Given the lack of a comprehensive and structured framework to link the various
elements of a manufacturing system, the Production System Design Laboratory at MIT
developed the Manufacturing System Design Decomposition (MSDD). This tool is useful
in designing, controlling, evaluating, and improving manufacturing systems. The
theoretical framework of the MSDD is put to practice in two Visteon plants: Indianapolis
and North Penn. The projects in which the author was involved vary greatly in terms of
product manufactured, stage of development, and analysis approach. However all these
encompass typical mass production plants evolving into lean manufacturers. Therefore,
the MSDD is very useful in framing this industry experience.
1.2.- Thesis outline
Chapter 1 defines the objective of this thesis and provides a summary of each of
the five chapters that conform it.
Chapter 2 describes the motivation for focusing in the area of manufacturing. It
describes the importance of the manufacturing industry as one of high impact in
Americans lives. It also covers the US manufacturing industry trends when seen from a
global perspective. Finally it lays out the need for a scientific-based methodology for
designing manufacturing systems.
Chapter 3 provides an overview of Japanese lean manufacturing practices, their
origin, and the principles that need to be in place for a system to mimic the Toyota
Production System. But in order to outperform their competitors, US firms need to do
more than just copying. With this spirit and encompassing offshore lessons, the
11
Production System Design and Implementation in the Automotive Components Industry
Production System Design Laboratory at MIT developed the Production System Design
Framework and the author presents it in this chapter. The cornerstone of this framework
is the Manufacturing System Design Decomposition (MSDD). The MSDD is a tool o aid
the designer of manufacturing systems to achieve the high-level objectives of a
manufacturing enterprise by decomposing these requirements into lower-level design
parameters using axiomatic design. This tool is helpful in communicating information
across the different levels of the organization, in separating means from objectives and
explaining the interrelation between the different elements of a manufacturing system and
how these achieve high-level requirements.
Chapter 4 covers the work performed by the author at Visteon Indianapolis, a
steering gear manufacturing plant. The two projects include rack and pinion steering gear
assembly systems at different development stages. The first project consists of a financial
analysis for evaluating two very different approaches for the assembly of the U222 rack
and pinion steering gears. These approaches are a high-speed asynchronous line and a
cellular approach. The proposed financial assessment method accounts for categories
commonly ignored in traditional accounting systems like scrap and inventory. The
superiority of the cellular approach is shown in terms of a net present value analysis.
Once selected this approach, the design parameters derived from the MSDD are
incorporated into a proposed layout and conceptual station designs.
The second project describes the launch of an assembly cell for the U204 rack and
pinion steering gears. An analysis of the equipment utilization is performed to identify
bottlenecks and equipment constraints. Given the existing machinery constraints at
launch, a short-term approach to improve labor efficiency is proposed by rebalancing the
work loops, and outlining minor changes. However, in order to reach target production
while still striving for labor efficiency, a long-term approach for the improvement of the
line is presented. This approach requires some modifications as outlined by the lowest
level design parameters from the MSDD. By embracing these changes and following the
standard work routines described, the cell can reduce approximately half of the wasted
operator motions present at launch.
12
Introduction
Chapter 5 presents the work carried out at Visteon North Penn Electronics Plant, a
manufacturer of electronic engine controllers for automobiles. The production process in
this plant is of particular interest for the scope of this thesis since two different
production approaches are used during one stage of the production of these modules.
These approaches are the typical asynchronous transfer line and a cellular approach. This
chapter explains the material and information flow throughout the plant. It then analyses
the production step in which the two approaches are used. The analysis is performed
using traditional performance metrics as the evaluating criteria. Further this analysis is
contrasted to an analysis made using the MSDD. Potential areas for improvement are
identified based in this latter approach. Finally, the equipment at North Penn is evaluated
through the lens of the MSDD.
13
Production System Design and Implementation in the Automotive Components Industry
Chapter 2: Manufacturing: The Industry and The
Science
2.1.- The Impact of the Manufacturing Industry
In order to better understand the importance of the US manufacturing sector it is
useful to appreciate its size and trends with a global perspective. Doing so provides a
motivation for focusing this thesis in this industry. The importance of the sector justifies
this work and the vast amount of literature touching manufacturing. Also, the trends of
the US sector reveal a need for deeper attention.
Although some may argue that the US economy is slowly moving to a service
economy, and that the manufacturing sector has steadily decreased, an actual analysis of
the employment figures reveals a different reality. According to Hopp and Spearman,
over half of the jobs in the US are tightly coupled to manufacturing. Therefore, the
potential economic consequences of loosing market share to the European or the Japanese
are enormous for the life of most Americans.
However, in this and the past decade, some indicators are in fact pointing into
trouble for the US manufacturing sector. Productivity growth relative to other
industrialized countries has slowed. Similarly, US shares in important sectors such as
automobiles, consumer electronics and machine tools have decreased. Trade deficits have
increased dramatically leaving the US as the largest debtor nation. Furthermore, in the
past two decades, the fraction of US patents granted to foreign inventors has doubled
[Hopp and Spearman, 1996].
Partly, the decline in the industry can be attributed to competition from post
World War II recovering economies; but also, to changing and increasing customers
needs. Given the daunting trend of the US manufacturing industry and its size and impact
in people's lives, it is necessary to understand the evolving conditions that have shaped
the shifts in the dominance of this industry. In Chapter 3, a review of the Japanese
manufacturing techniques is prepared. By understanding the elements that have led other
14
Manufacturing: The Industry and The Science
nations to achieve a competitive edge, and by providing a scientific-based structured
framework, US firms can begin to contribute individually to the re-emergence of the
manufacturing sector.
2.2.- Manufacturing as a Science
In order to regain market dominance it is necessary to understand practices that
have been successful in other environments. Further, it is necessary to develop a
methodology for analyzing the complex interactions that arise in a manufacturing system.
Such a methodology should enable US companies to outperform their offshore
competitors. To make this change possible, there must be an underlying science of
manufacturing that promotes cross-learning and continual improvement.
Although Frederick W. Taylor developed the scientific management framework at
the turn of the past century, neither he nor his successors set in place the theoretical
foundation for scientific manufacturing management. Most Operation Management
practices have shifted from one buzzword to the next without truly laying out a scientific
base. Hopp and Spearman argue that a major obstacle in the process of developing a
science of manufacturing is the involvement of people in factories. Given the complex
and multi-varied behavioral patterns that arise from human interactions, it is difficult to
reduce the behavior of a factory to a set of equations. They believe that since this is not
possible, a science for manufacturing can be established based on three principles:
" Intuition - Resulting from identifying and categorizing the basic behavioral
tendencies in manufacturing plants.
"
Synthesis - Resulting from assimilating information from various aspects of the
system and drawing meaningful conclusions. A framework can link the disparate
activities within a manufacturing system.
* Basics - Derived from exposure to areas such as probability, accounting, time
series forecasting, linear programming, and queuing theory.
15
Production System Design and Implementation in the Automotive Components Industry
To assist the synthesis principle, Hopp and Spearman developed a hierarchy of
manufacturing objectives, beginning with the goal of "high profitability" and
decomposed it into lower-level objectives and finally to the means for achieving these
goals ("less variability" and "short cycle times," for example). This hierarchy
demonstrates that certain tradeoffs exist when trying to achieve "ideal" manufacturing
system performance. Sometimes the means to achieve certain goals are contradicting
from one area to another. For instance, in order to have fast response to the customer, it is
required to have high inventory; also, in order to minimize cost, there should be as low
inventory as possible. Therefore there are contradicting directions to which the means
point.
High
profitability
Low
High
costs
sales
Low
unit costs
High
High
throu h ut utilization
Less
varabilit
High customer
service
Quality
roduct
Low
inventory
Fast
res onse
Low
utilization
Short
ccle times
High
invento
Many
products
More
variabili
Figure 1: Hierarchy of manufacturing objectives [Hopp and Spearman, 1996]
In line with their view, next chapter provides a structured framework for the
decomposition and synthesis of the multiple elements of a manufacturing system. By
presenting the interrelation between requirements and the means to achieve them, the
Manufacturing System Design Decomposition provides a structured framework to enable
a scientific approach for the design of a manufacturing system. The contradictive
directions of the framework presented by Hopp and Spearman is resolved by the MSDD
in part due to the interdependencies that result from the underlying foundation of
axiomatic design.
16
The Production System Design Framework
Chapter 3: The Production System Design
Framework
3.1.- Lean Manufacturing
3.1.1.- Background
As manufacturing systems evolve and competition increases, it has become a
necessity for all industries to analyze and to adopt successful manufacturing practices
elsewhere in the globe. Since the 1980's, American companies have begun to pay a lot of
attention to Japanese practices, and the term "lean manufacturing" appears today to be the
buzzword across many industries. But what does lean manufacturing really mean? How
can American companies adopt techniques developed elsewhere and still attain the same
benefits? What are the steps to implement lean manufacturing? And, is "lean" really the
future in manufacturing or is it just a temporary trend? In order to answer these questions,
it is necessary to understand what is lean manufacturing, where is it coming from, and
how is it likely to impact the manufacturing industry in the future.
In The Machine That Changed The World [Womack, Jones and Roos, 1991] a
comprehensive study of the evolution of the automobile industry is made. As the world's
largest manufacturing activity, the auto industry can be used to learn about general
industry trends. After World War I, Henry Ford and Alfred Sloan snatched the lead to
European craft producers and promoted the era of mass production. It was not until after
World War II when Eiji Toyoda and Taiichi Ohno began to develop the Toyota
Production System that marked the beginning of a new era: the "lean" production era.
The term "lean" was coined by the International Motor Vehicle Program (IMVP)
and MIT [Womack, Jones and Roos, 1991] to describe the system pioneered by Toyota
Motor Company in Japan.
17
Production System Design and Implementation in the Automotive Components Industry
After WWII, Japan could not possibly adopt the mass production system that was
at its full swing in the Western world. In order to compete, the Japanese were in a
situation that required a production system suited to their environment. The lack of
resources and the weak economy that they were facing made large capital investments
prohibitive. Toyota envisioned a system that continues to evolve today and that
represents nowadays the most benchmarked system in the world.
3.1.2.- Principles
The cornerstone of the Toyota Production System is the reduction of waste in
every possible form throughout the entire organization. Following two basic principles,
Toyota was able to reduce costs while delivering products faster and with better quality
than their Western counterparts.
One of the main principles implemented by Toyota is the idea of Just-in-time: "to
produce the necessary units in the necessary quantities at the necessary time" [Monden,
1993]. "Traditional" mass-producers schedule production based on forecasted demand of
final car sales. The scheduled amount of cars generates a production schedule for
upstream suppliers of car components and subcomponents. At each of these stages, the
production schedule accounts for a "fall-out" or defective-parts rate to ensure that the
right amount gets delivered to the subsequent stages. But the fall-out rate, as well as the
scheduled demand, is stochastic. Consequently, the amount produced at each stage,
doesn't necessarily match the amount sold, therefore resulting in inventory accumulation.
This system, known as a "push" system, is illustrated in Figure 2. According to Toyota,
the resulting inventory is waste. The monetary and human resources designated to handle
inventory are not in fact adding value to the product.
The system that Toyota implemented deals with inventory handling by taking a
different approach. There is a small amount of standard stock (standard work-in-progress
or SWIP) at each stage in the production stream. These stages include final cars, cars
components such as steering gears, engines, seats, etc, and subcomponents such as tie
rods, seats covers, etc.
18
The Production System Design Framework
Customer
Planned Vehicle Assembly
Requirements
Inventory on
Hand
Inventory
Dealer
L
MRP "Push" Production Schedule
FINISHED
Inventory
Second-tier
Subcomponents
[
Vehicle
Assembly
First-tier
Components
M
Inventory
Inventory
Inventory
ing]
Figure 2: Automotive "Push" Production and Scheduling System [Adapted from
Cochran, 1999]
When a car is sold, the car is retrieved from the assembler's SWIP. The necessary
upstream components are pulled in to the assembly plant to make another car of the same
type as the one withdrawn. This system, referred to as a "pull" system, is illustrated in
Figure 3. The retrieval from a final product from the SWIP signals the replenishment of
that specific type of product. This means that the necessary components have to be pulled
from the respective suppliers' SWIP. This in turn signals subsequent subcomponents
replenishment.
P
P
Signal
Withdrawal
P
Production
Kanban
Kanban
Kanban
--
........... Kanban
-
Delivery Pitcher
....
...
.
... ..
...
g
...
.
A
..
c
11
1
11
_L
1'
D
Heijunka Box
Second-tier
First-tier
Subcomponents
Subcoponents
SWIP
Vehle]
I
~
_
Assembly888g
SWIP
SWIP
Figure 3: Automotive "Pull" Production and Scheduling System [Adapted from
Cochran, 1999]
19
Production System Design and Implementation in the Automotive Components Industry
Whenever an item is retrieved from the SWIP at any stage, a signal is sent
upstream to request replenishment material to re-stock the missing part at the SWIP. This
is done with the aid of production and withdrawal replenishment cards or kanban.
Mistakenly, about 80% of the people think that kanban is the cornerstone of the Toyota
Production System [Shingo, 1989], when it really is just a tool to aid the principle of Justin-time, which represents in turn a remarkable difference with traditional mass production
systems.
Another important principle developed by Toyota is the idea of autonomation or
automation with a human touch. In Japanese, autonomation is translated "Ninbenno-aru
Jidoka" and abbreviated Jidoka, which might be loosely interpreted as autonomous
defects control [Monden, 1993]. This represents the second pillar of the Toyota
Production System as illustrated in Figure 4. Traditional mass producers tend to
maximize machine utilization for two fundamental reasons: to reduce investment per part
produced and to reduce labor per part produced. This results in faster and more
automated equipment. At Toyota, operator utilization is considered to be more valuable
than the equipment utilization [Shingo, 1989]. Therefore, an idle machine is preferred to
an idle worker. Since the manual work content in a sequence of operations usually varies
from process to process, having a person tied to one machine will naturally result in idle
time in some of the operators as they wait for the others to finish. This is the situation at
mass production plants, but their emphasis in machine utilization justifies it. However, at
Toyota, in order to increase the amount of value-added activities, the concept of
autonomation was introduced. By increasing the machine cycle times, the operators have
enough time to operate various machines at the same time, while the others are running.
To do this, it was necessary to separate the worker from the machine, allowing him/her to
walk away from the machine while the machine performs the operation automatically.
This is accomplished by manually feeding the parts to be processed into each machine
and by pressing a "walk-away" switch. The machine would then automatically process
the part and unload it, finishing its cycle and waiting for the operator to arrive and
retrieve the finished part and to load in the new one.
20
The Production System Design Framework
As a result, operators end up working in work loops. A cell-like or U-shape layout
is usually preferred to minimize walking distances. Standardizing these work sequences
is an important aspect to insurance of good quality and yield. Once the operator
utilization is maximized, in order to deal with waste in machine utilization, the equipment
is specified to run slightly faster that the operator's cycle time. The equipment is built to
run only at the necessary lower speed. Consequently, the resulting equipment is usually
less complex and less expensive that "traditional" mass production equipment,
eliminating the original problem that the Western producers try to address: reduce costs
by maximizing machine utilization.
TPS
Responsive4t6 Customer,
High QuaAlit bw Cost,
Vp1Ume andMix Flixi1ility
T
Leveland Balaace PrOductioii
The Foundation is Lead Time Reduction Enabled
by Less Than 10 Minute Set Up Time
Figure 4: The Toyota Production System Design Model [Cochran, 1999]
Another aspect of consideration at Toyota is the balancingof the operations to
meet the cycle time in each line and across the factory. This cycle time or takt time is
considered to be the heart bit of the factory, and is determined by the rate of demand of
the customer [Ohno, 1988]. As was previously mentioned, the equipment has to enable
the worker to interact with all the required machines in the loop and allow the operator to
21
Production System Design and Implementation in the Automotive Components Industry
complete a loop each takt time. Furthermore, all sequence of operations across the factory
and the supply chain, are specified to run at the pace of the customer. Every time a final
good gets retrieved, production upstream initiates to restock the SWIP at each station.
Customer demand dictates the time interval at which this occurs, and is referred to as takt
time.
Some may argue that still the demand is stochastic and therefore no accurate takt
time can be calculated in advance. Aware of this phenomenon, the system pioneered by
Toyota has the ability to adapt to fluctuations in demand and product variety. By leveling
the production with the aid of a Heijunka box, a scheduling board, Toyota is able to
smooth the production by damping fluctuations in demand and pacing production. Also,
with the same tool, the mix of final goods demanded by the customer can be
accommodated. Moreover, if the demand increases substantially, additional workers can
be introduced into the cell and the work content of each one of them can be reduced. The
relative simplicity of the equipment permits operators cross-training without a large
amount of effort. However, it is important to stress the need to standardize their work
loops and to carry out the sequence of jobs as designed.
Life-long employment is common in Japan. Workers develop a sense of
ownership for the company and are encouraged to suggest improvements. These
suggestions are helpful since the system itself is meant to be improved. The lessened
rigidity of the system allows small changes that improve the operators work environment
and in turn increase productivity.
Encouraging single-piece-flow, in contrast with the large batches at typical
Western mass producers, enhances quality at Toyota. Whenever a defect occurs, it is
immediately detected by the subsequent operations. Production is stopped and all
resources are drawn in to identify the origin of the defect. By asking the 5-why 's,
debuggers try to identify the root cause of the problem and solve it to prevent similar
mistakes to occur again. "At Toyota, there is only one reason to stop the line - to ensure
that it won't have to stop again" [Shingo, 1989]. Single-piece-flow allows operators to
readily identify defective parts and to search for a solution. Mass producers, in contrast,
22
The Production System Design Framework
have to produce, handle, and store huge amounts of defective parts before they even
realize they have a defective batch. They perceive large production batches as necessary
since it is to expensive and time consuming to change over from one product to the next
and back. Toyota instead developed quick change over techniques that made singlepiece-flow feasible. These techniques often referred to as Single Minute Exchange of
Dies (SMED) [Shingo, 1989], are possible since the equipment doesn't have to deal with
excessive tolerances that some mass producers specify. As previously said, the quality is
controlled by ensuring that the defects are not advanced and by introducing errorproofing devices orpoka-yoke.
3.1.3.- Implementation
As seen in the last section, there are many elements that the Toyota Production
System has invented or refined. Many of these elements are interrelated with one another,
some can be said to be common sense ideas, and some require deeper attention to be fully
understood and appreciated. Nevertheless, traditional Western companies have tirelessly
copied some of the concepts described before in an attempt to gain some of the benefits
that Toyota has. But partly because only some of the elements have been introduced in
isolation with their necessary complements, and partly because the depth of the ideas is
not fully understood, even less implemented, the efforts of Western companies have not
been as fruitful. In addition, the momentum of many years producing according to a
certain set of rules, and the mindset resisting change, hinders the efforts of Western
companies to evolve. Partial implementation of some concepts does not guarantee a
proportional partial gain in competitiveness. It is the implementation of a complete set of
elements that enable the system design to be successful. But appreciating all elements and
defining a feasible implementation path is not always clear. There are many elements
from within the organization that have to be integrated in harmony to avoid overstrain in
efforts. Some of the elements that are interrelated include people, equipment, tools,
materials, information, etc. It is the interaction of all of these elements that allows a
system to convert a set of inputs to the desired output.
At this point, it is useful to introduce some definitions for clarity purposes and for
coherence throughout this thesis. First, a system can be thought of as a set of elements
23
Production System Design and Implementation in the Automotive Components Industry
with definite inputs that are acted upon to produce a desired output [Pamaby, 1979].
Cochran makes the distinction between manufacturing and production systems as
follows:
"A Manufacturing System consists of the arrangement and operation of machines,
tools, material, people and information to produce a value-added physical,
informational or service product whose success and cost is characterized by
measurable parameters. The Production System consists of all of the elements
and functions that support the manufacturing system." [Cochran, 1999].
Production System is therefore a broader term than Manufacturing System. A
Manufacturing System encompasses all the elements that are directly involved in the
process of adding value to the inputs to yield the products of the system. A Production
System encompasses a Manufacturing System, together with the supporting elements and
resources associated with it. All the resources associated with managing, controlling and
measuring the performance of a Manufacturing System are considered to be part of the
Production System. Production System Design, therefore, involves not only the design of
all the elements of the Manufacturing System (people, equipment, information, etc.) but
also the definition of a performance measurement strategy, cost justification of the design
and overall design effectiveness (Figure 5).
Transportation
I.S.
ADept.
complex arrangement of
physical objects characterized
by
measurableparameters
Production
Engineer'n
Control
Accounting
Figure 5: Definition of a Production System [Cochran, 1999]
24
The Production System Design Framework
The Production System Design Framework [Cochran, 1999] presented in Section
3.3 presents a methodology to understand the impact of the various elements of the
Toyota Production System throughout the different levels of an organization. The
Manufacturing System Design Decomposition (MSDD), the centerpiece of the
framework, provides a structured way to understand the interrelation between the
physical implementation of the steps developed by Toyota with one another. It also
allows tracing the implementation of tools or steps back to the reason why they are being
implemented. The MSDD defines an interconnection between what must be achieved,
how it can be achieved and why it must be achieved at any manufacturing enterprise
[Cochran, 1999]. It represents a scientific-based analytical tool for the design,
implementation and improvement of any production system.
3.2.- Axiomatic Design
The underlying methodology behind the Manufacturing System Design
Decomposition is Axiomatic Design [Suh, 1990]. This methodology is a tool to aid the
designer in the sometimes not very structured design process, whether it is the design of a
physical object, a software package or a manufacturing system. The methodology
identifies three basic domains as illustrated in Figure 6. The customer domain is used to
capture the need of the customer, whether it is internal or external. Thefunctional domain
defines the objectives or functional requirements (FRs) to be achieved by the design. The
physical domain encapsulates the solutions or design parameters (DPs) to the
corresponding FRs. These three domains are interrelated with one another, and an
iterative mapping process has to be undertaken to ensure proper design implementations.
For any given FR, one can assign a corresponding DP that can fully satisfy the FR
and the process is completed. However, sometimes, these DPs need further
decomposition to convey the necessary information to the designer. Sometimes it is
necessary to assign new FRs to these DPs for which more detailed lower-level DPs are
mapped to transmit the necessary information. This process, referred to as "zig-zagging",
is illustrated in Figure 7. For any given FR, there is a DP, which in turns defines new
FRs, which need a new set of more detailed DPs.
25
Production System Design and Implementation in the Automotive Components Industry
What
Customer
Wants
(Internal &
External)
How
ON.~
FR'
Customer Domain
- Customer needs
Expectations
Specifications
Constraints, etc.
DP's
Functional Domain
e Design Objectives
-0
Physical Domain
Physical
Implementation
-0
-0
Figure 6: Mapping between Domains [Suh, 1990]
This process continues until the identified DP can be implemented without further
refinement of objectives and solutions.
"Zig"
FRI
F R11
FR12
DPI1
"Zag"
FR13
DPI 1
Functional Requirements
Functional Domain
DP12
D13
Design Parameters
Physical Domain
Figure 7: "Zig-Zagging" between FRs and DPs
The axiomatic design methodology is based on two fundamental axioms, which
should accompany the thought process of the designer. These axioms are:
26
The Production System Design Framework
*
The Independence Axiom: An optimal design maintains the independence of the
FRs. In this case satisfying a particular FR should not affect the feasibility of
satisfying another FR. At best, the DP for an FR can be adjusted without
affecting other FRs. If this is not the case, then one or all the FRs infringing on
one another should be reformulated to eliminate the interdependency.
*
The InformationAxiom: The second axiom states that an optimal design should
minimize the information content. Therefore the best design is one with no
coupling and with a minimum of information content.
Based on the independence axiom, the order of implementation or precedence can
be inferred. Ideally, a design should have, according to the independence axiom, one DP
for each FR. This is represented mathematically by the diagonal or uncoupled matrix in
Figure 8 and the order of implementation is arbitrary. Given that in reality some DPs
might affect several FRs, it is desired to implement those first and then implement the
DPs that affect less or one FR only. This way, the solution proposed to a problem is not
outdated. Other DPs should not further modify the FR being addressed by a given DP. A
partially coupled matrix represents this process, and the order of implementation is path
dependent. Additionally, if all DPs affect all FRs, the resulting matrix is coupled and the
implementation path is uncertain. If such a relationship exists, it is advisable to try to
decompose the DPs further until at least a partially coupled interrelationship is achieved.
FRI
R
Uncoupled:
RI
FR12
0
0 0
X 0
FR13J0
0
FDecoupled
XJ
DP11
DP12
13
or Partially Coupled
Rll1r
FR12 =
FR131
DPI
0 0' DP1
DP12
X X 0
X ._ fDP13J
,X
Coupled:
X X
FR12 = X X X
X XF
t
FRllIr5
DP 3
2
DP il FIgure
IPleR131tai
8 DP
R
DP1II
DP12J
DPs13
Figure 8: Implementation Relation between FRs and DI's
27
Production System Design and Implementation in the Automotive Components Industry
3.3.- The Production System Design Framework
Once familiarized with the elements of the Toyota Production System and with a
structured methodology for designing manufacturing systems, namely axiomatic design,
we can better understand and appreciate the Production System Design Framework.
The cornerstone of the Production System Design Laboratory at MIT is the
Production System Design Framework. It encapsulates the various elements that need to
be considered when designing, re-designing, implementing or evaluating manufacturing
systems.
Various attempts have been made at providing a structured approach to tackle the
complex interrelationships and tradeoffs that arise when designing a manufacturing
system. However, some of these efforts have lacked the ability to fulfill the requirements
of a comprehensive framework. Some have failed to communicate how lower-level
requirements affect the overall system performance [Hayes and Wheelwright, 1979].
Others have failed to explain what the means are to achieve higher-level objectives
[Hopp and Spearman, 1996]; others still fail to distinguish the means from the objectives
they are satisfying [Monden, 1993]. Often times manufacturing systems have been
designed by optimizing its various elements in isolation of the overall objectives using a
reductionist perspective as described by [Hopp and Spearman, 1996].
The framework shown in Figure 9 goes beyond previous shortcomings by
presenting tools that encapsulate the high-level thinking that should be incorporated in
lower-level applications. It also describes the interrelation and sequence of steps in which
the various elements should be implemented. Furthermore, the framework provides a tool
for continual evaluation of systems performance [Cochran, 1999]. The framework is
composed of the following elements:
" The Manufacturing System Design Decomposition
" The Manufacturing System Design Evaluation Tool
*
The Manufacturing System Design Matrix, and
28
The Production System Design Framework
*
The Production System Design and Deployment Flowchart and Steps for
implementation.
Of particular interest for the scope of this thesis is the Manufacturing System
Design Decomposition (MSDD). Next section provides an overview of the MSDD. For
further information about the MSDD please refer to [Cochran, Arinez, Duda, and Linck,
2000] and to Appendix A, and for more information about the other elements of the
Framework please refer to [Cochran, 1999].
Design and Deployment Framework
This Framework shows the interrelation between the Design and Deployment of a Production System. To learn more about what
. we do at the Production System Design Laboratory, please visit us at our website: http://web.mit.edu/psd/www
Deployment
Design
Design Decomposition
Design Matrix
Functional Requirements and Design Parameters
Ilustrates relationships
of a Production System
between DP's and FR's
System Design Flowchart
Shows imianntation arecedeice of Desiin Parameters
Deplovment Steps
2.
Mt-w
Dud&
Uafti cbphkfmmikit~ua
Examples
Illustrations of how DP's
satisfy FR's in practice in
different industries
Design Evaluation
Tool
Assessment of how well a PS
.
h
1.11.
rLiiscrtm
bm
pnweta
e-ass.
rl
is designed
Figure 9: The Production System Design and Deployment Framework
3.3.1.- The Manufacturing System Design Decomposition
The design of manufacturing systems is a complex subject that has received
attention significantly. Previous work has attempted to provide frameworks for
decomposing, clarifying or illustrating the interconnection between the various elements
that conform a manufacturing system [Cochran, Arinez, Duda, and Linck, 2000].
29
Production System Design and Implementation in the Automotive Components Industry
However, none of these attempts has comprehensively addressed four key elements for
the effectiveness of a framework [Cochran, Arinez, Duda, and Linck, 2000]. A successful
tool for manufacturing system design should:
1)
Separate objectives from means
2) Relate low level implementation to high-level requirements
3) Explain the interrelationship among the different elements of the system
4) Communicate information across the organization
The MSDD addresses the above requirements. By using axiomatic-based
decomposition, and deriving from previous research and experience in the field, the
framework represents an effective tool to provide a connection between what needs to be
achieved and how it can be achieved [Cochran, 1999].
Although the scope of applicability of the MSDD is at the shop-floor level, the
high-level objective is still to maximize return on investment (ROI). Derived from
equation (1) and following systematic decomposition as described in the previous section,
the FRs at next level are obtained.
ROI = Revenue - Cost
Investment
(1)
The first few levels of the MSDD are shown in Figure 10. The decomposition
process continues until the identified DPs don't require any further decomposition for its
implementation. The complete MSDD V5 is shown in Appendix A.
It is worthwhile to note how the high-level objective gets gradually translated into
lower-level implementable steps moving from general to specific required action. As it
does so, the branches can be categorized in functional areas as seen in Figure 10.
30
The Production System Design Framework
FR1
Maximize
return
long-term
on investment
DP1
Manufacturing
System Design
FRII
FR12
Maximize sales revenue
Minimize production
DP1 1
DP12
to maximize
Production
rou
satisfaction
Elimination of nonadding sources cs
of
customer
to
Manufacture products
target design specifications
Production processes with
minimal variation from
tarj0
FR121
FR122
Meet
Reduce waste in direct
Reduce waste
labor
labor
DP121
DP122
Elimination of non-value
adding manual7tasks
Reduction
lead time
DP113
Mean through
time reduction
Throughput time
reduction
variation
--
to
t
tasks
investment over
lifecycle
Investment
FR13
customer expected
ize
based on
long-term strategy
Deliver products on time
-------
FR-R1
DP13
lue
FR112
DPI12
DP-11
Mii
Irodutirsystemn
I
FR-111
FR 3
costs
inindirect
of indirect labor
a
FR123
Minimize facilities cost
DP123
Reduction of consumed
floor space
-----------
-----
FR-P1
Respond rapidly
production disruptions
Minimize production
disruptions
DP-R1
DP-P1
Procedurefor detection &
response to production
disruptions
Predictable production
resources (people.
equipment,
info
Quality
Jdentii'ing and
resolving
problems
Predictable
Output
Delay
Reduction
Figure 10: First Levels of the MSDD and Schematic Overview
31
Operating
Costs
Production System Design and Implementation in the Automotive Components Industry
Chapter 4: Visteon Indianapolis Steering Gear
Assembly
4.1.- U222 Project
4.1.1.- Background
One of the most significant barriers that arise as Western companies begin to
adopt manufacturing practices pioneered by the Japanese is the accounting system
present in these organizations. They usually fail to capture some or all of the benefits of
lean production, making the transformation from mass to cellular manufacturing appear
sometimes inefficient or unattractive.
Visteon Indianapolis faced this situation when a new business opportunity arose.
The decision to be made was to assemble steering gears for the new U222 project using
their usual high-speed mass-production lines or to implement lean cells. The benefits of
remaining as traditional mass producers were highlighted due to the possibility of
retooling one of the existing assembly lines. The line could assemble the new gears in a
high-speed manner, and retooling it would reduce the investment cost. Also, with the
current accounting system, the main driver to determine the profitability of projects is
labor and consequently the burden of fixed and variable costs. Under the high-speed
approach, these costs would be minimized improving the apparent profitability of the
line.
The assembly line that could be retooled for the new business was the CT 120.
This line produced the steering gears for the Escorts. Since this car was going to be
discontinued, one choice Visteon had was to transform this car's steering gear assembly
line, reusing some of its existing machines, into the new U222 line. This assembly line
would produce the steering gears for the Expedition and the Navigator. The demand for
these cars is expected to peak at approximately 300K gears/year by 2002.
Additionally, the F 150 gears might be incorporated into the U222 program
contributing with an extra 800K gears/year by 2003. The introduction of these gears into
32
Visteon Indianapolis Steering Gear Assembly
the same U222 program was dependent on the decision from the design team of having
rack and pinion (R&P) steering gears in the F 150 trucks as opposed to rotary valve (RV)
gears.
The first choice Visteon faced was to install a high volume assembly line with the
capability of producing the 1. 1M gears/year. The traditional equipment supplier of
Visteon had presented preliminary proposals for this approach. The different production
rate, before and after the incorporation of the F150-related demand, would be adjusted by
the incorporation of additional machines and testing equipment. The second choice is the
incorporation of a lean production system by installing cells incrementally as demand
increased.
This section includes a financial assessment of applying lean principles into the
design of the production system for the U222 program. The goal of this analysis is to
present a financial platform with which lean and mass production systems can be
contrasted and provide a recommendation for the project based on this comparison. The
recommendation covers various levels of scale, ranging from the overall system design,
to the actual assembly cell layout and to conceptual designs for stations and machines
within the cell.
4.1.2.- Net present value analysis
4.1.2.1.- Methodology
In order to perform a net present value analysis of a theoretical project, various
assumptions have to be made. The cash flows of the different alternatives are influenced
by the system in which the gears are produced as well as the type of gear produced (e.g.
size, material, components, etc). However, both the system for the specific U222 project
as well as the U222 gear itself are in the planning stage, and therefore, both of these
factors need to be estimated based on existing information. These assumptions shape the
projections of revenues and costs, which along with investment figures provide a clearer
comparative platform for the two investment alternatives. The margins and investments
are compared across the different alternatives using a net present value analysis.
33
Production System Design and Implementation in the Automotive Components Industry
4.1.2.2.- Assumptions for production scenarios
Given the background and methodology described above, the gears can be
produced in a high-speed fully utilized line, a high-speed underutilized line or in cells.
The cost breakdown for each of these scenarios is here omitted to maintain Visteon's
confidentiality. The cost breakdown was used to obtain the profit per gear under each of
the three scenarios using the following assumptions.
1) A High-speed line fully utilized is modeled from the PN150. This line is used to
predict how the U222 high-speed line would behave if it produced the 1.1 M
gears/year.
2) A High-speed line underutilized is modeled from the CT120. This line is used to
forecast the behavior of the U222 high-speed line during the first year of operation
when it is running at a low volume of 300K gears/yr.
3) Cells are based in the DEW98. This cell is used to project production numbers for the
U222 gears using a cellular layout.
Having obtained the profit per gear in each of these scenarios, a cash flow is generated
for the mass and the lean approach using the appropriate scenarios described above. In
order to generate it, these additional assumptions have been made:
4) The product life for the gears under any alternative will be equal and set to seven
years.
5) The salvage value for the equipment will be 50% of the investment at the end of
seven years. This assumption is based on the salvage value of the existing CT 120,
which can be retooled at the end of the product life for half of its original price. For
cells, retooling should be even less expensive and the salvage value should therefore
be higher, since fixtures and tools are simpler and less specific to the gear. However,
for this analysis this difference was not accounted.
4.1.2.3.- Assumptions for products
Included in the analysis are the following assumptions about the U222 gears:
34
Visteon Indianapolis Steering Gear Assembly
1)
A Navigator gear will have properties (selling price, material, freight etc.) similar to
the DEW98 gear, and the Expedition and the F150 gear to the PN150 gear.
2) The two factors determining the labor cost are the type of gear and the scenario in
which it is being produced. We first look the type of gear in question and then add the
labor cost of each of the departments involved in the production of this gear,
including the final assembly. For this last figure, we use the labor cost of the final
assembly in the scenario that we are analyzing, i.e. high-speed line, high-speed
underutilized line or cells.
3) Fixed and variable allocated costs will be distributed equally to every gear produced
across different scenarios, as opposed to the traditional perspective of assigning them
as a percentage of direct labor. This latter method works well when comparing across
different mass-production assembly lines. However, in lean manufacturing, the fact
that direct labor has increased doesn't necessarily imply that overhead or maintenance
have gone up proportionally. Scrap is one of the variable costs that does vary across
different scenarios and therefore has been taken out of this category to be treated
separately.
4) First-time-through numbers are used to determine the number of "Scrap gears" in the
cash flow section. Since these gears have to be processed again, they'll incur in a
cost. The cost that gets associated with each defective gear is the cost of processing
the gear through the assembly again. This is the total cost minus freight and the
material cost, since it is assumed that the material is reusable in most cases. For the
lean scenario we use a lean plant's numbers.
5) Inventory costs have been included in order to quantify the loss of selling the gears a
few days later as opposed to right away. This cost is calculated as the amount of
money that is not being generated during the number of days that it is sitting in the
floor ready to be shipped. The discount rate is Visteon's internal rate of return of
15%.
35
Production System Design and Implementation in the Automotive Components Industry
The assumptions for production scenarios and for products are used to draw
meaningful data for the cash flow analysis. However, at the end, Visteon's choices are to
implement a cellular approach or to retool their high-speed asynchronous line. Table 1
summarizes the specifications/requirements for these systems as assumed for this
analysis. The selection of capacity used is further explained in the next section.
Table 1: Summary of assumptions and requirements for production alternatives
Capacity (units)
1,100,000
300,000
Investment (M USD)
6.5 (retooling the CT120
line)
2.5 - 4 (depending on
supplier)
Base for cost data
PN150 line
DEW98 cell
Gears to produce
Navigator, Expedition,
F150 (all in same line)
Navigator, Expedition,
F 150 (with dedicated cells)
Projected FTT
90.5%
98.5%
Days of inventory
4.2
1.5
Salvage value
50%
50%
Project life (years)
7
7
TARR (Time Adjusted Rate
of Return)
15%
15%
4.1.2.4.- Capacity planning in cells
When selecting the capacity of each cell, it is necessary to keep in mind some
trade-offs that arise from investment, training, and balancing ability. Naturally, the
shorter the cycle time specified for each cell, the more it will be able to produce and the
36
Visteon Indianapolis Steering Gear Assembly
less number of cells will be required. However, as Figure 11 show, as cycle time
decreases, it becomes more difficult to balance the cell, resulting in layouts that resemble
typical high-speed asynchronous lines, where one operator is isolated to one machine. On
the other hand, as the cycle time increases, also there are more operations that need to be
performed, and consequently, more mistake-proofing devices that need be incorporated.
Based on the experience from Prof. David S. Cochran and the author in various
automotive components plants, a sweet spot has been identified between approximately
30 seconds and 2 minutes. When operating in this range of takt time, the workers are able
to perform various operations and thus maximize their available time. Also, the difficulty
to balance the line and the amount of training and mistake-proofing devices required is
minimum.
Difficulty to
balance the
cell
-
Required
training &
mistake-
Ideal cycle
time region
-proofing
10 sec 30 sec
1 min
3 min
2 min
4 min
Cycle time
Figure 11: Trade-offs and Ideal Cycle Time for Capacity Selection in Cells
[Cochran]
With the above considerations, the number of cells required to meet the total
demand were obtained by limiting the capacity of each of them to be in the ideal cycle
time region. In particular, the takt time for each of them was approximately 42 seconds
working two shifts.
To meet the capacity requirements, one cell would be required to meet the
demand from the Expedition and Navigator gears. Upon the introduction of the F 150
related demand, three more cells would have to be introduced to satisfy the projected
37
Production System Design and Implementation in the Automotive Components Industry
demand. On the other hand, if the high-speed asynchronous line approach is pursued, the
line would have to operate underutilized until the F150 program is approved and the line
can be fully utilized. This situation is illustrated in Figure 12. The dashed line represents
the demand if the F150 program is not introduced. In this case the amount of
underutilization for the high-speed line is very large while the cell accommodates better
to this demand minimizing capacity waste. These observations are quantified in the next
section.
High-speed line
capacity
Volume
(thousand
parts per year) 1100 .OR .t
900 - -
Demand with
F150 program
apa
Excess apacity J
if F150 program
t is introduced
Additional
r
600
excess capacity
if F150 program
is not introduced
60 ..
........-. .
-
300 . .
Cells
capacity
2002
2003
2004
\ Demand without
F150 program
2005
Time
Figure 12: Overcapacity using Cellular and High-Speed Systems
Partly, the reason why there is so much underutilization with the high-speed
approach is because of the lack of dedicated product lines. As Figure 13 shows, this
approach attempts to dedicate one line for all the different customers. The cycle time of
the high-speed line is 11 seconds. This cycle time results from trying to satisfy with one
single line the expected demand from the Expedition, the Navigator and the F150 gears.
Leaving aside the uncertainty of this latter demand, the short cycle time creates enough
complications. As Figure 12 shows, the balancing of this line becomes very difficult and
operators are isolated at each station. With volume fluctuations from the vehicle
38
Visteon Indianapolis Steering Gear Assembly
assembly plant, and with the line producing at a constant rate, the only decoupling
mechanisms are inventory accumulation or capacity underutilization. Both are very
costly, as the next section will show.
Vehicle
Assembly
Steering Gear
Assembly
6C
Repair
Loop
Air k
Leak
c
CT = 60 sec
Functional
Repair
Navigator,
e
CT =140 sec
Ct
sc
______
F5
~.
Sc Repair
3 Bench
A-r -
CT =42 sec
a1h4
C
ec
Sc
C38
c3.~
PartsC
-.
In
Parts
"ft
CT = 42sec
F5
Part
PartOutCT
F5
= 42sec
U222
High-Speed Line
CT =11 sec
Figure 13: High-Speed Approach to Satisfy Vehicle Assembly Demand
On the other hand, by dedicating a line to accommodate the volume from the
Expedition and the Navigator, and the rest to the F 15 0 related demand (Figure 14), a
better fit between capacity and demand can be achieved as shown in Figure 12. Further,
with dedicated lines, quality can be tracked from the final customer to the cell that it
came from, making defect detection easier and less costly. Also, it is easier to
accommodate design changes and additions by having a cell dedicated to a particular
product. Most importantly, given the uncertainty of the F150 related demand, only the
39
Production System Design and Implementation in the Automotive Components Industry
capacity that is required is installed. The additional demand can be satisfied by modular
replication of cells, making the design phase less costly.
Vehicle
Assembly
Steering Gear
Assembly
~
.A
U222 l1
Expedition
CT = 60 sec
--
Navigator
CT = 42 sec
CT =140 sec
C=2 e
C2F150
CT = 42 sec
CT = 42 sec
CT
=
42
sec
LU222Ce14
CT
Info
=
42sec
Part_________F150
CT = 42sec
CT = 42 sec
Figure 14: Cellular Approach to Satisfy Vehicle Assembly Demand
4.1.2.5.- Cash Flow
Based on the assumptions described above and the forecasted demand, a cash
flow was generated. Figure 15 shows an altered schematic representation of the net
present value of the mass and the lean approaches in both the case where the F150
program is introduced and the case where it is not.
The two possibilities regarding the introduction of the F150 program are
presented here. Since the design team still has to decide if it is going to be a rack and
pinion gear or a rotary valve gear, treating this uncertainty as an external variable makes
the comparison of both approaches more objective and reliable. Next section deals with
this analysis and shows the impact of this uncertainty in the NPV.
40
Visteon Indianapolis Steering Gear Assembly
Not producing gears for F150
trucks
Producing gears for F150 trucks
after 2003
Salvage value:
Loss from Expedition
& Navigator gears: Profits from Expedition, Navigator and F150
gears: 3.2M/yr
MASSMA
03
Retooling
CTl20
-04
-55
56s
3.2! iM
07r
58
-,
Retoolin
CTI20
Additional
equipment
1.5M
1.7M
5M~
NPV $5.3 M
IRR
2002
1st
-
20K/yr
NPV $-5 M
IRR -15%
34%/
4.5M/yr
Salvage value:
1.25M
1.7M/yr
5M
1
*
Cell 2.5M4
/
-5.6M/yr
.2002
-5M
first year
CELLS
2.5M
Salvage value:
Loss during
LEAN
Loss from Expedition & Navigator
gears:
o
300K/yr
S2502
LINE
Salvage value:
2nd cell
3rd Cell
4th Cell
's
'03
7
005
'07
Cell
NPV $9.1 M
2002
03
04
03
2.5M
05
57
8
0
NPV $4.9 M
IRR 66%
IRR 46%
7.5M
Figure 15: Cash Flow for U222 Project under Mass and Lean Approaches
The method used to calculate the above values is similar to the method used by
Visteon. However, with this approach, scrap was accounted and a cost penalty was
associated with reworking the part as shown by the simplified equation 2. Similarly,
under this approach, the inventory was included to account for parts waiting at the plant
to be shipped as opposed to shipping them as soon as they are produced. Another
difference with the method used by Visteon is that in the analysis presented here, the
uncertainty of the introduction of the F150 program is considered (Equation 3). On the
other hand, a typical program assessment would leave this variable up to the manager to
evaluate.
Profit/part = SP - (Mtrl+ L + FA + VA + F + RC + IC+ A)
where,
SP = Selling price
Mtrl = Material/part
L = Labor/part
41
(2)
Production System Design and Implementation in the Automotive Components Industry
FA
VA
=
Fixed Allocation
Variable Allocation
F = Freight
RC = Reprocessing Cost (Non - FTT parts)
IC = Inventory Cost
A = Assessment (~14% of costs)
The profits per part and the expected volume provide the income figures. The
capital investment with the corresponding depreciation is then deducted from each years'
cash flow. The expected Net Present Value for the project is calculated by assigning a
probability to the volume demanded. Equation 3 shows a simplified version of this
calculation.
NPV = NPV[(Pr ofit / part * Vol * P(F1 50)) - Investment]
(3)
TARR =15%
where,
Vol = Lumped Volume
P(F150) = Probability of producing F150 gears (assigned only to the
corresponding F150 volume
Investment = Capital investment from machinery and/or testing equipment
TARR = Time Adjusted Rate of Return
Given the above observations, and the shortcomings of the current evaluating
criteria at most traditional mass production plants, some modifications to the current
accounting system need to be evaluated. These would enable financial measures to truly
reflect the performance of systems. Some of these categories include:
42
Visteon Indianapolis Steering Gear Assembly
"
Re-processing of parts
*
Inventory accumulation inside and outside of line
*
Responsiveness
" Flexibility
" Quality benefits
*
Ergonomics and their effect in the operator
*
Savings due to modular replication
* Align operator incentives with corporate incentives - reward system
* Fixed and variable allocation in proportion to resources drawn
* Direct labor based on work performed as opposed to the level of automation
4.1.3.- Analysis
As can be seen from the previous section, cells provide a higher net present value
and internal rate of return than the high-speed line, both at low volumes if the F150 gears
are not produced and at high volumes if the F150 gears are produced. However, as Figure
16 shows, the difference in NPV gets accentuated as we move to lower volumes (towards
zero probability of producing F150 gears). Under this scenario, the low profits of the
underutilized high-speed line running at low volumes don't offset its high investment
cost, and the net present value even becomes negative.
Under the cost structure described in this analysis, the high-speed transfer line
will incur in losses if the F150 gears are not produced. The F150 business would
determine, once the line is installed, whether Visteon would be making a profit or not.
On the other hand, under a lean approach, the cash flow will yield a positive net present
value regardless of the F150 decision.
43
Production System Design and Implementation in the Automotive Components Industry
$10,000,000
$8,000,000
$6,000,000
Lean; ftl
$4,000,000
(L
z
$2 I 000 I000
$0
*r 0%,
W
20%
30 "
506/
60%
75%
%
100%,
-$2,000,000
7
-$4,000,000
-$6,000,000
Probability of producing F150 gears
Figure 16: Sensitivity of the NPV of the U222 Program to the Introduction of the
F150 Program using $2.5M cells
Further, if the cells installed were more costly than the ones included in this
analysis, the trend would still hold. This would be the case if the equipment supplier
increased the preliminary quoted price or if Visteon decided to contract with a different
equipment vendor. Figure 17 shows the same sensitivity analysis by using $4M cells
instead of $2.5M.
$8,000,000
$6,000,000
$4,000,000
$2,000,000
IL
z
$0
0%
10%
20%
50%
30%
-$2,000,000
60%
76%
6/
80%
%
100%
Mass
Assembly
-$4,000,000
1Lne
-$6,000,000
Probability of producing F150 gears
Figure 17: Sensitivity of the NPV of the U222 Program to the Introduction of the
F150 Program using $4M cells
44
Visteon Indianapolis Steering Gear Assembly
In this case, we observe that the net present value of both alternatives converge to
the same level if the F150 gears get produced. However, even a slight uncertainty that
this business will not be introduced harms mass NPV harshly.
Two important conclusions can be derived from this analysis. Within the range of
investment considered of 2.5 to 4 Million USD per cell, two characteristics differentiate
the above curves, one is a shift between the lean and the mass approaches and the other is
their different slope. The shift can be attributed to the fact that, when properly
implemented, cells allow the reduction of inventory. Also, if the proper mistake-proofing
devices and operator training are in place, the amount of scrap and rework can be
reduced. These two categories are quantified and illustrated in the shift between
approaches in the two graphs above. The steeper slope for the mass approach reveals that
the profitability of the line is more sensitive to externalities. In this case the externality
evaluated is the introduction of F150 gears; however, this can be extended to volume
fluctuations and other unforeseen events such as economic slumps or contract
cancellation or non-renewal.
4.1.4.- Recommendation
According to the assumptions made earlier, and as can be seen from the previous
analysis, a lean production system will provide a higher net present value than a highspeed assembly line. This holds true even with a 100% certainty that the F150 program
will be introduced, and assuming the cost of the cells to be much higher than initially
quoted.
In addition, when selecting a production method, other factors should be taken
into account such as responsiveness and flexibility, which were not quantified in the
previous analysis. The cost of retooling the CT120 line has increased more than 25%
from an initial quote due to modifications in the design of the new gear. Due to the high
automation of the machines and the precision required to perform at low cycle times, the
fixtures and tooling used is very specific to the gear in question, therefore making it
costly to adapt to a different gear. On the other hand, cells accommodate this flexibility
45
Production System Design and Implementation in the Automotive Components Industry
by utilizing simpler machines that can be quickly changed over and adapted for new
designs without incurring in outrageous costs.
In terms of responsiveness, the faster throughput time and lower work in progress
of cells (neither quantified in the analysis) provides an environment better suited for
today's customer needs. Customization and quicker lead-time are already playing an
important role for customers when ordering a new vehicle. In order to respond to the car
assemblers the way they will need to respond to their customers, chassis and other major
car components areas will have to quickly adapt to demand fluctuations and
customization, all this with high quality and without having to build large inventories to
remain competitive.
In conclusion, one cell should be built to meet the demand from the Expedition
and Navigator by the year 2002, and upon approval of the F150 program, up to three
more cells should be built incrementally to accommodate the increasing demand from the
assembly plant. Also, the following sections provide a more in-depth description of our
recommendation in order to avoid some of the problems that have been present in
previous attempts to implement lean manufacturing practices.
4.1.5.- Proposed Layout
In order to better convey to the equipment suppliers the principles that the U222
assembly system should embrace according to the MSDD, the author, together with Deny
Gomez and Prof. David S. Cochran developed the proposed layout depicted in Figure 18.
The proposed system includes two parallel rows of equipment. In between these rows, 7
direct workers perform all necessary tasks to assemble the gear in a takt time of 42
seconds. Material replenishment and preventive maintenance tasks take place from the
outside of the line to avoid production disruptions to satisfy FR/DP-T5. The material
replenishment cycle is approximately two hours.
The arrangement of the stations was constrained by requirements imposed in the
assembly sequence. Wherever there was flexibility in steps, the stations were grouped by
off-pallet, manual and automatic sections to reduce wasted walking distance as shown in
46
Visteon Indianapolis Steering Gear Assembly
Figure 18. Also, the width of the aisle and the machines was minimized where possible as
required by FR/DP-D2 1. By reducing wasted motions, labor costs are reduced and
operator-machine separation is encouraged thereby enabling adjustments in production
originated by volume fluctuations.
Fabefian
Test
Banish
AirLmakTast
Manual / Off-pallet
Manual / On-pallet
Automatic / On-pallet
Air~eakTest
Buslins
Yal.I hynt Valve
erim
Piom BeaIrsz
CmU Cecat
Cap
Tunrisiss
Pallet
Relmn
-i P..im Hoii
Rki Bas*sH
Rads
PeAssenAi~EaftIu
hsseztim In
66'
Figure 18: Proposed U222 Steering Gear Assembly Cell Layout
The one-piece-flow layout of the proposed system promotes high quality by
enabling immediate error detection and reducing further the throughput time in
accordance to FR/DP-T4.
It is worthwhile to note that the floor space consumed by this system is
approximately 25 times smaller than that of a high-speed line (WIN8 8), which has a
capacity of producing 4 times as much, representing a real space savings of
approximately 6 times.
Grouping operations by manual and automatic sections allows easier balancing of
the operator work loops. Figure 19 shows the standard work charts developed for this
project. Seven operators are required to run the line to meet a takt time of 42 seconds.
This type of charts is useful in picturing the work sequence of the operators. The
operations are grouped by workers. The manual, walking and automatic time is drawn to
the right of each operation. A horizontal line corresponds to manual time, an inclined line
47
Production System Design and Implementation in the Automotive Components Industry
from one operation to the next represents the walking time from station to station and the
dotted line represents the automatic time. Presenting information in this manner allows
the designer to ensure that all required tasks are performed under the required takt time.
Operators:
TIME
PART: U222 Gear (42 second Takt time)
PROCESS
Man lWalk 1 uio 1
3
1
0
OPERATION
#
10 Grab housing (valve) from container .
21
11
2
6
3
10
01
0
1
0
1
3
10
0
A
2
3
6
7
3
8
1
1
1
2
2
2
1
18
18
3
1
1
2
100 Rotate 90, install valve, Rotate -90, hit switch
240 Load outer tie rods, hit switch
250 Install res. cap, wipe gear, inspect
20
9
10
2
2
2
110 Load input bearing, seal and snap ring, hit switch
120 Install pinion nut, cap, spring and yoke plug, hit switch
7
28
1
1
10
20
0
20jUnload finished housing from Seal press and place into holder
r and el
inbn.unrnno
20 Lonad houising nm
20 Grab finished housing from holder, hit switch
30 Load housing (valve), grab and load housing (tube)
30 Unload housing assembly, hit switch
40 Load housing assembly, load bushings
40 Slide housing, hit switch
50 Grab rack bar from container
60 Load rack bar, load housing assembly
60 Unload gear and clamp on pallet (auto start rack insert)
250 Unload finished gear, hit switch
Pack gear into dunnage
70 Start and torque turnline, hit switch
80 Start and torque tumline, hit switch
90 Load rack bushing, hit switch
10
30
20
0 0
200 Load tie rods &spacer, hit switch
210 Rotate 90, Install boot, clamp and nut, hit switch
9
28
1
1
220 Rotate 180, Install boot, clamp and nut, hit switch
230 Install breather tube, hit switch
28
1
1
1
60
50
40
II
--
-
0
-
0
0
20
0
0
T
0
-
15
..
0
SE-
20
0
0
-
. .-
0
Figure 19: Standard Work Combination Chart for the U222 Assembly Cell
The work loops for this specific configuration with a 42 second takt time are
depicted graphically in Figure 20.
160
150
140
120110100 90
130
60 50 40 30 20 10
80 7(
XUR
0
(D 0
HH'
170
180
190
200 210220 230
UIJ
240
250
Figure 20: Work Loops for U222 Assembly Cell
48
Visteon Indianapolis Steering Gear Assembly
4.1.6.- Lessons leamed from the DEW98 Cell
A previous venture into cellular manufacturing also for R&P steering gear
assembly, the DEW98, has provided Visteon and the PSD Lab with a great source of
knowledge for subsequent generations including the U222 cell. In order to learn from this
experience, the author, together with Deny Gomez and Prof. David S. Cochran identified
flaws in the design and implementation stage to improve future endeavors.
The majority of the problems with the DEW98 cell are a consequence of the
operators not completing their work in a standardized, repeatable sequence, which leads
to delays in production and missed operations, which in turn result in quality issues.
However, the reasons why the operators do not complete their work in a standardized
pattern has more to do with the design of the equipment than with the operators
themselves. The equipment on the DEW98 cell has been designed in such a way that it
presents many ergonomic problems to the operators (long walking distances, difficult
access to parts, protrusions into the workspace, etc.) thus preventing them from
completing their work patterns in a repeatable fashion.
4.1.6.1.- Key Points
*
Operators do not complete their tasks in standardized, repeatable sequences.
*
The erratic nature of the work sequences is the cause of many delays and
production disruptions, as well as many of the quality problems.
*
Flaws in the design of the equipment are the reason why it is difficult and
sometimes impossible for operators to compete their tasks in standardized
patterns.
4.1.6.2.- Analysis
An observer of the DEW98 line can easily notice that the operators do not follow
a standard, repeatable sequence of steps to complete their tasks. It is not uncommon to
see operators work on two gears at one station before they move on to the next station
and complete the operations for those two gears there. It is also not uncommon to see
49
Production System Design and Implementation in the Automotive Components Industry
operators perform tasks that are part of another operator's standard sequence, and even
though cooperation between operations is positive and encouraged, the frequency with
which it occurs in the DEW98 cell suggests that something is amiss. In general, the
operators do not complete a sequence of operations (loop) under the takt time, but instead
the work sequences are very erratic.
It is important to stress how important standardized loops are to the success of the
cellular manufacturing concept. One of the very basic objectives of a cell is to produce at
takt time; i.e. to be able to assemble a final product every time the customer demands
one. And to assemble a product every takt time it is not only necessary that the automatic
stations are able to complete their operations under the takt time, but it is also critical that
each set of manual operations is completed in less than takt time.
It is also important to notice how disruptive it is for the operation of the assembly
cell when the operators do not complete their sequence of steps within one takt time. It is
not equivalent for an operator to complete the steps on one part every takt time and for
the same operator to complete the steps on two parts every other takt time. In the first
case the station downstream from the operator receives a part every takt time, but in the
second case it receives two parts following each other very closely about every other takt
time. In the case that the downstream station is the beginning of a manual sequence of
operations, then this sequence is also disrupted. When the downstream operation is an
automatic station, then a pileup is created and one part must wait for the machine to
finish the previous part so that it can be processed. It must be noted that we are striving
to make the flow of parts through the assembly process like "the flow of water through a
pipe," and therefore any disruptions that prevent such unobstructed flow are detrimental
to the performance of the assembly cell.
In the DEW98 cell operations are not completed in a standardized, repeatable
manner, but as mentioned above, the reason why this is so is because the equipment has
been designed in such a way that it prevents and sometimes even makes it impossible to
do so. Some of the ergonomic problems that the equipment in the DEW98 cell presents
to its operators are:
50
Visteon Indianapolis Steering Gear Assembly
* Machines protrude into the operator's workspace. On some stations the tooling
support pillars protrude to where the operator is supposed to stand (Pinion nut and
cap station.)
*
The spacing between stations and between the two sides of the cell is, in general,
too large. This translates into much time wasted walking (and not working) and it
also discourages the operators to separate from the machines/stations.
*
Finger switches do not encourage motion to the next station. They should be
walk-away switches and should be placed in a standardized location on all
stations to achieve the same operator motion.
* Material supply to the operator was an after-thought in the design, and it was
heavily constrained by the position of the control panels. Materials should be
delivered in a standard location, above the conveyor and in front of the operator,
to facilitate replenishment and standardize the operator's motions. The location
of the electrical panels should be secondary to that.
" The panels to set the pneumatic tools should be given the lowest priority in terms
of placement, after space for the material and for the control panel has been
allocated.
These undesirable characteristics in the design of the equipment are the root cause
for many of the quality problems and delays that the DEW98 cell has experienced. It is
critical that these issues are addressed in the design of future cells. With that goal in
mind the author with Deny Gomez and Prof. David S. Cochran developed a set of
specifications intended to avoid repeating the same mistakes in the design of the U222
equipment which can be found below.
It is worthwhile to note that the ergonomic problems posed by the equipment in
the DEW98 cell is a very important cause for the quality and delay problems being
experienced, but it is not the only cause. The concept of having an operator isolated
outside the cell performing the rack swage operation is certainly preventing attaining the
51
Production System Design and Implementation in the Automotive Components Industry
full benefits of cellular manufacturing (flexibility and balanced loops). The fact that the
DEW98 is attempting to implement lean manufacturing concepts while completely
surrounded in a mass production environment is also an important cause of problems.
The pattern of replenishment of parts to the DEW98 cell, as well as the quality of the
incoming product are both critical aspects of the performance of the DEW98 cell, and
neither of these two aspects has changed from traditional approaches. So, although a
good deal of the problems encountered in the DEW98 cell can be avoided by designing
equipment to achieve different specifications, it is also important that these other issues
are addressed before the full benefits of lean manufacturing are achieved.
4.1.7.- Conceptual Station Designs
Derived from the lessons learned from the DEW98 experience and with the hope
of communicating clear guidelines to the equipment suppliers, the author, together with
Deny Gomez developed conceptual designs for every station in the cell. The intention
was to communicate how the various FRs and DPs from the MSDD that relate to
equipment design would be specifically embodied in steering gear assembly equipment.
Figure 21 highlights the leaf FR/DP pairs that relate to equipment design [Arinez,
2000]. Some of these pairs include:
*
Incorporation of mistake-proof devices
" Design of machine for serviceability
*
Reduction of transfer batch size
* Design of quick changeover for material
* Independence of access for production and maintenance
* Machines designed to run autonomously
* Machines configured to reduce walking distance
*
Ergonomic interface between the worker, machine and fixture
52
Visteon Indianapolis Steering Gear Assembly
*
Minimize facilities cost
*
Minimize investment
Figure 21: FR/DP Pairs Related to Equipment Design [Arinez, 2000]
As shown above, the equipment requirements are derived from systematic
decomposition of the high-level objectives. Therefore, by satisfying these requirements,
the equipment is in turn satisfying the ultimate objectives of the enterprise [Cochran and
Dobbs, 2000].
Figure 22 thru Figure 27 show the conceptual designs for the U222 project, which
attempt to capture the lessons learned from the DEW98 cell and most of the requirements
imposed by the equipment design FR/DP pairs from the MSDD.
53
Production System Design and Implementation in the Automotive Components Industry
Slanted
Housing Assembly Station
y
TubesIn
sea[ Press
Housings (valve)
Station
Processed
Housings
Holder
Housing Unlood
Housing Looding
I~
Llz
Figure 22: Proposed Conceptual Station Designs for the U222 Project
54
Visteon Indianapolis Steering Gear Assembly
Racks In
Rack Insert Station
Bushing Press Station
Containers
In
Containers
Out
LiFt Assist D~evice
Gear
Unloading
Gear
Loading
Retrun
Cart
Walk-away switch
Housings Chute
Pivoted
Unloading
Pallet
Figure 23: Proposed Conceptual Station Designs for the U222 Project
55
Production System Design and Implementation in the Automotive Components Industry
Turnline Station 1
Suspended
Pneumatic
Too(
Turnline Station 2
Parts In
W/Lk-owoy Switch
Figure 24: Proposed Conceptual Station Designs for the U222 Project
56
Visteon Indianapolis Steering Gear Assembly
Rack Bushing
Station
Bushing
tatvin n
,.Valve
Ye/t tinon
Seat
Press Station
Press
Shaft
Spindle
Valves n
Parts
In
Pallet Lift
Walk-mwy
Containers Out
Sliding
Actuator
Figure 25: Proposed Conceptual Station Designs for the U222 Project
57
Switch
Production System Design and Implementation in the Automotive Components Industry
Breather Tube Crimp
Outer
Tie Rod Station
Boot
Insertion
Station
oots
Boot Insertion Station
Tie
Rod Station
Spindle
Containers
Out
Suspended
Tool
Walk-away Switch
Pallet
LiFt
Figure 26: Proposed Conceptual Station Designs for the U222 Project
58
M
Visteon Indianapolis Steering Gear Assembly
Final Inspection Station
Packout
Palet
Return
Cart
Rails built into
the floor
Figure 27: Proposed Conceptual Station Designs for the U222 Project
59
Production System Design and Implementation in the Automotive Components Industry
4.2.- U204 Project
4.2.1.- Introduction
The U204 line began to produce rack and pinion steering gears in June 2000. As
Visteon Indianapolis' second lean implementation, it has provided tremendous learning
experiences to all people involved in it.
At its launching stage, the rate at which problems have appeared may have been
discouraging and frustrating to some people. With increasing demand pressure from the
auto assembly plants, it has been difficult to take the time to analyze what can be done to
fix the line thoroughly and, further, what could be done differently in the next generation
of cells to facilitate the ramp-up stage.
This section attempts to give a comprehensive analysis of the launch state of the
line, a possible short-term approach to improve its effectiveness through labor efficiency,
and a long-term solution that will enable the line to produce as projected.
Furthermore, with the hope of assisting future designs and launches, the
shortcomings present at the launch of this line are analyzed through the MSDD. By
highlighting areas for improvement in a structured framework, the author intends to point
out issues that need special attention in subsequent designs.
4.2.2.- Launch state
4.2.2.1.- Equipment utilization
By analyzing how much manual time and how much automatic time was being
spent at each station, it was possible to identify that, regardless of the number of
operators, the automatic equipment would soon become the bottleneck. At launch,
random sources of variation were the dominant cause preventing the line from running
smoothly. However, when the line stabilizes it should be noticed that the bottlenecks lie
in the automatic equipment, specifically the functional and air leak test as Figure 28
shows.
60
Visteon Indianapolis Steering Gear Assembly
Isolator bushing
Installshipping plugs, final inspection,
unclan
gearaand
p
Install be rod
Install
breather
ends,
tube, crimp tie
Grease
& installtie rod
Grease
& install tie
installation
palltreturn
rod
- -1 -
stamp
pin
boots
-
clamps and jam nuts
boots.
and jam
rod boots, clamp.
S/A Install/torque
i
clips
1I
nuts
tie-rodx
Stake yoke
Effectice functional
test(including
t
and transfer
both
Auto mesh
load
me)
final set
and install
Auto
*
yoke pre-torqna
leak
r
-
Auto burnhrackhntath
Grease
on ap
pin
thtranstanti::
test
Manual
m~niaPaliatnrotatin
Torque
pinion nutand
Inset Tlneman
rack
bushing
Rack
Retenton ring
I&
aI
al
SwaaInag
insertstadon
M*
-
-
S/A rack Insert
Install
Load
turnines
pinion
& check
valve
and
seal
bearing
House
-j
-T
-"
M!!-
loading staton
Housing*
s
t-
:lip
anitr in va
Ina tI cI
Ins
tall
4i
-
S/A Install inputbearing,seal andsnapring
-*00
containers
0
10
30
20
40
50
60
80
70
Time (sec)
Figure 28: Work Content per Station at Launch
Since this is a synchronous line with little de-coupling (return elevator stores up
to 4 pallets), whatever disruptions occur at one side of the line are translated to the other
side by slowing the upstream traffic. Therefore, it is not surprising not to see parts
accumulation uniquely ahead of these tests. The bottlenecks are not necessarily evident at
this points because it is a synchronous line whose disruptions at one station will have
impact as well in up and downstream stations
4.2.2.2.- Work pattern
Faced with increased pressure from the assembly plant on one hand, and the
debugging of problems at the line on the other hand, Visteon had to add more operators
to the line to meet the required output. After launch, the line had been running with 14
operators on average per shift for three shifts, producing an average of 880 parts per day.
However, the excessive number of workers made operator-machine separation
unnecessary, hindering one of the key benefits of lean manufacturing. The work pattern
for the 14 operators running the line at launch is shown in Figure 29. It is evident from
61
*
..
T(~~'~'~.K
-
-
Production System Design and Implementation in the Automotive Components Industry
this figure that the work content for each operator varied drastically, signaling a
necessary rebalancing of the line.
U284 Gear
14
Operatos:
TIME Igoc
PART:
PROCESS
Operator I
-equec
1.(1) Start shortline
1 1(0) Load housing
8
M
aa
2 (1) Return spindle home and load components (pinion bearing and sea) into station
2 (1) Waiting for first part of auto cycle
2 (1) Perform intermediate rotation
(1) Last part of auto cycle
3 (2) Tq. shrt tumline (It), St. long turnline (both), tq. shrt tumline (rght),tq. Long tumline
3 (2) Install check valbs (transfer time included as auto time)
5
37.3
14.25
(both)
8
5.333
10.5
lak test (ncluding
~Illi}
F"fl.-LIHIM 4
HO
If 1
i ILT I
Ii
4.
Ililili
11) tfl
45.17
6
;
H
25
rI~tITnI111111111 rn1rn
'I
4.5
6.5
5.5
.;T1
1MrIi
'ut fl,
'IL
6
4
5.333
2
19.83
transfer time)
14.25
3.75
5.667
5.5
5.333
13.33
2
4.t333
2.667
(11). Load yoke components
(11) Rotate 90 deg,
(11) Press palm buttons to raise pallet
(11) Torque pinion cap
(11) Mark gear with pin
(11) Install yoke
(11) Press palm buttons to lower pallet
(11) Rotate back 9 deg.
(11) Press palm buton to reatease
.............................
....... -222
.............
. ..........
.............
.......
......
..........
-VT
......
..
I..................
.......
. I---................
.............
......
....
-+ ...........
ff it H111
t il 111111,
44
t11i I I !1]
......... .
t ...
ii1
31
load and final set
...................
...............
.J 111 ...
29.5
(12) Auto burnish rack teeth
(14) Mesh
rot
7-
U
1t
26
to lit pallet, tq. Bushing, ins. Bush. for next op., press P.B.
(bA) Raise pallet lit & rotate 90 deg
(EA) Retrieve tools & insert valve. replace tools, P.S. to lower lit & rel
Auto
4
4tt
q4~t
Hi
61
53
SO
d
M
6.5
10.25
(5)Press P.B.
(9)
20
2tt
16.5
6 @B5) Seat clip
6 (1B) Press palm buttons to seat clip
6 (95) Palrm buttons to raise pallet
6 (9B) Torque nut
6 (BB) Apply grease
6 (61) Press patm buttons to lower pallet
6 L)EM Rotate & release pallet
6 (7) Load input bearing, seal and snap ring
7
7
7
7
7
7
7
7
7
10
1o
1.25
17.33
4 (4) Push pallet into rack insertion station, load rack into station, press P.S.
4 (A) Grab rack and assemble ret. ring
4 )B9) Change racks and press PB to swage racks
5
5
Walk
aecAuto0
15.5
10.5
1.....4
if
(16
&7
Functional
Test
16A and 16B combined (including trasfer time)
6B.25
B (20) Insert tie rods into machine
E
(20)
.......* -4
is
(19) Stake yoke
Place plastic clip and cut with tool
B (20) Place travel restrictor and press palm
10
(21) Grease both boot groaes (can be performed as part approaches)
(2i) Retrieve and grease boot (ialf operation can be done after pallet has let and before new begins to approach)
(21) Rotate 70 deg.
(21) Install boot
(21) Insert tinnerman clip, start jam nut, place tinn. Clip with tool, torque jam nut with tool
(21) Rotate pallet back 70 deg.
(21) Press palm button to lower & ret. pallet
(21) Mark gear with pin
(22) Retrieve bong and place it on greaser (can be performed as part approaches)
(22).Rotate paltet 70 deg.
(22) Install boot
(22) Insert tinnerman clip, start jam nut, place tinn. Clip with tool, torquejam nut with tool
(22) Rotate pallet back 70 deg.
(22) Press palm buttons to lower & rel. pallet
(23) Retrieve breather tube & dip (can be done prior to pallet
(23) Install breather tube
3(2)
Crimp tie rod boots clips
(23) KIt tie rod ends to pallet
(23) Press palm buttons to ret. pallet
() Rotate pallet 180 deg. ahead of eirtest
B
1.5
B
23
2.5
0.5
7
.
T .........
.......
..........
............
fill
6
........ ......
44-
T.
...... ..
.
.....
....
T
4!
T
....
.
4-1
4
6
1.5
9.333
6
10.5
6.667
21if
.....
..........
... ........
..........
.......
.. ..
..
..
...... ..
..... ..
i5
5
T
. ..
. T
4
.
.. ...
(24)
(24)
(25)
(25)
(25)
Install left tie rod and
Torque let tie rod end
Install plugs
Mark gear with pin
Unclamp and *hand over" to bushing operator
(26)
(2)
(2)
(26)
(261
Bushings in
Finish loading gear
Press palm buttons (and wait for auto cycle)
Unload, and mark gear with pin.
Pack gear to dunnate
14
10.5
15
. .....
T
T
13
15.5
44i ii
14.5
1.5
13.25
11.55I ...........
1;2'
44.
4.
4 ~9.
'j
rH
i1~j~
1It~ii:
~rL1 Ifl
.~
Figure 29: U204 Work Pattern at Launch: 14 Operators
62
.....
4
. .. . ...
!I
TT TT !
5
6
9
ltt
............
........
--------------
....... .......
29
(24) Place centering tool & raise
(24) Install right tie rod end
(24) Press palm buttons to lif pallet
(24) Torque right tie rod and
(24) Press palm buttons to lower pallet
......
..........
2.5
0.5
B
8.5
10
I
..
.......
8
i.5
orrival)
..
--T - - -T
T
Ii H H
11.75
17.8
.
buttons
I'll.....
V
.......
[LI
Visteon Indianapolis Steering Gear Assembly
4.2.2.3.- Operator utilization
Figure 29 also suggests that since operators had to wait on the slowest cycle to
complete, there was a great amount of idle time. The distribution of value added, nonvalue-added, and idle operations is shown in Figure 30. Again, this is a signal that
pointed into re-balancing the work-loops to decrease the idle time and also to reduce the
time spent at non-value added operations.
Non-value-added activities are understood as necessary activities which do not
add value to the product. For instance, walking, pressing palm buttons, rotating pallets,
etc. Idle time occurs when the operators are simply standing waiting with no part to
process. Among the sources of idle time there is defective incoming parts, lack of parts
and excessively long auto-cycle times. These sources of variation are translated to
adjacent stations through the synchronicity of the line.
80
70
A
60
E> Idle
50
Non-value added
operations
4)%
%"11110
40 A
Value added
operations
30
20
10
0
1
2
3
4
5
6
7
9
8
10
11
12
13
14
Operator No.
Figure 30: Distribution of Work at Launch with 14 Workers
63
Production System Design and Implementation in the Automotive Components Industry
The overall worker utilization is shown in Figure 31. It is worthwhile to notice
that after launch, more than half of the time the operators were performing non-valueadded operations or remained idle.
* Value added operations
* Non-value added operations
-Idle
34%
45%
21%
Figure 31: Overall Worker Utilization at Launch with 14 Workers
4.2.3.- Short-term approach: Improving labor efficiency
As shown above, and as perceived by any observer of the line, there was a great
amount of idle time and non-value added operations at launch. Neither of these two can
be blamed on the operators. In order to understand what is originating these
inefficiencies, we must take a close look at the work content at each station. As seen in
the previous section, some of the automatic equipment is not meeting the takt time, and in
some cases its almost twice as long.
In the short run, since this equipment is limiting the desired output, it is
considered a waste to have excess amount of workers that will not have any impact on the
output. The bottlenecks are still there, and no matter how many workers are introduced
the output will still be constrained by the automatic equipment. In the contrary, excess
amount of workers creates confusion and cluttering inside the line, which unlocks a
frustrating effect.
64
Visteon Indianapolis Steering Gear Assembly
With the equipment limitations in mind, a short-term approach to improve the
effectiveness of the line was proposed. With this, the number of operators and their work
patterns would be adjusted to the capacity of the assembly system. The line would be
producing the same output as at launch but with fewer workers.
In addition to adjusting the workforce and re-balancing the work patterns, some
minor changes should be implemented to achieve the labor efficiency sought in the short
term. These changes are outlined in Section 4.2.3.3.
4.2.3.1.- Work pattern
By rebalancing some operators' work-loops and with some minor changes
described in Section 4.2.3.3 the number of operators can be reduced from 14 to 10. The
work pattern for these operators is shown below in Figure 32. This Figure reveals a more
even distribution of work among all operators.
4.2.3.2.- Operator utilization
In addition to the better distribution of work among all operators, there is a clear
reduction in idle time per worker shown in Figure 33.
Percentage-wise, there is also a significant improvement as shown in Figure 34.
Furthermore, this difference can be better appreciated in absolute terms as Figure 35
shows. The savings from 4 operators per shift translate into approximately 850K
USD/year.
65
Production System Design and Implementation in the Automotive Components Industry
PART: L24 Gear
PRACES
pe-or
1
1
1
1
2
2
2
2
St N OPERATI
(
.as
(-1) Start shortline
(D) Load housing
(D)Install check valve
(1) Load pinion bearing and seal (auto cycle)
d
aa
Operators:
10
liNt lert
ka
TIME
PMCWS Auto
an alk
20
1.72
2.13
(5A Grab rack and assemble ret. ring
(58) Change racks and press PB to swage racks
(4) Push pallet into rack insertion station, load rack into station, press P.H.
(2) Tq. shrt turnline (It), St. long tumline (both), tq. shrt turnline (rght). tq. Long tumline (both)
30
40
0
7
70
BO
_14 ....
------ --------- --
6
45.17
...............
..
3 () Press P.B. to lift pallet. tq. Bushing, ins. Bush. for next op., press P.B.
3 ("A Raise pallet lift & rotate 90 deg
3 (GA) Retrieve tools &insert vale, replace tools, seat clip, P.B. to lower lift &rel.
4
4
4
4
4
4
4
60so
50so
40
3D
20
10
to
a
.
1. 1
t
Press palm buttons to seat clip
(B) Palm buttons to raise pallet
(B) Torque nut
(58) Apply grease
(58) Press palm buttons to lower pallet
(B) Rotate & release pallet
(7) Load input bearing, seal and snap ring
------
...............
.....
..............
......
...
(58)
TA
LW
- ........... ..........
(9) Auto leak test (including transfer time)
5 (11) Load yoke components
5 (11) Rotate 90 deg.
5 (11) Press palm buttons to raise pallet
5 (11) Torque pinion cap
5 (11) Mark gear with pin
5 (11) Install yoke
5 (11) Press palm buttons to lower pallet
5 (11) Rotate back 90 deg.
5 (11) Press palm button to realease
5 (2) Insert tie rods into machine
4,5j~j~jjJj]~**
it I
it
[fL
+4+"
lit
I....., ....... H+
- ...
[ .. ....
. ...
29.5
(12) Auto bumish rack teeth
3'
(14) Mesh load and final set
Be
(16 & 17) Functional Test 16A and 16B combined (including tranfer time)
..
25
. . .. .
L
. . .O
I
NI
-
(19) Stake yoke
(21) Retrieve and grease boot
(21) Grease one boot groove
(21) Rotate 70 deg.
(21) Instaill boot
(21) Insert tinnerman clip, start jam nut, place tinn Clip with tool, torque jam nut with tool
(21) Rotate pallet back 70 deg.
(21) Mark gear with pin
(21) Press palm button to lower & rel. pallet
(21) Place travel restrictor and press palm buttons
1. ...
35. 33
l
I T II
......... .i .
Retrieve boot and place it on greaser (can be performed as part approaches)
(22) Grease one boot groove
(22) Place plastic clip and cut with tool
(22) Rotate pallet 70 deg.
(22) Install boot
(22) Insert tinnerman clip, start jam nut, place tinn. Clip with tool, torque jam nut with tool
(22) Rotate pallet back 70 deg.
(22) Press palm buttons to lower &rml. pallet
(22)
............
HIM
1
Hil I
...............
.......
IIIIII
HIM
.
HIM
I
11
Ili
11
H Hill,
.....
Place centering tool & raise
(24) Press palm buttons to lift pallet
.. ..........
........
.............
.
(24) Torque right tie rod end
(24) Torque left tie rod end
(24) Press palm buttons to lower pallet
(25) Install plugs
(25) Mark gear with pin
.......
... .........
..........
...
+
.....
.......
R2) Bushings in
Load gear from holder
Press palm buttons
Unclamp gear from previous station
Leave gear at holder
Unload finished gear to current unload area and mark gear with pin.
Pack
sear In dunnae
1.
(2)
(2)
P5)
R6)
R2)
a
Hill
Jill
(23) Retrieve breather tube & dip
(23) Install breather tube
(23) Crimp tie rod boots clips
(23) Grab tie rod ends
(23) Install right tie rod end
(23) Install left tie rod end
(23) Press palm buttons to rel. pallet
(8) Rotate pallet 1tI deg. ahead of airtest
(24)
1111
.. I
14.5
IT . ......
44.
-- ,[
-
I.'
Figure 32: Short-term Recommended Work Pattern: 10 Operators
66
Visteon Indianapolis Steering Gear Assembly
80
70
60
50
.
o Idle
) 40
E
* Non-value added
operations
p30
20
* Value added operations
10
0
1
2
3
4
5
7
8
6
Operator No.
9
10
Figure 33: Distribution of Work per Cycle with 10 Operators
34%
45%
* Value added operations
* Non-value added operations
E Idle
21%
11%
Before
(14 op)
27%
62%
After
(10 op)
Figure 34: Distribution of Work per Cycle with 10 Operators
67
Production System Design and Implementation in the Automotive Components Industry
1200
-
1000
31.1 % Labor savings
850k USD per year
353
800
79
600]dl
0Idle
E
P 400
Non-value added
operations
*Value added
operations
200
0
Launch work disribution
Proposed work disibution
Figure 35: Overall Distribution of Work Time per Cycle with 10 Operators
4.2.3.3.- Required action for labor efficiency
In order to make this improvement possible and achieve the aforementioned
benefits, it is recommended to follow the work-loops defined in the previous section in
addition to implementing the changes described in Appendix B.
4.2.4.- Long-term approach
Although the short-term solution described above provides a more economical
way to maintain the original output level, it is desired to analyze a long-term scenario.
Under the long-term approach, the U204 cell should be able to run in a two-shift pattern
to avoid paying premiums and target output must be reached. The number of operators
would result from these other trade-offs.
4.2.4.1.- Takt time calculation
Taking together the demand from both the Mazda and Ford trucks, the required
demand at the assembly plant is expected to ramp-up to 1310 parts per day. Working two
shifts and allowing for a 15% equipment downtime buffer, this translates into a takt time
of 34 sec.
.
'(2shifts (436min
1 day
available time I day , 1shift
1310 parts
total demand
68
60sec
1 min
.
Visteon Indianapolis Steering Gear Assembly
4.2.4.2.- Work pattern
Having calculated a target takt time, the work-loops are again rebalanced to
conform to it. As can be seen in Figure 37, some work-loops fall in the 34-40 seconds
range allowed for downtime. Ideally, if there was no downtime, this arrangement would
be appropriate and target production would still be achieved; but under real
circumstances, because equipment failure occurs, the target takt time should be no longer
than 34 seconds to allow for a downtime buffer.
Once all operators are performing as predicted by Figure 37, kaizening down
some operator's cycle times will be necessary to prevent any equipment downtime from
affecting the overall output.
4.2.4.3.- Operator utilization
In addition to the more even distribution of work among all operators illustrated
in Figure 37, there is a significant reduction in idle time per worker as shown below in
Figure 36. The reduction in idle time waste is illustrated percentage-wise in Figure 38.
mIdle
Non-value
25
E
added
operations
20
Value added
operations
10-
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Operator No.
Figure 36: Distribution of Work per Cycle to Achieve Target Production
69
Production System Design and Implementation in the Automotive Components Industry
Onerators:
PART: UI2tt4 Gear
PART: L1204 Gear
PROCESS
PROCESS
Onp. -#t .I=
o
lit .oN., .* APERATIOl
PRTO
1 (-1) Start shortline (both ends)
1 (0) Load housing
1 () Install check valve
-e
qe
d.,.rts
sdecie
Takt time = 34 sec
uffer for 15% downtime
14
TIME 1ec)
Man iWalk lAuto I
15.5 1.3E
in
4-444.
]+HT I
10.5
6
1.3E
2 (1) Load pinion bearing and seal (auto cycle)
2 (2) St. long turnline (both ends)
14
15.92
3 (3) tq. shrt tumline (rght end), tq. Long tumline (both ends)
3 (5C) Retrieve swaged rack from new station (between at. 3 and 4)
3 (4) Load rack into station, push pallet into rack insertion station, press P.B.
21.92
1
2 1.11
B 2.11 45.17
4 (5) Press P.B. to lift pallet, tq. Bushing, install bushing for next op, press P.B.
4 (5) Seat clip
f!
;;, 'I'
30
s
n
36*4
2
-7
1
4 44444 44
i t 14+
It
fHil
1.10
HTL
hilliTT
1
41)7
llfli
htl4
4
26
4.5
5
26.5
5 (A) Raise pallet lift & rotate 90 deg
5 (A) Retrieve tools & insert valve, replace tools, P.B. to lower lift & rel.
144
1.5
19.83
6
2
1.5
6 (60) Press palm buttons to seat clip and raise pallet
6 (7) Load input bearing, seal and snap ring, press PB to initiate auto cycle
6 (6p) Torque nut and apply grease (simultaneously with both hands)
6 (68) Press palm buttons to lower pallet
6 (66) Rotate& release pallet
.1. .
10.51
11111
iA hll1
4 1-4444
i f lI
(9) Auto leak test (including transfrir time)
ll~1iilI-11
12.25
1.5
2
5.5
10.63
1.5
1.5
1.5
2
7 (11) Load yoke components
7 (11) Rotate 90 dog.
7 (11) Press palm buttons to raise pallet
7 (11) Torque pinion cap
7 (11)_Install yoke
7 (11) Press palm buttons to lower pallet
7 (11) Rotate back 90 deg.
7 (11) Press palm button to realease
7 (11) Mark gearwith pin
3.1
7.
t t. IL
-
34.75 .I
.. ......
(12) Auto bumish rack teeth
I
4.. .14414I
E
4
22 ...
..
(14) Mesh load and final set
(16 & 17) Effective transfer time (counting the two tests)
(16 & 17) Functional Test 16A and 16B combined (including tranfer time)
(19) Stake yoke
8
6.667
23
(20) Insert tie rods into machine and press p.b
6 (20) Place travel restrictor. put strap and press p.b.
9 (21)
9 (21)
9 (21)
9 (21)
9 (21)
9 (21)
9 (21)
9 (21)
8
4
Retrieve and grease boot
Grease one boot groove
Rotate 70 dog.
Install boot
Insert tinnerman clip, st. jam nut, place tinn. Clip with tool. tq. jam nut with tool
Rotate pallet back 70 deg.
Mark gear with pin
Press palm button to lower & ret. pallet
10
10
10
10
10
10
10
(22) Retrieve boot and place it on greaser (can be performed as part approaches)
(22) Grease one boot groove
(22) Rotate pallet 70 deg.
(22) Install boot
(21) Insert tinnerman clip, at. jam nut, place tinn. Clip with tool. tq. jam nut with tool
(22) Rotate pallet back 70 deg.
(22) Press palm buttons to lower &rel. pallet
11
11
11
11
11
(23)
(23)
(23)
(23)
(23)
Retrieve breather tube & dip (can be done prior to pallet arrival)
Install breather tube
Crimp tie rod boots clips
Kit tie rod ends to pallet
Press palm buttons to rel. pallet
12
12
12
12
12
(24)
(24)
(24)
(24)
(24)
Place centering tool & raise
Press palm buttons to lift pallet
Torque right tie rod end
Torque left tie rod end
Press palm buttons to lower pallet
1.5
(26) Bushings in
(26) Unload finished gear from st. into current unload area, load new gear from holder
(26) Press palm buttons
(26) Mark finished gear with pin.
Pack sear to dunnage
II I if III
Miltl
6
14.4
1.5
2.5
0.5
6
4
1.5
6
14.4
1.5
0.5
6
6.5
10
6
1.5
ii
ft
T
6
6.667
10
10
5
KnftI ItI-H
1114141
111
5
13 (25) Install plugs
13 (25) Mark gear with pin
13 (25) Unclamp gear and leave it at holder at next station
14
14
14
14
14
14
1
1
11
+
Pq~fbTh
~
~
~
I1
Ti~
.....111
...
....
[4 [1 IH
Il'
6
15
iI
21
14.6
2
Figure 37: Work Pattern to Achieve Target Output with 14 Operators
70
i
THITI4Hm114141114
if41 i
15.5
1.5
4
7
-H
uifi
iH
It
i1
1
r
1~~h
Visteon Indianapolis Steering Gear Assembly
34%
* Value added operations
0 Non-value added operations
o Idle
1045%
9%
21%
-.00
Before
(14 op)
27%
64%
After '
(14 op)
Figure 38: Overall Distribution of Work per Cycle to Achieve Target Production
In absolute terms there is also a reduction in time spent at non-value added
operations as Figure 39 shows. Taking together the time savings from non-value-added
operations and idle time, the cell can become almost twice as effective. This is achieved
by reducing the average takt time per person by approximately 36 seconds as shown in
Figure 40. By implementing the steps outlined in the following section and following the
work-loops defined in the previous section, wasted time can be cut in half.
1200
1000
48%Cell
Waste
800
Reduction
.
600
] Idle
400--
* Non-value added
operations
200
*Value added
operations
0
Launch work distribution (14 operators)
Proposed work distribution (14 opertators)
Figure 39: Overall Distribution of Work Time per Cycle to Achieve Target
Production
71
~ IAi~T
~
-~
Production System Design and Implementation in the Automotive Components Industry
70
Possible takt
time
60
reduction =
36 seconds
040
Ef
Avg idle time
20-
20
Avg time on Value
Added
Ops
10-
Avg time on NonValue Added Ops
0
Launch
takI time (14
operators)
Proposed takt time (14 operators)
Figure 40: Average Takt Time per Worker at Launch and with Proposed
Improvements
It is worthwhile to note here that this long-term solution is based on the same
number of workers that originally ran the line at launch. The savings originate from fact
that less time is consumed in producing the same output, therefore enabling the line to
achieve target production.
4.2.4.4.- Required action to achieve target output
To attain the savings mentioned above, some issues need to be addressed.
Appendix C describes in detail the required station-by-station changes. In general, these
include:
" Auto cycle times have to be within takt time including manual work.
* Continuously pressing palm buttons is over-killed and makes man-machine
separation impossible at short takt times
*
Poor material handling and station ergonomics prevent smooth operator-part
interaction and discourage loop work
*
Sequential, standardized work sequences have to be defined to produce
consistently according to takt time and ensure high quality
72
Visteon Indianapolis Steering Gear Assembly
4.2.5.- Analysis with the MSDD
Many of the pressing problems that occurred at launch can be traced to weakly
satisfied FR-DP pairs from the MSDD. As can be seen in Figure 41, a quick glance at the
non-satisfactory shaded FRs reveals great potential for improvement.
Most of the corrective action given the launch state of the line was proposed in
the previous sections. However it is far more economical and effort efficient to analyze
what wasn't implemented well at the launching state and ensure proper conformance to
the lowest level FRs in future cellular generations. Table 2 presents the leaf FR-DP pairs
that were not satisfied at the launch of the U204 cell.
7
0
[
0
[10
Predictable
Output
Delay Reduction
Operational
Costs
n
10
Quality
000000
Identifying
and
[J
Investment
Resolving
Problems
U
Leaf FR's fully satisfied.
Leaf FR's not satisfied.
Figure 41: Overview of Unsatisfactory FRs at the Launch of the U204 Cell
Table 2: Unsatisfactory FR-DP pairs at the launch of the U204 line
FR/
DP
MSDP
MSFR
Q11
Eliminate machine assignable causes
Failure mode and effects analysis
Q121
Ensure that operator has knowledge of required tasks
Training program
73
Production System Design and Implementation in the Automotive Components Industry
Q122
Ensure that operator consistently performs tasks correctly
Q123
Ensure that operator human errors do not translate to defects Mistake proof operations (Poka-Yoke)
Q14
Eliminate material assignable causes
Supplier quality program
Q2
Center process mean on the target
Process parameter adjustment
R121
Identify correct support resources
Specified support resources for each failure mode
R122
Minimize delay in contacting correct support resources
Rapid support contact procedure
R123
Minimize time for support resource to understand disruption
System that conveys what the disruption is
R13
Solve problems immediately
Standard method to identify and eliminate root cause
P121
Ensure that equipment is easily serviceable
Machines designed for serviceability
P131
Reduce variability of task completion time
Standard work methods to provide repeatable processing
time
P133
Do not interrupt production for worker allowances
Mutual Relief System with cross-trained workers
P142
Ensure proper timing of part arrivals
Parts moved to downstream operations according to pitch
TI
Reduce lot delay
Reduction of transfer batch size (single-piece flow)
T221
Ensure that automatic cycle time <= minimum takt time
Design of appropriate automatic work content at each station
T222
Ensure that manual cycle time <= takt time
Design of appropriate operator work content/loops
T23
Ensure that part arrival rate is equal to service rate
Arrival of parts at downstream operations according to pitch
T4
Reduce transportation delay
Material flow oriented layout design
T52
Ensure that production activities don't interfere with one
another
Ensure coordination and separation of production work
patterns
T53
Ensure that support activities (people/automation) don't
interfere with one another
Ensure coordination and separation of support work patterns
D11
Reduce time operators spend on non-value added tasks at
each station
Machines & stations designed to run autonomously
D12
Enable worker to operate more than one machine / station
Train the workers to operate multiple stations
D21
Minimize wasted motion of operators between stations
Configure machines / stations to reduce walking distance
D23
Minimize wasted motion in operators' work tasks
Ergonomic interface between the worker, machine and fixture
D3
Elminate operators' waiting on other operators
Balanced work-loops
123
Minimize facilities cost
Reduction of consumed floor space
13
Minimize investment over production system lifecycle
Investment based on a long term system strategy
74
Standard work methods
I
-
-, -
-.
Visteon Indianapolis Steering Gear Assembly
Most of these unsatisfactory FRs can be addressed by incorporating the
improvements outlined in Appendix C. As stated there, each of the improvements can be
related to a weak FR. Table 3 summarizes the state of the U204 line at launch as
classified by the different areas of the MSDD. It also states the possible state it can
achieve when implementing the outlined station-by-station improvements.
Table 3: Unsatisfactory FRs at Launch and with Proposed Long-Term Approach
Leaf FRs
Launch
Afterchanges
Quality
Identifying and resolving problems
Predictable output
Delay reduction
Operations cost
Total
6 of 9
4 of 7
4 of 8
7 of 12
7 of 10
28 of 46
1 of 9
2 of 7
1 of 8
4 of 12
1 of 10
9 of 46
With the proposed physical changes all areas from the MSDD can be improved as
can be seen in the above table. However, there are some changes that require surrounding
areas to the cell to be corrected. For instance in order to reduce defective incoming
materials and improve the quality, the flow where parts are coming from must be
traceable. The operators have to have knowledge of the required assembly tasks. In order
to improve other areas like delay reduction and identifying and resolvingproblems,
standard work routines should be designed and implemented. Also, whenever disruptions
occur at any station, a standard procedure for rapid identification and correction of
problems should be outlined. Equipment must be specified to operate within takt time
accounting for both manual and automatic time. To fully utilize labor, and to improve the
operationscost area, operators must be cross-trained and capable of operating multiple
stations. Also, the work loops should be balanced to minimize wasted time by enabling
worker relief when standard work disruptions occur. Some of the equipment is
unnecessary (return automated chute for tie-rod containers) and costly. As part of the
predictableoutput area, it is required to have a motivated work force performing standard
work; however, with unbalanced loops, equipment malfunctions and ergonomic
problems, this is difficult to achieve.
75
Production System Design and Implementation in the Automotive Components Industry
4.2.6.- Conclusion
Given the nature of cells, there is not much decoupling between stations to
dampen production disruptions when they occur. Every source of disruption has an effect
on the output of the line. It is due to this exposure of problems that waste can be
eliminated. Thanks to the "fragility" of the cell, no defective parts are being handled,
transferred, processed, nor stored.
Evidently, these benefits are hard to appreciate at the launching stage of a cell.
The debugging curve faces a much steeper slope at the beginning than a high-speed
transfer line. The high-speed lines with rework loops and parallel processing have the
"ability" to hide defects, and surely they are capable of handling rejects but there is a cost
associated with that.
In conclusion, the U204 line can be brought to the desired stage by implementing
working solutions as the ones describe above, and defining and implementing standard
working practices. For upcoming cellular ventures, it is recommended that a structured
design approach be followed. Particularly, the MSDD is suggested here as a means to
identify what needs deeper attention in the coming generations. Conformance to the
lowest-level FRs of the MSDD guarantees satisfaction of the highest-level goals of the
enterprise as a whole.
76
Visteon North Penn Electronic Engine Controller Manufacturing
Chapter 5: Visteon North Penn Electronic Engine
Controller Manufacturing
5.1.- Introduction
This section includes the work performed by the author together with Carlos
Tapia and Prof. David S. Cochran at Visteon's North Penn electronics plant, located in
Lansdale, Pennsylvania.
The production system analyzed at North Penn manufactures Electronic Engine
Controllers (EECs). These devices are responsible for engine control functions generally
described as power train control. Basic functions controlled by the EECs include spark
timing control (what a distributor and timing belt used to do), transmission control, and
engine management (air-fuel mix and diagnostics).
The manufacturing process consists of four stages: Two Surface Mount Device
(SMD) processes, lamination, and packing. During SMD processing different
components populate both sides of a circuit board. At lamination, the populated circuit
board is assembled with the connector and the casting case; here also the software is
programmed into the board. During packing, automated transfer lines and robotic arms
prepare ready-to-ship products.
The Visteon North Penn Electronics plant is particularly interesting for our
analysis as they perform the lamination of the EECs using two very different approaches.
One is a traditional asynchronous high-speed transfer line and the other is a synchronous
lean cell.
The scope of this project is partly to analyze the material and information flow of
this product as it is processed through these different systems. Also, by using the
Manufacturing System Design Decomposition, a thorough assessment of the two
lamination methods is made. The results of this analysis are compared to the evaluation
of these two systems using traditional performance metrics. By using the MSDD
evaluation method, we are able to point out potential areas for improvement at the
77
Production System Design and Implementation in the Automotive Components Industry
implementation level of the cell, which are not immediately obvious from just observing
performance results.
Furthermore, by using the Equipment Evaluation Tool [Gomez, Dobbs and
Cochran, 2000], which was derived from a subset of the MSDD related to this area, an
analysis of some of the stations at the two systems is performed. This shows how
equipment designed with a systems perspective leads to improved overall system
performance.
5.2.- Material and Information Flow
5.2.1.- Material Flow
North Penn produces 200-300 different types of EECs to accommodate the many
different variations of Ford vehicles in North America. They also supply the controller to
other vehicle manufacturers, such as Mazda. The process starts with about 20 different
circuit board configurations that constitute the different product family groups. These
boards are populated with electronic components in different patterns to make up to 60
different types of ready-to-assemble boards (approximately 3 patterns per family group)
as shown in Figure 42. Each one of these 60 populated boards represents a product
family. At lamination, the boards can be programmed differently to form the vast
variation of EECs.
78
Visteon North Penn Electronic Engine Controller Manufacturing
Lamination
SMD
Family Group
(20)
Families
(60)
Final products
(200-300)
Figure 42: EEC production steps
5.2.1.1.- SMD Processes
During the SMD processes, the electronic components are mounted to a Printed
Circuit Board (PCB). The SMD process lines are grouped in pairs dedicated to group
families. Each of these transfer lines populates either side, top or bottom, of the PCB.
Figure 43 shows a schematic representation of one of these lines. The transfer lines are
configured to run autonomously with the exception of manual loading and unloading.
Three operators are in charge of these tasks, as well as performing changeovers,
replenishment of material and supervision of the automatic machines.
Parts that have been populated on their upper side are directly transferred to the
contiguous bottom process line located across the hall. Between the two transfer lines,
there is a small amount of inventory. The Work-In-Process (WIP) between the two lines
can be up to a half day of production (472 parts for each line pair, or about 5000 parts in
total).
79
Production System Design and Implementation in the Automotive Components Industry
The operating pattern for these lines is three shifts of 8 hours. There are two 15min. breaks and a lunch break of 30 min. per shift. This leaves a total of 21 available
hours in a day.
A schematic representation of a generic SMD transfer line is shown in Figure 43.
The boards are stacked manually in a loading station at the beginning of the line. They
are automatically transferred one by one to the conveyor that transports them through the
different processes in the line. The first process is the application of solder material. The
material is applied to the PCBs in a printer-like fashion according to a pattern that
matches the electric contact points of each specific board type. The next step is visual
inspection of the solder pattern: measuring whether the right amount of solder paste has
been applied at the right location. This is done at the automatic visual tester.
M
Machineven
Automatic
PCB Unloading
SMD)
Machine
Manual
PCB Loading
Tester
Figure 43: SMD Process Sequence
After the inspection, adhesive is applied to the PCBs to prepare them for
subsequent chip placement at the next station. The adhesive is needed to hold the main
chip in place due to its size and weight. The other electronic components are held in
position by the solder paste previously applied. With the solder and adhesive in place and
after successful testing, the boards go through the first SMD machine. The function of the
SMD machine is to place the different electronic components required for a particular
product. These components are supplied in reels, which are mounted on the SMD
machines.
80
Visteon North Penn Electronic Engine Controller Manufacturing
Most of the SMD machines used at North Penn have reels on both sides of the
conveyor and there is a dispenser hand for each reel location. As the part moves through
them, each hand inserts the components as needed in a process termed "board
population".
Once populated, the boards go through the flux oven, which melts the solder paste
to create a solid electrical connection. The average throughput for this process is
approximately 20 minutes. Finally the boards are automatically unloaded into containers
of 27 parts. These containers wait until an operator transfers them across the hall, to
initiate a similar process in the bottom side of the PCB.
5.2.1.2.- Lamination Process
The populated boards can be routed through two different lamination processes.
The first includes two high-speed transfer lines (cycle time: 10 sec.) and the second
consists of a "lean" cell (cycle time: 50 sec). Each transfer line has an annual capacity of
1.6 million parts, whereas the cell can only produce 350,000 parts per year. Their relative
size comparison is shown in Figure 44. At a first glance, there is an evident difference in
the two systems, as can be seen in Figure 45 and Figure 46. The subsequent sections
herein will quantify these differences and provide a rationale for a more sound system
assessment.
Transfer line process sequence
Most of the EECs at the plant are assembled in one of the two high-speed transfer
lines. Figure 45 shows a diagram of the process sequence and Table 4 describes it.
81
Production System Design and Implementation in the Automotive Components Industry
a to
Am
gl 1sqt
--
-
'A-.=
Transfer line
Cell
Figure 44: Relative Size Comparison between Cell and Transfer Line
CaConttnues...
100
... .,,.tiue..... Contnuing
4)
7}VTouftmgk
SC1)
AOoema6
ve
0)aBarm-O&
Soldet~andwaaje
~
Figure
rnfrlielyu
~
45:r
~
~
82
~
zw
cooain
Visteon North Penn Electronic Engine Controller Manufacturing
Table 4: Transfer Line Process Steps
1. Automatic unload of PCBs from their containers and into the conveyor.
2. In-circuit test of the PCBs. In the future, this step will take place at the SMD process lines.
This step is required to ensure that the necessary components are put in place.
3. Mating of the PCB with the prepared casting-connector. Previous to this point, an incoming
branch delivers the assembled casting-connector. The lamination material has also been applied
to this subassembly in preparation for its mating with the PCB.
4. Screw board to casting. A six-spindle fully automatic screwing machine fastens both
components.
5. Solder connector to board. A rotating machine picks up the boards and dips the connector in a
curtain of molten solder.
6. Bar code reading to identify the PCB family
7. Solder and warping visual inspection
8. Voltage-stress-test station
9. Conformal coating (dip in silicon and oven curing)
10. Bottom plate placement and screwing
11. Gasket application
12. Bum-in process to induce failure of weak components
13. Unload
"Lean" cell process sequence
The "lean" cell represents an innovative process alternative at North Penn for the
lamination of the PCBs. The EECs produced by this cell are mainly supplied to Mazda.
Although the product is the same as that produced by the transfer lines, the process in the
cell, illustrated in Figure 46, and described in Table 5, slightly varies from the transfer
line sequence described above. The variations in the process were introduced to satisfy
special customer requests. The flexibility of the cellular approach allowed these
modifications.
Table 5: Lamination Cell Process Steps
1. Laminate castings. A similar process to the transfer line is used. The larger cycle time of the
cell allows slower operation and reduced machine complexity.
2. Visual inspection for good adhesive beads
3. Assemble connector to casting (Casting/Connector Assembly Station)
4. Place PCB onto casting subassembly
5. Screw-down of PCB into casting subassembly. The larger cycle time of the cell allows a twospindle machine to be used as opposed to the six-spindle machine at the transfer line.
6. Program input
7. Solder connector to board
8. Visual inspection of solder
9. Native-mode test
10 Reset test
11. Ambient test
12. Dip coat and curing (conformal coating)
83
Production System Design and Implementation in the Automotive Components Industry
13. Placement of bottom and top cover plates into subassembly
14. Screw-down cover plates. Same machine as step 5
15. Cold chamber process (batches of 20 parts)
16. Dry in oven
17. Hot fmal test
Notice that there is no bum-in process. After the hot final test, the EECs are
transferred to the transfer lines to be processed at the bum-in oven. They are incorporated
into the line just before the gasket application station, at the discretion of the line
workers.
The cell requires three operators but can be operated with one, two or more
operators as volume changes. The standard work routines for the three-operator
configuration are shown in Figure 46. The operating pattern for these lines is two 8.5hour shifts and one 7-hour shift, with minimal overlap.
12) Dip
Coater
Incoming
bottom plates
(10'x 20')
14) Bottom plate
Screwdown
9,10) Testers
40 ft
cmc
15) Refrigerators
+
10 f
7,8) Solder
E:1 FPot
6) Programmer
5) Board
Screwdown
16) Dry in oven
17) Hot Final
Test
-
E
l
Incoming
Castings
4) Incoming
Boards
1) Laminate
Dispense
2,3) Incoming
Connectors
30 ft
-
*
Figure 46: Lamination "Lean" Cell Layout
84
Visteon North Penn Electronic Engine Controller Manufacturing
Some of the benefits of the cellular approach can be already appreciated from
Figure 46. The fact that operators perform their tasks while walking as opposed to sitting
is better from an ergonomics point of view. By working in small teams, people tend to
develop a sense of ownership for the parts they build, improving the morale in their
environment. Also, as the following sections will show, the equipment is more easily
accessible promoting cross learning and enabling better balancing to accommodate
fluctuating volume requirements.
5.2.1.3.- Packing
Once the EECs are fully assembled, they are automatically unloaded from the
lamination transfer line. Then they proceed through the system of conveyors into final
packing, where a robotic arm prepares the boxes ready to be shipped. Each of these boxes
contains 18 EECs.
Some boxes are sent directly to the staging area to be shipped and others are held
in inventory for some days at the Automatic Storage / Retrieval System (AS/RS). WIP in
the pack area is about 600 units in the automated line and 750 modules in the AS/RS.
5.2.2.- Information Flow
5.2.2.1.- Scheduling
The scheduling and planning of production is supported by the software package
Rhythm TM [i2 Technologies]. This software takes existing orders and assists in leveling
production by volume as well as by mix. The plant can rely on a month of solid orders
made by the customer to plan its production. RhythmTm assumes an infinite capacity plant
and it is up to the production scheduler to level production with the aid of the software.
The production is scheduled at two points: SMD first pass (top) and lamination as
shown in Figure 47. The reason for scheduling at SMD is because it is perceived as the
most constrained process. The setup times have a significant impact on the capacity of
the lines. For scheduling purposes, the throughput time from SMD to packing is assumed
to be two days. Information on volume and mix is sent to the first SMD line. At this
point, the information about the mix is only specific to product families; final product
85
-~
--
1F~LfL
--
-
-
-
Production System Design and Implementation in the Automotive Components Industry
variations are not yet determined. Products flow in a FIFO manner through the rest of the
downstream operations. When the EECs reach lamination, the bar code is read to
determine the product family type. With this information, the specific software is
programmed into the EEC according to the production schedule determined by
JyfiTM
M.
Rhythm
Purchase orders for the supplied materials are produced using an MRP system.
The interaction between the production processes, suppliers and scheduling center is
represented in the value stream mapping in Figure 47.
Rhythm. MRP ProductIon Planing and
Planned Vehicle Assemnby
Requirements (Weeldys Schedule)
J *-
Schaeing Systemn
2l 0P
-
L=*i w
LamlnatloiV
t
0ed
Wvh
Asewnbly
SMD
TSMttnie
32 we
SMD
Taktdme:
32 se
Figure 47: Value Stream Map of the EEC Production
5.3.- Lamination Analysis
5.3.1.- Observed performance at lamination: transfer line and
cell
In order to better appreciate the difference in performance between two
lamination systems, several categories were quantified and summarized in Table 6. These
values, normalized by capacity, were obtained from the observed performance of the two
86
-
-
Visteon North Penn Electronic Engine Controller Manufacturing
systems. It is interesting to note that even when the cell requires more direct labor, it
outperforms the transfer line in all the other metrics considered.
Given that traditional accounting systems in mass production plants strive to
reduce direct labor, most of the other benefits that cellular manufacturing promotes are
often overlooked. Table 6 attempts to illustrate this point. Assuming that we give all
categories a similar weight, we can then calculate the average for all these relevant
metrics. The results show that, by minimizing required resources and reducing waste, the
performance of the cell is on average 63% better than the performance of the transfer line
Even when this evaluation method suggests that the preferred production
approach is the cellular one, there is not enough information that can be derived from
these numbers to improve the performance of the system. The next section takes a
different approach to analyze the performance of the two systems. By using the
Manufacturing System Design Decomposition, potential improvement areas at the
implementation level are identified.
Table 6: Observed performance at the lamination transfer line and cell
Floor Area (sq. ft.)
1
1.37
WIP within Lamination
1
1.02
Throughput time (hrs)
1
2.33
Capital Investment (M)
1
1.57
Direct Workers
1
0.44
Indirect Workers
1
2.19
Defects (assignable to lam. process)
1
2.50
Average
1
1.63
Good Parts/labor-hour (w/indirect labor)
1
0.76
Capacity
1
1.00
87
Production System Design and Implementation in the Automotive Components Industry
5.3.2.- Analysis of Lamination Processes using the MSDD
In this section, an analysis of the lamination processes is performed using the
MSDD. This methodology allows the identification of potential areas for improvement.
The justification for the use of the MSDD as a design tool can be reinforced by tracing
the degree of conformance of each system to the MSDD. The conformance to the MSDD
can be compared to the performance of each system as defined by traditional metrics
described in the previous section, which usually define system performance.
5.3.2.1.- Evaluation of the High-Speed Lamination line using the MSDD
This section shows the degree of conformance of the high-speed lamination line
used at North Penn to the MSDD. The process of evaluation is to consider only the leaf
FRs as shown in Figure 48. The reason for evaluating only these FRs is that it is
sufficient to show that one leaf FR is not satisfied to show that the parent FR is not
fulfilled. Also, since these FRs are at the lowest level in the MSDD, they can be easily
evaluated because an implementable DP can be assigned to them.
L
0 00
Quality
[
OO
Identifying
and
Resolving
Problems
0M M
Predictable
Output
000
00
Delay Reduction
Operational
Costs
Investment
Leaf FR's fully satisfied. Grade: 1.
I
Leaf FR's not satisfied. Grade: 0
Figure 48: High-Speed Line Evaluation Using the MSDD
88
Visteon North Penn Electronic Engine Controller Manufacturing
We use a grade of 1 to represent an FR that is fully satisfied and a grade of 0 for
an FR that is weakly or not satisfied at all; the grades are shown schematically above in
Figure 48. The conformance to the MSDD FRs by areas is summarized in Table 7
5.3.2.2.- Evaluation of "Lean" Cell Lamination system using the MSDD
The methodology used to assess this system is the same as that used to evaluate
the high-speed transfer line. The satisfaction of the leaf FRs of the MSDD is shown
schematically below in Figure 49. The conformance to the MSDD functional
requirements for the cell is also summarized in Table 7.
By using this analysis, we can observe where the system can be improved. Due to
the nature of the MSDD, the leaf FRs can be traced to implementable solutions.
Therefore, special attention can be paid to low performing FRs. The next section outlines
the low performing leaves based on this approach for present and future cellular
implementation improvements.
E
Quality
Identifying
and
Delay Reduction
Predictable
Output
Operational
Co sts
Resolving
Problems
Leaf IFR's fully satisfied. Grade: 1.
Leaf FIR's not satisfied. Grade:
0
Figure 49: Lean Cell Evaluation Using the MSDD
89
Inve stment
Production System Design and Implementation in the Automotive Components Industry
Table 7: Achievement of MSDD leaf FRs at Lamination
T.L.
FRsMiettfon MoldIn
Quality
Identifying and resolving problems
Predictable output
Delay reduction
Operations cost
Total
3 of 9
1 of 7
4 of 8
2 of 12
1 of 10
11 of 46
"CeI
5 of 9
3 of 7
8 of 8
10 of 12
9 of 10
35 of 46
5.3.3.- Recommendations for cellular implementation derived
from the MSDD
Based on the experience of this first cellular implementation in the lamination
area, and using the MSDD-based analysis from the previous section, some
recommendations can be made for present and future cellular ventures as well as for
overall plant design.
It is first worthwhile to note that, in comparison with Figure 48, the schematic in
Figure 49 reveals a design meant to satisfy overall system goals. This is an indication that
the designer of the cell maintained broad systems thinking while trying to incorporate key
components of a world-class production system.
However, based on the MSDD analysis, some leaf FRs were not satisfied.
Identifying these FRs gives valuable information for improving the performance of
current and future systems. Due to the nature of the leaf FRs, a corresponding DP can be
implemented and tracked to them. The FR/DP pairs that received a grade of 1, and
therefore the ones that provide room for improvement and attention, are outlined in
Table 8.
Table 8: Low Performing FR/DPs at the Lamination Cell
FR/DP
Q11
DP
FR
Eliminate machine assignable causes
90
Failure mode and effects analysis
Visteon North Penn Electronic Engine Controller Manufacturing
Q14
Eliminate material assignable causes
Supplier quality program
Q31
Reduce noise in process inputs
Conversion of common causes into
assignable causes
Q32
Reduce impact of input noise on process Robust process design
output
R113
Identify what the disruption is
Context sensitive feedback
R121
Identify correct support resources
Specified support resources for each
failure mode
R122
Minimize delay in contacting correct
support resources
Rapid support contact procedure
R123
Minimize time for support resource to
understand disruption
System that conveys what the disruption
is
T31
Provide knowledge of demanded
product mix (part types and quantities)
Information flow from downstream
customer
r32
Produce in sufficiently small run sizes
Design quick changeover for material
handling and equipment
I2
Eliminate information disruptions
Seamless information flow (visual
factory)
In order to address some of these low performing FRs, it is necessary to zoom out
to analyze the plant again with a broader perspective. As can be seen in Table 8, some
problems with traceability of defective incoming parts, information disruptions, and
repairing procedures still remain. Although the lamination cell was able to improve
dramatically on its predecessors as Figure 48 and Figure 49 show, the current plant's
value stream mapping reveals potential areas for improving from a systems perspective
on these low performing FRs.
Figure 50 shows a modified version of the current value stream map at North
Penn. The fundamental difference is the alignment of SMD with lamination. With this
proposed mapping, a lamination line can be dedicated to each SMD pair. This can be in
turn dedicated to a particular product or customer. Doing this eases defect traceability and
correction. Also, it presents greater flexibility to accommodate changes in product design
91
Production System Design and Implementation in the Automotive Components Industry
or in customer requirements. Further, by laying out the plant in such a way, scheduling
can be simplified. As Figure 50 shows, only one scheduling point is required at
lamination, thereby enabling a pull production system which allows inventory and cost
reduction.
Asembly
Vehicla
(WaftklShd~ule)
S Panned
Rtcqukrnments
KW
Pull Producian Planing and
Schedulng System
ZZNW
b~ iF7r
VW~
SMD
SMD
Takt tim
32 ec
Takt Um:
32 ec
Lamination/
Assembly
TAkttM=
32 se
Figure 50: Proposed Value Stream Map
5.4.- Equipment Design
5.4.1.- Equipment comparison based on the MSDD
One of the most striking differences between the transfer line and the cell is the
equipment used in both assembly systems. One can attribute such differences to the
concepts derived from the Toyota Production System [Monden, 1993, Ohno, 1988,
Shingo, 1989] and how these concepts define the way the equipment should be designed.
However, practices that have given good results in some companies do not necessarily
yield the same results in all companies. Generalizing specific machine design guidelines
from company to company naturally restricts the potential to go beyond competitors. In
order to understand why the equipment should be designed in a "lean" way, and to allow
one to improve on other world-class equipment designs, we look again into the MSDD.
92
Visteon North Penn Electronic Engine Controller Manufacturing
As previously described, the MSDD allows us to trace high-level objectives of a
manufacturing system into lower-level physical implementations at the shop floor. The
FR-DP pairs that in some way affect equipment design and operation have been
identified [Arinez, 1999] to understand the cause-effect relationship between goals and
implementable steps.
To evaluate how well the FR-DP pairs related to equipment design are satisfied,
we use the Equipment Evaluation Tool (EET) [Gomez, Dobbs and Cochran, 2000]. The
EET can be used to assess how well a particular piece or set of equipment conforms to
the requirements imposed by the equipment related FRs. It can be used to ensure that
equipment designs better align with overall manufacturing system objectives. The tool
can also be used to identify problems in existing equipment and to set goals for the
improvement of equipment to better satisfy the requirements placed on it by the MSDD
[Gomez, Dobbs and Cochran, 2000]. The criteria used by the EET are presented in
Appendix D.
5.4.1.1.- Application of the Equipment Evaluation Tool
The Equipment Evaluation Tool is used to analyze the differences in equipment
design at the transfer line and the cell. Three similar processes were selected from both
lamination systems and measured with the EET. The processes analyzed are: PCBCasting screw-down, solder application, and conformal coater loading. Figure 51 to
Figure 53 show the processes being evaluated and Table 9 summarizes the results of the
evaluation.
Table 9: Evaluating of processes at both lines using the EET
"Lean" Cell
Transfer line
PCB-Casting screw-down
4.9
2.5
Solder application
4.7
3.6
Conformal coater loading
5.1
3.1
Process
93
Production System Design and Implementation in the Automotive Components Industry
The casting screw-down process is shown in Figure 51. The larger cycle time of
the cell allows for simpler equipment. A 2-spindle screw-gun can perform the operation
performed by a 6-spindle surrogate. With lower complexity, the 2-spindle machine is
more reliable and easier to maintain. The design is more flexible and can better
accommodate changes in the design of the product. The evaluation shown above reflects
these advantages.
The next process evaluated, the solder application, is illustrated in Figure 52. The
equipment used is, again, simpler and more accessible. The simplicity of this equipment
results from designing it to operate at a longer cycle time.
When loading the conformal coater, two greatly different processes are used. As
Figure 53 shows, the cellular approach fully utilizes labor capabilities. Using an operator
to perform this task enables simultaneous visual inspection, which helps to anticipate
production disruptions. Also, the same operator is used to unload the coated boards. On
the other hand, automating this task requires a multi degree-of-freedom robotic arm. An
additional station for automatic inspection is required. Also, a second robot for unloading
the boards is needed. The robot itself is a complex piece of machinery requiring constant
maintenance. But it represents a safety hazard for humans. Therefore, the robot should be
confined from human contact making access for repairs more intricate.
From these results, we can observe that complexity results from speed. Although
simpler, the equipment at the cell is consistently better fit to meet system-wide goals.
Higher marks are earned for the equipment used at the cell implying that this equipment
better satisfies the FRs related to equipment design, which in turn enables the system to
achieve higher-level objectives.
94
Visteon North Penn Electronic Engine Controller Manufacturing
Cellular Equipment - Cycle Time : 50 sec
Transfer Line Equipment - Cycle Time: 10 sec
~2!Spinles
Figure 51: PCB-Casting screw-down
Figure 52: Solder application at the cell
Figure 53: Loading conformal coater
95
Production System Design and Implementation in the Automotive Components Industry
5.5.- Conclusions
Throughout this study, some objectives were met. First, the Electronic Engine
Controller (EEC) production process at Visteon North Penn Electronics Plant was
explained by following the material and information flow.
Second, derived from the two different lamination processes, an analysis of the
process using a traditional transfer line was contrasted to that of a cellular approach. The
observed performance of the two systems was compared to an evaluation based on the
Manufacturing System Design Decomposition. Although the two approaches yielded
similar overall system assessments, the latter identified areas of potential improvement
for the current and future cellular implementations.
Finally, by using the Equipment Evaluation Tool, some of the equipment used in
both systems was evaluated. The higher marks attained by equipment designed for cells
implies that equipment designed with a systems perspective leads to improved overall
system performance.
96
Conclusion
Conclusion
A theoretical framework and the author's experience in the automotive
components industry are presented in this thesis. The Manufacturing System Design
Decomposition is the tool used to frame these industry applications. A commonality
between the different applications in which the author was involved is that they all
involve typical mass production plants adopting lean manufacturing practices. By dealing
with the interdependencies of the various elements of a manufacturing system, the
MSDD represents a broadly applicable, valuable tool to improve and guide the design of
systems.
The first project involves the author's experience in an automotive steering gear
manufacturer plant. A financial analysis is performed to prove the superior profitability
of cells over traditional mass production methods. Recommended conceptual machine
embodiments are presented drawing from past ventures into cellular manufacturing, and
the preferred station and work loop distribution is included.
Within the same plant, the author had the opportunity to participate at the launch
of another assembly cell. Based on standard work analysis and potential improvements
derived from the MSDD, the effectiveness of the cell is analyzed. A short-term approach
to improve its performance by reducing labor to avoid overcrowding is proposed.
However, in the long-run, it is proposed that the same number of operators run the line
while producing approximately twice the output as at launch. This can be done with the
incorporation of standard work routines and station improvements derived from the
equipment-related FRs from the MSDD.
Finally, the last project includes the work performed at an automotive electronics
manufacturing plant. Here, a mass and a lean approach for processing one stage of
production are contrasted using traditional performance metrics and conformance to the
MSDD leaf FRs. Both approaches show the superiority of the lean approach; however,
with the latter, potential areas for improving future ventures are presented.
97
Production System Design and Implementation in the Automotive Components Industry
References
Arinez, J. F., Cochran, D. S. "An Equipment Design Approach to achieve Production
System Requirements" Proceedings of the 33rd CIRP International Seminar on
Manufacturing Systems. June 5-7, 2000.
Arinez, Jorge F. and David S. Cochran. "Application of a Production System Design
Framework to Equipment Design." Proceedings of the 3 2 "d CIRP International Seminar
on Manufacturing Systems. Leuven, Belgium, May 24-26, 1999.
Cochran, David S. "The Production System Design and Deployment Framework."
Proceedings of the 1999 SAE International Automotive Manufacturing Conference.
Detroit, MI, May 11-13, 1999.
Cochran, D. S., Arinez, J. F., Duda, J. W., Linck, J., "A Decomposition Approach for
Manufacturing System Design" Journal of Manufacturing Systems, 2000
Cochran, D. S., "Production System Design" Oxford University Press, 1999.
Parnaby, J. "Concept of a Manufacturing System." International Journal of Production
Research. Vol. 17-2 (1979): 123-135
Gomez, Deny D., Daniel C. Dobbs and David S. Cochran. "Equipment Evaluation Tool
Based on the Manufacturing System Design Decomposition." Proceedings of the Third
World Congress on Intelligent Manufacturing Processes and Systems. Cambridge, MA,
June 28-30, 2000.
Hayes, R.H., and Wheelwright, S.C. "Link manufacturing process and product
lifecycles", Harvard Business Review, January-February 1979.
Hopp, W. and Spearman, M. (1996), Factory Physics, Irwin/McGraw-Hill, Boston, MA.
Monden, Yasuhiro. Toyota Production System: An Integrated Approach to Just In Time.
2 " ed. Norcross, Georgia: Industrial Engineering and Management Press, 1993.
98
References
Ohno, Taiichi. Toyota Production System. Beyond Large-Scale Production. Portland,
OR: Productivity Press, 1988.
Shingo, Shigeo. A Study of the Toyota Production System From an Industrial
Engineering Viewpoint. Trans. Andrew P. Dillon. Portland, OR: Productivity Press,
1989
Suh, Nam P. The Principles of Design. New York: Oxford University Press, 1990.
99
-
I,.
Production System Design and Implementation in the Automotive Components Industry
Appendix A: Manufacturing System Design
Decomposition v5.1
Level I
FRI
Maximize longterm return on
Investment
PM1
Return on
investment over
system lifecycle
Elm
PR
DPI
Manufacturing
System Design
Level 11
FRI I
Maximize
sales
revenue
FR12
Minimize
manufacturing
costs
PM1I
Sales revenue
PM12
Manufacturing
costs
FRI3
Minimize
investment over
production
system lifecycle
PM13
Investment over
system lifecycle
Level Ill
FR11l
Manufacture
products to target
design
specifications
PM111
Process capability
DP13
Investment based
on a long term
strategy
DP12
Elimination of
non-value adding
sources of cost
DP1I
Production to
maximize
customer
satisfaction
..1, I .
FRI12
Deliver products
on time
PM1I12
Percentage
on-time deliveries
FRI13
Meet customer
expected lead
time
PM113
Difference bet.
throughput time
and customer's
expect. lead time
FR121
Reduce waste in
direct labor
FR122
Reduce waste in
indirect labor
FR123
Minimize facilities
cost
PM121
Percentage of
operators' time
spent on wasted
motions and
waiting
PM122
Amount of
required indirect
labor
PM123
Facilities cost
-.
=
DP-111
Production
processes with
minimal variation
from the target
DP112
Throughput time
variation
reduction
DP113
Mean throughput
time reduction
DP121
Elimination of
non-value adding
manual tasks
100
DP122
Reduction of
indirect labor
tasks
DP123
Reduction of
consumed floor
space
Appendix A: Manufacturing System Design Decomposition v5.1
FR111
Level Ill
Manufacture
products to target
design
specifications
DP111
8311111
JAIIH
111111 ,,,H
Level IV
Production
processes with
minimal variation
from the target
FR-Q2
Quality
FR-Q1
Operate
processes within
control limits
mean on the
target
PM-Q2
Difference
between process
mean and target
in process output
DP-Q2
Process
parameter
adjustment
DP-Q3
parts with an
assignable cause
DP-QI
Elimination of
assign. causes of
variation
FR-Q1
FR-Q12
Eliminate machine Eliminate operator
assignable causes assignable causes
FR-Q3
Reduce variation
Center process
PM-Q1
# of defects per n
Level V
Rv
PM-Q3
Variance of
process output
Reduction of
process noise
FR-Q13
FR-Q14
FR-Q31
FR-Q32
Eliminate method
assignable causes
Eliminate material
assignable causes
Reduce noise in
Reduce impact of
process inputs
input noise on
PM-Q11
PM-Q12
PM-Q13
PM-Q14
PM-Q31
process output
Number of defects
per n parts
assignable to
equipment
Number of defects
per n parts
assignable to
operators
Number of defects
per n parts
assignable to the
method
# of defects per n
parts assignable
to the quality of
incoming material
Variance of
process inputs
PM-Q32
Output variance I
input variance
DP-Q1I
Failure mode and
effects analysis
DP-Q12
Stable output from
operators
DP-Q13
Process plan
design
DP-Q14
Supplier quality
program
DP-Q31
DP-Q32
Conv. of common
causes into
assign. causes
Robust process
design
FR-Q121
Ensure that oper.
has knowledge of
required tasks
PM-Q121
# of defects per n
parts caused by an
op.'s lack of und.
about methods
FR-Q122
Ensure that oper.
consist. performs
tasks correctly
PM-Q122
# of defects per n
parts caused by
non-standard
m
FR-Q123
DP-Q121
DP-Q122
DP-Q123
Training program
Standard work
methods
Mistake proof
operations (PokaYoke)
Ensure that
operator human
errors do not
translate to defects
PM-Q123
# of defects per n
parts caused by
human error
101
Level VI
Production System Design and Implementation in the Automotive Components Industry
Level Ill
FR112
Deliver products
on time
mill,
DP112
Throughput time
1 iijill
variation reduction
Level IV
I
FR-R1
Identifying and
Resolving Problems
Respond rapidly
to prod'n disrupt.
PM-RI
Time between
occurrence and
resolution of
disruptions
DP-RI
Proc. for detection
& response to
prod'n disruptions
Level V
#0,,
_
-, &",*
FR-RI1
j 0,4,QW W
k1AWAQ*-,T - )!j
-&;"
FR-R12
Comm. problems
to the right people
PM-R12
Time between id.
of what the disrup.
is & support res.
understanding it
Rapidly recognize
prod'n disruptions
PM-RI11
Time between
occurrence of
disruption & id. of
what the disrup. is
DP-R12
Process for
feedback of
4A
FR-RI 3
Solve problems
immediately
PM-R13
Time bet. support
res. understanding
what the disr. is &
problem resolution
DP-R13
Standard method
to id. &eliminate
root cause
Level VI
DP-RI I
Subsystem config.
to enable op.'s
detection of disr.
FR-R111
Identify
disruptions when
they occur
PM-R1II
Time between
occurrence and
recognition that
disrupt. occurred
FR-R112
Identify disrupt.
where they occur
PM-R112
Time between id.
of disruption and
id. of where the
disruption
occurred
FR-R113
Identify what the
disruption is
PM-R113
Time between id.
of where disrupt.
occurred and id. of
what the
disrupt on is
FR-R121
Identify correct
support resources
PM-R121
Time between id.
of what the
disruption is and
id. of the correct
support resource
FR-R122
Minimize delay in
contacting correct
support resources
PM-R122
Time between
identification and
contact of correct
support resource
FR-R123
Minimize time for
support res. to
understand disrup.
PM-R123
Time bet. contact
of support res. &
support res. und.
what dsruption is
DP-R111
Increased operat.
sampling rate of
equipment status
DP-R112
Simplified material
flow paths
DP-R113
Context sensitive
feedback
DP-R121
Specified support
resources for each
failure mode
DP-R122
Rapid support
contact procedure
DP-R123
System that
conveys what the
disruption is
operatio i's state
,;A_,Ii.,!1L 11 -1a -
I
102
I
-
Appendix A: Manufacturing System Design Decomposition v5.1
FR112
Level 11,
Deliver products
on time
DP1 12
Throughput time
variation reduction
Level IV
I
FR-PI
Minimize prod'n
disruptions
PM-PI
# of occurrence of
disruptions &
Amount of time
lost to disruptions
Predictable
Output
DP-PI
Predictable prod'n
resources (people,
equipment, info)
Level V
FR-P13
Ensure predictable
worker output
FR-P14
PM-P12
PM-P13
# of occurrences &
length of
unplanned eqpt.
downtime
# of disruptions &
amount of time
lost due to
operators
PM-P14
# of disruptions &
amount of time
lost due to mat'l
shorta es.
FR-P11
Ensure availability
of prod'n info.
PM-P1 I
# of occurrences &
amount of time
lost due to info.
disruptions
FR-P12
Ensure predictable
equipment output
DP-PI11
Capable and
reliable info.
system
DP-P12
Maintenance of
equipment
reliability
DP-Ph3
Motivated workforce performing
standard work
FR-P121
Ensure that
equipment is
easily
serviceable
PM-P121
Amount of time
required to
service equipmt.
FR-P122
Service
equipment
regularly
PM-P122
Frequency of
equipment
servicing
Ensure material
availability
DP-P14
Standard material
replenishment
system
Level VI
FR-P142
Ensure proper
timing of part
arrivals
PM-P142
Parts demanded
- parts delivered
FR-P131
Reduce
variability of task
completion time
PM-P131
Variance in task
completion time
FR-P132
Ensure
availability of
workers
PM-P132
# of occurrences
& amount of
operator
lateness.
FR-P133
Do not interrupt
prod'n for worker
allowances
PM-P133
# of disruptions
& amount of
time lost due to
op. allowances
FR-P141
Ensure that
parts are
available to the
mat'l handlers
PM-P141
# of occurrences
of marketplace
shortages
DP-P131
Std. work to
provide repeat.
processing time
DP-P132
Perfect
Attendance
Program
DP-P133
Mutual Relief
Syst. with crosstrained workers
DP-P142
DP-P141
Standard work in Parts moved
according to
process bet.
pitch
sub-systems
0
DP-P121
Machines
designed for
serviceability
DP-P122
Regular
preventative
maint. program
103
-
-
-
-
Production System Design and Implementation in the Automotive Components Industry
Level III
FR113
Meet cust omer
expected lead
time
DP113
Mean thro ughput
time redu ction
JA 1111111
Level IV
I
I
Level V
Level VI
FR-T2
Reduce process
delay
(caused by ra> r)
PM-T2
Inventory due to
process delay
FR-T1
Reduce lot delay
PM-TI
Inventory due to
lot size delay
Delay
Reduction
J
L
-
DP- T1
Reduction of
transfer batch size
(single-piece flow)
DP-T2
Production
designed for the
takt time
FR-T21
Define
takt time(s)
PM-T21
Has takt time
been defined?
(Yes / No)
FR-T22
Ensure that prod'n
cycle time equals
takt time
PM-T22
Difference bet.
production cycle
time and takt time
FR-T23
Ensure that part
arrival rate equals
service rate (ra=rS)
PM-T23
Difference bet.
arrival and service
rates
DP-T21
Definition or
grouping of cust.
to achieve ideal
range of takt times
DP-T22
Subsystem
enabled to meet
desired takt time
(design and op.)
DP-T23
Arrival of parts at
downstream
operations
according to pitch
FR-T221
Ensure that auto.
cycle time <
minimum takt time
PM-T221
Has this been
achieved? (Yes I
No)
FR-T222
Ensure that
manual cycle time
takt time
PM-T222
Has this been
achieved? (Yes I
No)
FR-T223
Ensure level cycle
time mix
PM-T223
Is average cycle
time less than takt
time in desired
time interval?
DP- T221
Design of approp.
auto. work content
at each station
DP- T222
Design of approp.
operator work
content/loops
DP-T223
Stagger prod'n of
parts with different
cycle times
104
Appendix A: Manufacturing System Design Decomposition v5.1
Level Ill
FR113
Meet customer
expected lead
time
DP13
Mean throughput
time reduction
Level IV
I
Delay
Reduction
(continued)
Level V
FR-T3
Reduce run size
delay
PM-T3
Inventory due to
run size delay
FR-T4
Reduce
transportation
delay
PM-T4
Inventory due to
transportation
delay
FR-T5
Reduce
systematic
operational delays
PM-T5
Prod'n time lost
due to interference
among resources
DP-T3
DP-T4
DP-T5
Production of the
desired mix and
qty. during each
demand interval
Material flow
oriented layout
design
Subsystem design
to avoid
production
interruptions
FR-T32
FR-T31
Produce in
Provide
sufficiently small
knowledge of
demanded product run sizes
mix (part types
PM-T32
and quantities)
Actual run size PM-T31
target run size
Has this
information been
provided?
(Yes/No)
DP-T31
DP-T32
Information
flow from
downstream
Design quick
changeover for
material handling
and equipment
customer
FR-T52
FR-T51
Ensure that
Ensure that
support resources producton
don't interfere with resources (people/
automation) don't
production
interfere with one
resources
another
PM-T51
PM-T52
Production time
lost due to support Production time
lost due to
resources
production
interfering with
resources
production
interfering with
resources
one another
DP-T51
Subsystems and
equipment
configured to
separate support
and production
access
requirements
105
DP-T52
Ensure
coordination and
separation of
production work
patterns
FR-T53
Ensure that
support resources
(people/
automation) don't
interfere with one
another
PM-T53
Production time
lost due to support
resources
interfering with
one another
DP-T53
Ensure
coordination and
separation of
support work
patterns
--
-
lull
-
Production System Design and Implementation in the Automotive Components Industry
Level II
FR12
Minimize
manufacturing
costs
Level III
DP12
Elimination of
non-value adding
sources of cost
Level IV
FR121
Reduce waste in
direct labor
FR122
Reduce waste in
indirect labor
FR123
Minimize facilities
cost
DP121
Elimination of nonvalue adding
manual tasks
DP122
Reduction of
indirect labor tasks
DP123
Reduction of
consumed floor
space
FR-DI
Direct Eliminate
Labor
operators' waiting
on machines
PM-Di
%of operators'
time spent waiting
on equipment
DP-D1
Human-Machine
separation
FR-D2
Eliminates wasted
motion of
operators
PM-D2
%of operEators'
time spent on
wasted mo tions
FR-D3
Eliminate
operators' waiting
on other operators
PM-D3
% of operators'
time spent waiting
on other operators
Indirect
Labor
DP-D2
DP-D3
Balanced workDesign of Nork
stations & loops to loops
facilitate o p.'s task
FR-11
Improve
effectiveness of
prod'n managers
FR-12
Eliminate
information
disruptions
PM-li
PM-12
Amount of indirect Amount of indirect
labor required to
labor required to
manage system
schedule system
DP-li
DP-12
Self directed work
Seamless
teams (horizontal
organization)
information flow
(visual factory)
FR-D1I
Reduce time ops.
spend on nonvalue added tasks
at each station
PM-D1I
% of op.'s time
spent on non
value-adding tasks
while waiting at a
station
FR-D12
Enable worker to
operate more than
one machine I
station
PM-D12
Percentage of
stations in a
system that each
worker can
operate
FR-D21i
Minimize wasted
motion of
operators between
stations
PM-D21
Percentage of
operators' time
spent walking
between stations
FR-D22
Minimize wasted
motion in
operators' work
preparation
PM-D22
Percentage of
operators' time
spent on wasted
motions during
work preparation
FR-D23
Minimize wasted
motion in
operators' work
tasks
PM-D23
Percentage of
operators' time
spent on wasted
motions during
work routine
DP-DII
Machines &
stations designed
to run
autonomously
DP-D12
Workers trained to
operate multiple
stations
DP-D21
Machines I
stations configured
to reduce walking
distance
DP-D22
Standard tools I
equipment located
at each station
(5S)
DP-D23
Ergonomic
interface bet. the
worker, machine
and fixture
106
Level V
Appendix B: Recommended action for short-term efficiency
Appendix B: Recommended action for short-term
efficiency
oended- Action
-Recom
Station
1
Automate spindle rotation
0
Install check valve impact gun
5A, 5B
Reverse order to conform with clockwise loop
6A
Seat clip at this station
11
Replace marker for quick grab/mark/return marker
21
Replace marker for quick grab/mark/retum marker
21
Mechanical stop to return pallet 70deg
21
Change poke yoke sequence
22
Add another greasing holder just as st. 21
22
Move plastic clip and tool to cut it from st. 20 (tie rod station) to this station
22
Mechanical stop to return pallet 70deg
22
Change poke yoke sequence
24
Adjust air pressure for pneumatic guns at stations 21 and 22 to reduce adjustment time at this st.
25
Replace marker for quick grab/mark/retum marker
26
Add decoupling holder
6B
Hang grease supply instead of having to lift it
6A
Bring bushings containers out chute closer to the operator
6A
Increase valves in chute angle
7
Ensure that proper weight is added to the back of input seal and other components chute, pars don't side
11
Bring material supply closer to the operator
20
Bins-in stopper is too high, bins are too heavy for a woman to lift. Reduce the height of the stopper or improve the bin
retrieval system.
20
Return to Gilman conveyor for tie-rod bins out - Replace for gravity fed chute
21
Rotate in the opposite direction or move boots to the left (the structure is in the way but can be pulled back)
107
Production System Design and Implementation in the Automotive Components Industry
21
The type of gloves used by the operator make it difficult to grab the nuts and components. Use instead tighter-fit gloves.
22
Rotate in the opposite direction or move boots to the right
22
The type of gloves used by the operator make it difficult to grab the nuts and components. Use instead tighter-fit gloves.
6A, 6B,
Introduce mechanical stops for easier rotation
11
Add additional stops at pallet return elevator to decouple two sides
Rotate air leak test 180 deg. This will relief operators work content for future work distributions without adding auto rotation
108
Appendix C: Reconnended action for long-term efficiency
Appendix C: Recommended action for long-term
efficiency
Sta#Qic
Re0no4
FIPPIar
i%0 ed'Iactoion
addressed,
The dimension on the pallets & the release/lock handles should be checked
D11
Need greaser at housing load, similar to boot greaser.
D22
Automate spindle rotation
D11
Install a wider storage tub for turn line part YL8C 30702 EA.
D23
2
Calibrate tool to stay open after second press of power button
D11
3
D21
4
Group the rack ret. ring assembly and swaging (St 5A & 5B) between the two lines (U204 and U152).
Deliver racks into st. 3
Install check valve impact gun at this st.
5
Increase raising and lowering speed at st. 5
T221, D22
6A
Replace left white delrin piece for protruding piece to ease rotation
T222
Bring bushings containers out chute closer to the operator
D23
Increase valves-in chute angle
Q11
Hang clip as pallet approaches
Q122
Move the seat press to the 'full lift' position
D11
Hang grease supply instead of having to lift it
D23
Mechanical stop to return pallet 90deg
D11
Increase raising and lowering speed
T221, D22
Replace continuous pressing of palm buttons to seat clip and palm buttons to raise pallet for walk-away
buttons
Ensure that proper weight is added to the back of comp. Chutes
D11
Have a lubricator container with a sponge right in front of the brthr chute
D11
Add a simple feature to pallet to locate the tie rod ends
D11
Place larger drip pan under line
Q13
Move oil pan to the south,
Q13
Seal the plastic pipe breather tube tray below panel box #23
Q13
Use a breather tube installation tool similar to the one used in the Winn 88 Line
D22
Consider the use of special gloves
D23
9
Rotate air leak test 180 deg and perform required gear rotation at st. 11
123, 13
11
Bring closer yoke components to the operator
D23
Mechanical stop to rotate pallet 90 deg
D11
0
1
6B
7
23
109
D11
Q122
Production System Design and Implementation in the Automotive Components Industry
Mechanical stop to rotate pallet 90 deg back
D11
Speed up the raising of pallet after pressing p.b
T221, D22
Speed up the lowering of pallet after pressing p.b
T221, D22
Speed up the release of pallet while pressing palm buttons
T221, D22
Ensure proper counterweight is in place for yoke installation
Q122
Replace marker for quick grab/mark/return marker
D23
Modify tooling to hold parts better
Q11
Consider having the supplier bevel the leading edges of yoke
Q11
Add a mirror under the yoke assembly area
D23
Reduce functional test cycle time or add additional stops to reduce trsfr time
T221
Add access gate after st. to get to repairs/reject and get in and out at the nt. End
P121, R122
Bins-in stopper is too high. Reduce the height of the stpr. or improve the bin retrieval sys.
D23
Return to Gilman conveyor for tie-rod bins out - Replace for gravity fed chute
123, 13
Replace the travel restrictor (N807853) supply pan with a dispensing tube
D22
Consider using tubing spacer clips in place of the cable ties, or hang cable tie gun from balancer
D22, D23
Rot. in the op. direction or move boots to the left (the structure is in the way but can be pulled back)
D23
Replace marker for quick grab/mark/return marker, or
D23
Eliminate the paint mark identification for the test stamp (saving aprox. 5 sec)
D11
Mechanical stop to return pallet 70deg
D11
Check poke yoke sequence
Q123
Adjust air pressure for pneumatic gun to control nut distance
P131, D23
Split tie rod grease between both boot stations
D22
Rotate in the opposite direction or move boots to the right
D22, D23
Mechanical stop to return pallet 70deg
D11
Check poke yoke sequence
Q123
Adjust air pressure for pneumatic gun to control nut distance
P131, D23
Increase the pitch of the boot hopper
Q11
Split tie rod grease between both boot stations
D22
24
The tool to hold the input shaft should be suspended
D23
25
Replace marker for quick grab/mark/return marker
D23
Place shipping plugs to the right
D23
Add decoupling holder to allow processing flow
P142, R112
16 & 17
20
21
22
26
Check poke yoke sequence
IQ123
110
Appendix C: Recommended action for long-term efficiency
Improve marking system
D23
Increase the speed of the conveyor
D11
Revise travel logic at functional test area
Q123
Safety dept. needs to evaluate the real need for palm buttons as opposed to walk-away switches
D23
Add pallets stops at pallet return elevator and functional test area
Q32
111
Production System Design and Implementation in the Automotive Components Industry
Appendix D: Equipment Evaluation Tool
E
E
Tn
U
>
'U
I
7
Quality
*m
I
I
01%
Labor
-v-
Ir-
Ir-
I
Direct
Delay
Reducfion
Time
- Variation
L
C
(D,
0~
IF
-
FR: Maximize long-term return on investment / DP: Manufacturing system design
DP: Manufacturing system designFR: Minimize production costs / DP:
on investrnent/
return customer
long-term
satisfaction
to maximize
/ DP:Maximize
Production
FR: Maximize sales revenueFR:
Elimination of NVA sources of cost
FR: Meet customer expected lead-time I FR: Reduce waste in
FR: Deliver products on time / DP:
FR Manufacture products to target
direct labor / DP:
DP: Mean throughput time reduction
Throughput time variation reduction
design specifications / DP: Production
Elimination on NVA tasks
processes with minimal variation
FR: Elimin.
FR: Stabilize Process / DP: Elimination FR: Respond rapidly to FR: Minim. FR Reduce FR: Reduce
ops. waiting
run size
process
production disruptions I production
of assignable causes of variation
on eqpmnt
disruptions / delay / DP: delay / DP:
DP: Procedure for
DP: Humandetection and response to DP: Predict. Production Prod'n of
machine
production balanced to desired mix
production disruptions
separation
takt time and quantity
resources
FR: Ensure
FR. Elimin.
predictable
op. assig.
equipment
causes /
output / DP:
DP: Stable
Reduce time
Maint. of Ensure that
output from
operators
equipment
operators
spend on
production
Eliminate
non-value
Reduce
Eliminate Ensure op.
Communic. Ensure that cycle time is Produce in
Eliminate
Rapidly
wasted
sufficiently systematic added tasks
balanced
Minimize
machine
method
errors don't recognize problems to equipment
motion of
small run operational at each
is easily
with takt
the right
facilities
assignable assignable translate to production
operators
station
delays
sizes
serviceable
time
cost
people
causes
causes
disruptions
defects
__________
L--,
-
I
:e
===No
L-,V
I __________
J
E
I __________
A.
I
I __________
I __________
__________
I
__________
__________
__________
FRs driving equipment design and operation but not
directly evaluated (evaluated using their parent FRs)
FRs used as evaluation criteria
112
Minimize
investment
over
production
system
lifecycle
-