INTEGRATION OF SMED AND TRIZ IN IMPROVING PRODUCTIVITY AT
SEMICONDUCTOR INDUSTRY
KARTIK SREEDHARAAN KUMARESAN
UNIVERSITI TEKNOLOGI MALAYSIA
INTEGRATION OF SMED AND TRIZ IN IMPROVING PRODUCTIVITY AT
SEMICONDUCTOR INDUSTRY
KARTIK SREEDHARAAN KUMARESAN
A project report submitted in fulfillment of the
requirements for the award of the degree of
Master of Engineering (Industrial Engineering)
Faculty of Mechanical Engineering
Universiti Teknologi Malaysia
MAY 2011
iii
In dedication to my beloved parents and wife
iv
ACKNOWLEDGEMENT
My sincere appreciation and thanks to my supervisor Assoc. Prof. Dr.
Muhamad Zameri b. Mat Saman for his supervision, constructive critics and the
continuous guidance to furnish this project paper to an accepted quality.
I am also indebted to my sponsor Intel Technology Sdn. Bhd. whom has
supported me financially throughout my tenure as a post graduate student. Also
would like to extend my appreciation to Mr. Ismail Ishak and Mr. TJ Yeoh from
Intel KMCO, who provided technical assistance at various occasions throughout this
project. Without their continued support and interest, this thesis would not have been
the same as presented here.
I would take this opportunity also to recognize my parents, wife, family
members and friends for their unconditional love and support.
Last but not least, none of this would be possible without the blessing and
grace of the Almighty God ‘Anbe Sivam’.
v
ABSTRACT
A case study on a test handler’s changeover process was conducted in a
semiconductor organization (Intel Technology Sdn. Bhd.). The test handler being a
constraint operation in the production supports the testing of two of the mainstream
chipset products. Though the test handler is capable to support multiple chipset
products but due to the equipment configuration complexity, the changeover process
today requires an average 4 hours to fully complete. The long changeover duration
degrades the overall productivity especially inability to meet customer demand
timely, lower utilization and rising cost issues. These identified issues are potential
factors that could impact the sustainability of the organization in long run. This case
study focuses on improving the changeover process using techniques from Single
Minute Exchange of Die (SMED) and Theory of Inventive Problem Solving (TRIZ).
Both the techniques have individual strengths and weakness and thus the focus will
be to integrate them to complement each other to enhance the changeover process
duration further. Problems in the current process like non standard practices,
complex hardware setup and waste activities that plagued today are process were
identified and categorized accordingly. Later, appropriate techniques from SMED
and TRIZ were proposed to counter these issues systematically. SMED will be used
mostly for task simplification while TRIZ will be used for hardware part redesigns
and overall process optimization. The end of mind of this study is to achieve a lean
and optimized changeover process that can be performed below 30 minutes with no
safety, quality or output concerns.
vi
ABSTRAK
Kajian kes ini bertumpu amnya pada pengubahsuaian mesin di Intel
Technology Sdn. Bhd yang merupakan pemgeluar cip komputer terbesar di dunia.
Mesin yang digunakan di operasi pemeriksaan cip silikon secara automatik ini
mampu mengendali pelbagai jenis cip tetapi memerlukan pengubahsuaian tertentu.
Proses pengubahsuaian mesin ini boleh memakan masa sehingga 4 jam untuk
disiapkan sebelum digunakan balik untuk operasi. Disebabkan ini, organisasi ini
mengalami kemerosotan produktiviti and juga kerugian kos-kos lain. Untuk
mengatasi masalah ini, teknik-teknik daripada SMED and TRIZ diperkenalkan untuk
menyelesaikan isu –isu seperti ketidakselarasan aktiviti dan penukaran alat ganti
yang kompleks. Teknik SMED and TRIZ dikaji secara teliti sebelum dicadangkan
untuk penyelesaian. Teknik-teknik SMED banyak digunakan untuk
mempermudahkan activiti kerja dan teknik-teknik TRIZ banyak berguna untuk
mereka bentuk alat ganti yang lebih mudah dan efiksyen untuk ditukar ganti.
Cadangan-cadangan ini setelah dilaksanakan dapat membantu mengurangkan masa
pengubahsuaian daripada 4 jam kepada 30 minit dengan tiada sebarang masalah
kualiti, keselamatan ataupun pengeluaran.
vii
TABLE OF CONTENTS
CHAPTER
1
TITLE
PAGE
DECLARATION
ii
DEDICATION
iii
ACKNOWLEDGEMENTS
iv
ABSTRACT
v
ABSTRAK
vi
TABLE OF CONTENTS
vii
LIST OF TABLES
xii
LIST OF FIGURES
xiv
LIST OF ABBREVIATIONS
xviii
INTRODUCTION
1.1
Overview
1
1.2
Background of Problems
1
1.3
Statement of Problems and Justification
3
1.4
Objective of Study
3
1.5
Scopes
3
1.6
Significance of Study
4
1.7
Thesis Structure
4
viii
2
LITERATURE REVIEW
2.1
Overview
5
2.2
Lean Manufacturing
5
2.2.1
History and Background
7
2.2.2
Lean Concepts
8
2.2.3
Lean Tools
9
2.2.4
The ‘bright side’ of Lean
10
2.2.5
The ‘Anti’ Lean Sentiment
11
2.3
2.4
2.5
2.6
3
Optimization of Changeover Process
13
2.3.1
Changeover and PM
14
2.3.2
Benefits of Quick Changeover
15
2.3.3
Alternatives to Quick Changeover
15
Single Minute Exchange of Die (SMED)
15
2.4.1
Advantages of SMED
18
2.4.2
Disadvantages of SMED
19
Theory of Inventive Problem Solving (TRIZ)
20
2.5.1
History and Background
20
2.5.2
Conceptual Basis of TRIZ
21
2.5.3
Common TRIZ terminologies
24
2.5.4
Tools in TRIZ field
25
2.5.5
Application and Implementation of TRIZ
28
2.5.6
Shortcomings of TRIZ
29
Summary
31
METHODOLOGY OF RESEARCH
3.1
Overview
32
3.2
Research Objective
32
3.3
Analyzing Research Methods
33
3.4
Description of Research Methods
35
3.4.1
36
Secondary Data Study
ix
3.4.2
4
Primary Data Study
37
3.5
Limitation of Research Methods
40
3.6
Summary
41
PROBLEM IDENTIFICATION
4.1
Overview
42
4.2
Background and Justification
42
4.2.1
Low Tester Utilization
43
4.2.2
Drive for Flexible Manufacturing
44
4.2.3
Driving Cost Competitive Advantage
45
4.3
Case Study Company
46
4.3.1
Background
47
4.3.2
KMCO
47
4.4
Product Background
48
4.5
Process Background
50
4.6
Equipment Background
52
4.7
Changeover Process Historical Study
54
4.8
Changeover Process Flow in Detail
55
4.9
Problems and Gaps Identification
85
4.9.1
Pre Changeover Activities
87
4.9.2
Preliminary Soft Setups
88
4.9.3
Hardware Part Setups
90
4.9.4
TIU Replacement
99
4.9.5
PnP Teaching Process
100
4.9.6
Dry Cycling Validation
102
4.9.7
TP Download Phase
105
4.9.8
Standard Unit Validation
106
4.9.9
Wrap Up Activities
107
4.10
Summary
109
x
5
PROPOSED FRAMEWORK
5.1
Overview
110
5.2
Strategy and Execution for Counter Measures
110
5.3
Process Flow Optimization
113
5.3.1
Upfront Setup Improvement Proposals
116
5.3.2
TP Download Improvement Proposals
122
5.3.3
Validation/Calibration Improvement Proposals
125
5.3.4
Post Setup Activities Alignment Proposals
127
5.4
6
Optimizing Hardware Setups phase
130
5.4.1
Proposals to Identify Fungible Parts
132
5.4.2
Proposals to Identify Non Fungible Parts
132
5.4.2.1 Non Fungible Parts With Multi Function
132
5.4.2.2 Non Fungible Parts With Hardware Redesign
133
5.5
Human Dynamic and Procurement Improvement Proposal 135
5.6
Summary
136
RESULT AND DISCUSSION
6.1
Overview
137
6.2
Implementation of Pre Setup phase
137
6.3
Improving of TP Download phase
146
6.4
Improving of Hardware Setup phase
149
6.4.1
Identified ‘fungible’ parts
149
6.4.2
Identified ‘non fungible’ parts
153
6.4.3
Redesigning hardware parts
155
6.4.3.1 Nest redesign
155
6.4.3.2 One turn screw design
162
6.5
Improvement of PnP teaching phase
171
6.6
Improvement of mechanical unit validation
176
6.7
Improvement of standard unit validation
182
6.8
Elimination of wrap up phase
188
6.9
The new optimized test handler changeover process
189
xi
6.10
7
Return of Investment (ROI) analysis
191
6.10.1 Capital and cost savings
192
6.10.2 Total utilization indicator improvement
194
6.10.3 Learner and Efficient Training
196
6.11
Critical Appraisal
197
6.12
Future Recommendation / Studies
198
6.13
Summary
199
CONCLUSION
200
REFERENCES
201
xii
LIST OF TABLES
TABLE
TITLE
PAGE
2.1
The seven deadly waste in Lean context
6
3.1
Categories of observations
39
4.1
Physical attributes comparison
50
4.2
Physical attributes comparison to NPI
51
4.3
Hardware parts involve in the changeover
64
4.4
Summary of the changeover process
86
4.5
Problems identified at the pre changeover phase
88
4.6
Problems identified at the preliminary soft setup
89
4.7
Sequence of the contactor chuck replacement
92
4.8
Problems identified during hardware part changes
98
4.9
Problems identified during the TIU replacement phase
100
4.10
Problems identified during the PnP teaching phase
102
4.11
Problems identified during the dry cycling phase
104
4.12
Problems identified during the TP download phase
105
4.13
Problems identified during the standard unit validation
107
4.14
Problems identified during the wrap up activities phase
108
5.1
The 5 pre requisite items and current practice
117
5.2
The ‘Beforehand Cushioning’ proposal
118
5.3
The activity dependency table
119
5.4
Technician idle time
126
5.5
‘Merging’ technique on identifying NVA activities
127
5.6
The current hardware setup phase
131
5.7
Technician headcount per shift
136
xiii
6.1
Pre setup improvement result
138
6.2
The new ‘End Lot’ process sequence
141
6.3
The TP Download improvement result
146
6.4
The identified ‘fungible’ hardware parts
151
6.5
The identified non ‘fungible’ hardware parts
153
6.6
The steps in changing the ‘nest’ size
158
6.7
The actual changeover steps with new ‘nest’ design
160
6.8
The functionality of the one turn screw
164
6.9
The result of the hardware setup proposals
165
6.10
The time study result of new improved hardware setup
167
6.11
The improvement result of PnP teaching phase
172
6.12
The result of dry cycling phase proposals
177
6.13
The result of the standard unit validation proposals
183
6.14
The result of the wrap up phase elimination
188
6.15
Time study of the overall new optimized changeover
process
190
6.16
The actual tool savings by factories for Q3’11
193
6.17
The contribution of TRIZ
197
xiv
LIST OF FIGURES
FIGURE
TITLE
PAGE
2.1
General changeover process
13
2.2
Shingo’s original SMED model
16
2.3
Shingo’s conceptual stages and techniques
17
2.4
SMED in Lean context
18
2.5
The 40 Inventive Principal
21
2.6
The 4 element model of TRIZ
23
2.7
The contradiction matrix
26
2.8
Tools in TRIZ field
27
3.1
Research Methodology flow
35
4.1
Breakdown of tester utilization
43
4.2
The corporate utilization goal
43
4.3
Example of demand trend
44
4.4
Cost breakdown of Nebula
45
4.5
Inventory impact analysis
46
4.6
Intel Kulim’s topography
47
4.7
KMCO’s Factory View
48
4.8
Nebula and Nexus sample units
49
4.9
Product roadmap illustration
50
4.10
The process flow
52
4.11
The M4542AD Dynamic Handler
53
4.12
Test module setup
53
4.13
Time trend of conversion from Q2’10 to Q4’10
54
4.14
The generic changeover flow
56
xv
4.15
Generic activities in pre changeover phase
57
4.16
Generic activities in preliminary soft setups
59
4.17
The change of AEPT state
59
4.18
The change of ‘sticky’ tag indicators
60
4.19
The ‘Andon’ light
60
4.20
The use of barricades
61
4.21
Reset temperature in the GUI
61
4.22
Manual turn off by pressing ‘TEMP’ button
62
4.23
Example of change kit box with hardware parts
62
4.24
Complete tool sets
63
4.25
Example of a TIU
67
4.26
Generic flow of TIU replacement phase
68
4.27
Rear of the handler
69
4.28
The ‘Clamper’ button
69
4.29
The position wheels to move the TIU and test head
70
4.30
Docked TIU on test head
71
4.31
Generic flow of PnP teaching phase
72
4.32
The command menu
73
4.33
Starting the teach sequence
74
4.34
Completing the teaching sequence
75
4.35
Existing the PnP teaching phase
76
4.36
Generic flow of dry cycling phase
77
4.37
Generic flow of TP download phase
79
4.38
Change of CTSC environment
80
4.39
TP download status
81
4.40
Choosing STD1 summary
81
4.41
Generic flow of standard unit validation
83
4.42
Generic flow of wrap up activities
84
4.43
Time study of pre changeover activities
87
4.44
Time study of preliminary soft setups
89
4.45
Time study of hardware part changes
91
4.46
Time study of TIU change phase
99
4.47
Time study of PnP teaching phase
101
xvi
4.48
Time study of dry cycling phase
103
4.49
Time study of TP download phase
105
4.50
Time study of standard unit validation
106
4.51
Time study of wrap up activities
108
5.1
The changeover duration factors
111
5.2
The generic proposal model
111
5.3
Extraction of proposal techniques
112
5.4
The generic ‘non optimized’ changeover flow
113
5.5
The ‘serial’ flow of activities
114
5.6
Illustration of ‘segmented’ of TIU phase
120
5.7
The changeover process flow with upfront setup proposals 122
5.8
Segmented TP download phase
5.9
The changeover process flow with TP download
123
improvement proposals
124
5.10
The new aligned post setup phase
129
5.11
The changeover process flow with wrap up phase
elimination proposal
130
5.12
Current contactor chuck setup
134
5.13
Current ‘nest’ design
134
5.14
X- pitch blocks
135
5.15
The current screw design
135
6.1
The new pre setup activity flow
142
6.2
Illustration of pre setup phase development
143
6.3
Actual time study of new pre setup phase
144
6.4
Illustration of changeover process change with pre setup
phase improvement
145
6.5
TP Download activity breakdown
148
6.6
Illustration of changeover process change with TP download
phase improvement
148
6.7
The transfer pick up assembly
150
6.8
The Nebula and Nexus tray difference
150
6.9
The new ‘nest’ design
157
6.10
Old versus new ‘nest’ design comparison
157
xvii
6.11
New chuck (L) and old chuck (R ) comparison
158
6.12
The current X pitch block and screw design
163
6.13
The X pitch block with new screw design
163
6.14
Illustration of hardware setup executed as parallel activity 170
6.15
Illustration of changeover process change with optimized hardware
setup
171
6.16
Actual time study of the improved PnP teaching phase
174
6.17
Illustration of PnP teaching phase with activity breakdown 175
6.18
Illustration of the changeover process change with
improved PnP teaching phase
175
6.19
The improved dry cycling activity flow
180
6.20
Illustration of the improved dry cycling phase buildup
180
6.21
Actual time study of the improvised dry cycling phase
181
6.22
Illustration of changeover process change with improved
dry cycling phase
6.23
182
Activity flow of the improvised standard unit validation
phase
186
6.24
Actual time study of the improved validation phase
186
6.25
Illustration of the changeover process change with improvised
standard unit validation phase
6.26
187
The new flow of post changeover activities replacing
wrap up phase
189
6.27
The new optimized changeover flow
191
6.28
The capital purchase ROI analysis
193
6.29
The cost breakdown analysis (post improvement)
194
6.30
The post improvement utilization analysis
195
6.31
The post improvement technician idling time reduction
196
xviii
LIST OF ABBREVIATIONS
ABBREVIATIONS
FULL NAME
AEPT
Automated Equipment Performance Tracking
ATE
Assembly Test Equipment
GU
Goal Utilization
HMLV
High Mix Low Volume
MA
Machine Availability
NA
Not Available
NVA
Non Value Added
PM
Preventive Maintanence
SDT
Schedule Downtime
SPEC
Specifications
SEMI E10
Semiconductor Equipment and Materials
International (E10 is spec)
SMED
Single Minute Exchange of Die
TIU
Test Interface Unit
TP
Test Program
TRIZ
Theory of Inventive Problem Solving
USDT
Unscheduled Downtime
VA
Value Added
CHAPTER 1
INTRODUCTION
1.1
Overview
This chapter describes a high-level overview of this project. This includes the
background of the project, problem statement, objective, scopes and the lastly
significance of this project.
1.2
Background of Project
The ever-growing technological envelope and the shrinking of product life
cycle have ultimately changed the overall face of today’s global economy where
trends are more volatile and impulsive with end-customers are more vivid in their
choices and selection of products. The ‘ripple’ of these effects has strongly
influenced in the semiconductor industries especially manufactures supporting High
Mix Low Volume (HMLV) products. It is well noted that the number of transistors
that can be placed inexpensively on an integrated circuit has doubled approximately
every two years (Gordon Moore, 1965) which precisely describes a driving force of
technological and social change in the late 20th and early 21st centuries and the trend
has continued for more than half a century and is not expected to stop until 2015 or
later. This has directly impacted the once flamboyant semiconductor industries which
2
are now facing competitive pressures to meet the ever-changing demand from end
customer and at the same time the challenge in reducing the overall operation cost.
Some 3 years ago, when Intel’s Kulim Microprocessor and Chipset
Operations (KMCO) mega-factory was erected as the biggest offshore facilities (was
then taken over by the mega factory in Vietnam in 2nd Quarter 2010) and ramped-up
aggressively for HMLV manufacturing (include assembly and testing), there were 2
main mounting challenges for the factory was;
i.
To make a breakthrough in the factory process cycle time (time taken
to manufacture a product from start of assembly to finish product ship
out)
ii.
To demonstrate a low cost competitive advantage especially when
compared against Internal competitors (other Intel factories i.e.
Penang, China, Costa Rica and US) and external competitors whom
are mostly EMS or other sub-contracted factories
The above 2 challenges are linked together by one similar gating issue which
is the conservative or the traditional manufacturing flow which focuses on batchbased production that in return produce large inventory build-ups, high storage cost
and overall lower equipment utilization. This manufacturing method opposes exactly
the concept of Lean Manufacturing which dictates on identifying and eliminating
those waste or Non Value Added (NVA) activities in accordance to achieve optimum
performance. Lean advocates for continuous flow and manufacturing flexibility of
readily adapting to the market shifts (Anthony Inman, 2010). The ability and
competency to be flexible is much easier to be said than done as the complexity to
design such facility could be both costly and sophisticated especially on long term
sustaining.
This project will fully focus on the case study of reducing the changeover
time for an Automated Testing Equipment (ATE) called the ‘Extreme Test Handler’
in a semiconductor industry by integrating 2 well known problem solving
3
methodologies; the Single Minute Exchange of Die (SMED) techniques together
with the Theory of Inventive Problem Solving (TRIZ) principals.
1.3
Problem Statement and Justification
Today, the non flexible factory environment practices batch build of products
and equipment dedication policy at the Test operation area due the long hours of
non-optimized changeover process for the Test Handler which in result causes low
equipment utilization, high inventory accumulation, higher cost per unit and the
rising cost to purchase new capital equipments.
1.4
Objective
The objective of the project is to reduce the changeover duration for a test
handler from a current 4.0 hours to 0.5 hours by integrating SMED techniques and
TRIZ principals.
1.5
Scopes
The scopes of this project will cover as below:
i.
Only techniques from SMED and TRIZ will be used. All the
TRIZ are based only from the ‘Principal’ tools
ii.
The study is limited only to the chipset products codenamed
‘Nebula Peak’ and ‘Nexus Peak’
iii.
The focus area is the Test operation of the company where the
focused equipment M4542AD Dynamic Test Handler (Extreme
Handler)
4
1.6
Significance of Study
The highlights of this project are as below:
i.
Changeovers or conversion for test equipments such as the Test
Handlers are rarely studied or practiced upon compared to other
assembly equipments such press, lathe or even conveyor based
equipments. This could be due to fact that these types of test
equipments are only commonly found at high end semiconductor
industries with structural and functional die testing capabilities.
Also to note is that this type of equipment are more complex and
sophisticated technically to change or redesign.
ii.
This project explores the opportunity to apply the SMED
techniques to perform changeover for the Test Handler and further
enhance the shortcoming of the prior techniques by integrating
them with TRIZ principals.
1.7
Thesis Structure
This thesis consists of 7 chapters. Chapter 1 is the introduction of the project
with background, problem statements, objective, scopes and the significance of the
study. Chapter 2 presents the literature review on Lean manufacturing, rapid
changeover, Single Minute Exchange of Die (SMED) and Theory of Inventive
Problem Solving (TRIZ) which also consist the critical analysis of each techniques.
Chapter 3 is basically the detail of the methodologies used throughout the project.
Chapter 4 focuses on the problem identification in the case study where both
qualitative and quantitative data collected are presented. Chapter 5 shows the counter
measure proposals with respect to the SMED and TRIZ techniques. Chapter 6 is
where the end results are presented together critical appraisals and future study
recommendations. Chapter 7 is the conclusion and summary of this project.
CHAPTER 2
LITERATURE REVIEW
2.1
Overview
This chapter will focus on reviewing and critically analyzing literatures
related and relevant to this project. The core objective is to lay out foundation for this
project by extracting the significant findings, the opportunities and the shortcomings
or gaps highlighted in the studies.
2.2
Lean Manufacturing
The term „Lean‟ has derived into multiple different meaning since its
introduction during the 20th century. According to Intel‟s Lean Learning Centre
(2003), „Lean‟ is an integrated approach in designing and improving work towards a
customer focused ideal state through engagement of all people aligned by common
principles and practices. Others prefer the simple and basic concept that Lean is to
identify and eliminate wastes from every aspect of the business (Levinson and
Rerick, 2002). Waste or commonly known as „muda’ in Japanese is defined by
Hiroyuki Hirano as „everything that is not absolutely essential‟ (Santos, Wysk and
Torres, 2006). Table 2.1 below summarizes the typical 7 deadly waste in Lean
context (Liker and Meier, 2006). The term “Lean production” was first introduced in
the book titled „The Machine that Changed the World‟ (Womack et al., 1990) and the
6
term Lean Manufacturing flourished naturally after the success of automobile makers
Toyota in the 20th century applying Lean in their shop floor. Lean manufacturing
focuses on continuous improvement of the manufacturing processes which leads to
superior processes that require less human effort, less manufacturing space, less
operational capital and less time to make products with fewer defects to precise
customer requirements (Marchwinski and Shook, 2003).
No
Waste
Table 2.1: The seven deadly waste in Lean context
Description
1
Producing items earlier or in greater quantities than needed
by the customer. Generates other wastes, such as overstaffing,
Overproduction storage, and transportation costs because of excess inventory.
Inventory can be physical inventory or a queue of
information.
2
Waiting
Workers merely serving as watch persons for an automated
machine, or having to stand around waiting for the next
processing step, tool, supply, part, etc., or just plain having no
work because of no stock, lot processing delays, equipment
downtime, and capacity bottlenecks.
3
Transportation
Moving work in progress (WIP) from place to place in a
process, even if it is only a short distance. Or having to move
materials, parts, or finished goods into or out of storage or
between processes.
4
Taking unneeded steps to process the parts. Inefficiently
processing due to poor tool and product design, causing
Overprocessing unnecessary motion and producing defects. Waste is
generated when providing higher quality products than is
necessary.
5
Excess
Inventory
Excess raw material, WIP, or finished goods causing longer
lead times, obsolescence, damaged goods, transportation and
storage costs, and delay. Also, extra inventory hides problems
such as production imbalances, late deliveries from suppliers,
defects, equipment downtime, and long setup times.
6
Unnecessary
Motion
Any motion employees have to perform during the course of
their work other than adding value to the part.
7
Defects
Production of defective parts or correction. Repairing of
rework, scrap, replacement production, and inspection means
wasteful handling, time, and effort.
7
2.2.1
History and Background
The concept of waste being built into jobs can be identified as far back as pre
20th century when Frank Gilbert observed inefficiencies in the brick laying worker‟s
repetitive motion of bending and lifting the bricks from ground to the wall. He then
improves the task by placing the bricks at wall level which resulted in almost
threefold of efficiency improvement. Subsequently, more work and motion study
was done including works like The Principles of Scientific Management (Taylor,
1911) which introduced standardization and marked the early example of early lean
techniques. Henry Ford continued the focus on waste elimination by developing the
mass assembly manufacturing system which startled the world because it exhibited a
higher degree than most people would have thought possible with the seemingly
contradictory requirements of true efficiency which are; constant increase of quality,
great increase of pay to the workers, repeated reduction in cost to the consumer. This
brought a new era of mass production in short throughput time with high profit to the
manufacturer. By enforcing very strict specification and quality criteria on the
manufactured component, he eliminated the requirement for skilled workers almost
entirely which help to reduce the manufacturing cost by 60% -90%.
Lean manufacturing concept is a generic process management philosophy
derived mostly from the Toyota Production System (TPS) which is considered as the
cradle of Lean. TPS is renowned for its aggressive focus on the 7 deadly wastes
elimination. It all started with in the 20th century when it‟s brain child Toyota,
Sakichi Toyoda implemented an improvement to auto-stop the looms when the tread
broke in his textile factory which then became the seed for autonomation or ‟Jidoka‟.
But then, it was Taiichi Ohno who initiated the Lean school of thought by putting
together all the experience, themes and the school of thought into what is known as
the Lean Manufacturing. Another industrial revolutionist called Shigeo took Lean to
a greater height when he reduce the setup time of the hull assembly on a 65,000-ton
super tanker from 4 months to 2 months, setting a new record in shipbuilding
(Vardeman et al., 2010).
8
2.2.2
Lean Concepts
Lean concepts are not necessarily a tool that we would implement but a state
or condition that we would strive for. Some of the commonly used concepts and
which are related to changeover are as below:
i.
Autonomation: Also referred to as jidoka, it is adding human element of
being able to identify problems and either stop for correction or selfcorrect before moving onto the next steps
ii.
Continuous Flow: The ideal state for any process is to move away from
traditional batching of work, whether material or information, and flow
for continuous, one element at a time. This reduces many types of waste
particularly waiting
iii.
Pull: In order to improve continuous flow and reduce the waste of
overproduction, process should be „pull‟ what they need from the
previous step in the process and only that triggers new actions
iv.
PDCA: Plan- Do-Check-Act means that whether solving a problem or
building a plan everyone should follow this process to ensure learning and
success towards the goal
v.
Visual Management: Is both a tool and a concept. The ideal state is all
employees, operators and management, should be able to manage all
aspect of the process at a glance using visual data, signals and guides.
vi.
Value Added: The Japanese term of waste is known as Muda (Sayer and
Williams, 2007) and can be divided into Type-1 muda (Non-value-added
but necessary) and Type-2 muda (Non-value-added and not necessary).
Value Added task are only those tasks which (1) the customer is willing
to pay for, (2) transform the product or service, (3) are done right after
first time
vii.
Waste Elimination: Eliminating waste from the process is the goal of
many lean tools and should be an ongoing effort in itself. This comes in
form of 7 types of waste best remembered using the acronym
TIMWOOD; Transportation, Inventory, Motion, Waiting,
Overprocessing, Overproducing and Defect
9
2.2.3
Lean Tools
Lean tools are proven practices that help to move closer to the ideal state and
are consistent with the Lean Principles. There are various methods and tools to assist
the changeover process and help to implement the Lean production systems. Some of
the universally accepted tools as per documented by the Association for
Manufacturing Excellence are as below:
i.
5S: adapted from the Japanese words that start with „S‟ but have been
rewritten as Sift, Sweep, Sort, Sanitize and Sustain
ii.
Error Proofing: Error Proofing is also known as poke-yoke or mistake
proofing. It involves redesign of equipment or process to prevent
problems from occurring
iii.
Six Sigma: Method and a set to tools to reduce variations in a process
particularly quality using mostly statistical tools. It‟s primary method is
DMAIC: Define, Measure, Analyze, Improve and Control
iv.
Scoreboards: are part of Visual Management of safety, quality, delivery
and cost metrics including analysis and action plans use to help shop floor
teams to manage their own process
v.
Layout: An arrangement of work(machines, people, method and material)
so that processing steps are sequentially done in next door steps one at a
time in order to efficiency, reduce waste and improve communication
vi.
Kanban: Often in the forms of cards, are signals that a downstream, or
customer, process can use to request specific part from the upstream, or
supply or process
vii.
Andon: The andon cord is the ability for an operator to pull a cord that
triggers a horn and light which tells their team leader that they need help
or support safety, health, and environment
viii.
Honshin kanri: A strategic planning process to establish strong agreement
and align people in a common direction with agreed upon methods to
improve
10
ix.
Kaizen: Kaizen is a structured process to engage those closest to the
process to improve both the effectiveness and efficiency of the process.
Its goal often to remove waste and add standardization
x.
Preventive Maintenance: Simplifying and structuring maintenance
activities to prevent problems rather than react to them can increase
capacity to improve continuous flow
xi.
Setup Reduction: The time takes to changeover equipment from one
product to the next is a major barrier to continuous flow and setup
reduction seeks the reduction or elimination of that time. This is also
known as SMED or Single Minute Exchange of Dies
xii.
Value Stream Mapping: This is structured process helps manager
understand the flow of both material and information through their
operation and develop plans to move them closer to ideal state
The ‘bright side’ of Lean
2.2.4
T Ross & Associates Environmental Consulting Ltd. (2003) mentioned in
their report, that Lean production typically represents a paradigm move from
conventional “batch and queue” functionally aligned mass production to “continuous
flow” product-aligned pull production. This shift requires the implementation of justin-time production principle and employee-involved, system-wide, continual
improvement. Several potential outcomes of implementing Lean are identified as
follow:
i.
Reduced Inventory Level: Include raw materials, work-in-progress(WIP),
finished product along with associated carrying costs and loss due to
damage, spoilage, off-specification etc
ii.
Decreased Material Usage: Include product inputs, including energy,
water, metals, chemicals, etc by reducing material requirements and
creating less material waste during manufacturing
11
iii.
Optimized Equipments: Capital equipment utilized for direct production
and support purposes using lower capital and resource intensive machines
to drive down costs
iv.
Reduce need for Factory Facilities: Physical infrastructure primarily in
the form of buildings and associated material demands by driving down
the space required for product production
v.
Increase Production Velocity: The time required to process a product
from initial raw material to delivery to a consumer by eliminating process
steps, movement, wait times, and downtime
vi.
Enhance production flexibility: The ability to alter or reconfigure
products and processes rapidly to adjust to customer needs and changing
market circumstances enabling the implementation of a pull production,
just-in-time oriented system which lowers inventory and capital
requirements
vii.
Reduced complexity: Complicated products and processes that increase
opportunities for variation and error by reducing the number of parts and
material types in products, and by eliminating unnecessary process steps
and equipment with unwanted features
The ‘Anti’ Lean Sentiment
2.2.5
It has not always been glitz, glamour and glory for Lean as there has been
some known and obvious disadvantage of the Lean concepts and the setback from its
implementation. Apple Inc has been described as successful giant organization that is
known to be „Anti Lean‟ in its business practices. Some of the disadvantages of
Lean (Wen Jie and Lang Wood, 2010) could be summarized as:
i.
Supply Chain Problem because only a small amount of inventory is kept
on hand, lean manufacturing depends heavily on suppliers that can
provide products for the manufacturing process dependably and without
interruption. Problems like employee strikes, transportation delays and
12
quality errors on the part of suppliers can create manufacturing holdups
that can be fatal. Vendors may be unable or unwilling to supply parts or
products on a tighter schedule or in smaller amounts. These needs can
burden suppliers with unprofitable costs and create tensions that
ultimately affect the manufacturing process and can cause frequent
changes of suppliers, or even difficulties finding suppliers who can
provide on the necessary schedule at all
ii.
High Cost of Implementation of lean manufacturing often means
completely dismantling previous physical plant setups and systems.
Training employees can be lengthy and acquiring managers experienced
in lean manufacturing process can add considerably to company‟s payroll
expenses. The purchase of machinery that increases efficiency, and the
setup of smaller work cells can add to long-term debt. Small and mediumsized businesses, in particular, may find the cost of changeover to lean
manufacturing processes prohibitive
iii.
Customer Dissatisfaction Problems because lean manufacturing processes
are so dependent on supplier efficiency, any disruption in the supply
chain and therefore, on production can be a problem that adversely affects
customers. Delivery delays can cause long-lasting marketing problems
that can be difficult to overcome
iv.
Requires Close Management where Lean manufacturing requires close
management that a lean production system must be constantly monitored
to find potential future problems and maintain production efficiency. This
close management requires regular communication between line level
employees and management. In many organizations, there is a divide
between management and workers, resulting in a hesitation to share
information. Close management also can run the risk of management
being too active in fine-tuning employee tasks, often called
"micromanagement"
v.
Employee Pushback, in some cases, organizations may be culturally
resistant to change. Employees may balk at changing how they do their
jobs. In some cases, a lean production modification may cause one set of
employees to have more work, while reducing the work load on another
13
set. This can trigger resentment and employees who fight the
implementation of a lean production system. In some cases, the pushback
may come from management. This is especially true in organizations
where managers see their organizations as separate "fiefdoms" under the
larger organization. Managers sometimes can attempt to sabotage the
implementation of a system because of a fear of losing power and control
of the department and groups that they head
2.3
Optimization of Changeover Process
Changeover or conversion is defined as the need of equipments to support for
2 or more products simultaneously. Figure 2.1 shows a general changeover process
and the total elapsed time that is measured from the ramp down period of current
product to the time the new product is fully ramped up.
Figure 2.1: General changeover process
It is necessary to optimize line changeover efficiency especially in a High
Mix Low Volume (HMLV) electronics assembly environment before Lean
Manufacturing is implemented (Douglas Farlow, 2005). Traditionally, improvements
in changeover are approach only through the evaluation and elimination of the Non
Value Added (NVA). Studies have shown the existences of the 7 „deadly‟ waste in
an inefficient changeover and highlighted the goal of an efficient changeover is to
14
reduce waste specifically transportation and motion (Womack and Jones, 1996). Also
Total Productive Maintenance (TPM) has also identified quick change over time as
one of it‟s six key principles (Nakagima, 1988).
2.3.1
Changeover and Preventive Maintenance
The role of better changeovers to facilitate reducing downtime losses has
been widely reported (Klospic and Houser, 1997) including changeovers application
in TPM and the 6 big losses. However, there has been little assessment of
changeover improvement techniques been applied for maintenance procedures. There
are many similarities between activities that happen at changeover and maintenance
as summarized below:
i.
Involve substitution of components
ii.
Dismantling to gain access
iii.
Removal of component or assemble
iv.
Replacement of component
v.
Adjustment to reinstate the desired level of performance
Poor quality of parts can impact changeover and maintenance performance.
With better maintenance practice, changeover process can be optimized with the
reduction of lengthy downtime due to component damage or the wear and tear
situations (McIntosh et al., 1999). Though typically, maintenance and changeovers
happens separately at different times frame which increases tool downtime but there
are opportunities to integrate changeover and maintenance activity especially for
those with low variability high frequency changeover/maintenance scenario.
15
2.3.2
Benefits of Quick Changeover
Changeover reduction is an extremely powerful tool which improves a plants
ability to provide better utilizing its own assets and improve productivity of
employees (Rubrich and Watson, 2004). Also some of the advantages of
manufacturers implementing changeover reduction (Feld, 2000) are:
i.
Improved product quality
ii.
Work in process (WIP) reduction
iii.
Reduction in set up costs
iv.
Improved utilisation of labour and equipment.
v.
Shorter Manufacturing lead time
vi.
Quicker response to customer demand
2.3.3
Alternatives to Quick Changeover
The following are some of the available options to explore to achieve quick
changeover;
2.4
i.
Single Minute Exchange of Die (SMED)
ii.
Group technology/cell formations
iii.
Design Standardization
iv.
Mechanization or automation
Single Minute Exchange of Die (SMED)
Single Minute Exchange of Die or better known as SMED took its first step
in 1950‟s as a concept from the brain child of Shigeo Shingo‟s efficiency experiment
at Toyo Kogyo Mazda plant in Hiroshima, Japan. It is also worth to note that, it was
16
Dr Shigeo‟s who first initiated the concept of „Defect=0‟ or „poka-yoke’ in 1961
(Vardeman, 2010). Shingo‟s SMED was highly inspired by Taylor‟s Principle of
Scientific Management. SMED emphasises that changeover improvements is sought
primarily by rearranging internal and external elements where the whole changeover
process can be completed below 10 minutes. According to Shingo, an internal
activity is something that can be performed only when the tool is stopped or down
for production while external activities can be performed in parallel when the tool is
up and running production. Figure 2.2 shows the original approach by Shingo‟s to
achieving quick changeover while Figure 2.3 shows an enhanced SMED conceptual
stages with techniques:
Figure 2.2: Shingo‟s original SMED model
Shingo‟s 3 Main Stages:
i.
Stage 1: Separate internal and external activities.
ii.
Stage 2: Converting internal activities to external where and when
possible
iii.
Stage 3: Engineering, streamlining or standardising all remaining
activities so that they can be completed in the fastest possible time
Shingo‟s 12 main techniques:
17
i.
Using Checklist
ii.
Perform Function Check
iii.
Improving Die Transportation
iv.
Preparing Operation Conditions in Advance
v.
Function Standardization
vi.
Using Intermediary Jigs
vii.
Improving Storage and Transportation for Blades, Dies, etc
viii.
Implement Parallel Operations
ix.
Use Functional Clamps
x.
Eliminating Adjustments
xi.
Least Common Multiple System
xii.
Mechanization
Figure 2.3: Shingo‟s conceptual stages and techniques
By 1980, SMED was ear-marked as one of the most revolutionary concept in
manufacturing after it successfully integrated into the Toyota Production System
(TPS) when the die punch set up time in the cold forging process was reduced form
one hour 40 minutes to just 3 minutes with significant reduction in operating costs
18
and improve in the lead time (Vardeman et al, 2010). SMED became the cornerstone
of Lean Manufacturing especially in setup time reduction by waste elimination and
enabling smaller batch sizes of lots to be processed, demonstrating JIT and as an
element for continuous improvement or „kaizen’ (Karlsson and Ahlstrom, 1996) as
shown in Figure 2.4.
Figure 2.4: SMED in Lean context
2.4.1
Advantages of SMED
According to Shigeo, SMED will help lower the skill level requirements
when the tool changeover are quick and simple which in return eliminate the need to
hire or train highly skilled workers. As earlier discussed, competitive advantages like
quick customer response could be achieved through SMED where rapid changeovers
will create a flexible manufacturing environment. Shingo‟s SMED promotes world
class manufacturing through delivering outstanding business results and
improvement in customer satisfaction levels (Varderman, 2010). Typically
improvements in changeover is approached only through the evaluation on the Non
19
Value Added (NVA) steps but through the introduction of SMED as a Lean tool to
reduce changeover time, even the Value Added (VA) steps are accessed and waste
are eliminated systematically at all levels. Today, SMED can be applied to every
situation and urges that everyone would benefit from its principles (Brad McEnroe,
1985) where the main benefits are the introduction of parallel operations and a
reduction in wasted adjustments by having fixed numerical settings, using gauges
and using referencing plans. Wasted adjustments, can consume up to 50% of set-up
times, when not tackled.
Besides this, another advantage here is improved employee involvement.
Both SMED and Lean by culture welcome employee involvement towards
continuous improvement. Employee involvement (Kearney, 1997) is one of the
dominant elements of world class manufacturing that can make both small and large
companies competitive in the global market. Organization especially shop floors with
SMED implemented has shown a positive change in employee‟s attitude where
involvement of employee in both problem solving and decision making has yield
outstanding results especially in set up time reduction cases.
2.4.2
Disadvantages and gaps in SMED
The biggest 2 gaps found in the SMED methods and techniques today are;
i.
Sustainability – substantial evidence showing that some changeover times
has continued to improve, some had stabilized while a few has took a
reverse and degraded. The main reason for the degradation in changeover
time was due to sustainability issues. Management commitment and
continuous focus on changeover reduction are the key enablers on
sustainability (McIntosh et al., 2001)
ii.
Low focus on hardware redesign- SMED does not sufficiently promote
some important options, particularly those who seek to reduce the
duration of the changeover tasks or eliminate them altogether. This
20
particularly arise when a system in need for a design change for
continuous improvement (McIntosh et al., 2000)
2.5
Theory of Inventive Problem Solving (TRIZ)
TRIZ (pronounced TREEZ) is the Russian acronym for „’Teoriya Resheniya
Izobreatatelskikh Zadatch‟‟ or the Theory of Inventive Problem Solving. TRIZ
introduced by Russian engineer and scientist Genrikh Altshuller in 1946, is a
problem solving method based on logic and data, not intuition, which accelerates the
ability to solve problems creatively. TRIZ also provides repeatability, predictability,
and reliability due to its structure and algorithmic approach (Orloff, 2003). This
proven algorithmic approach to solving technical problems began when Altshuller
studied thousands of patents and noticed certain patterns. From these patterns he
discovered that the evolution of a technical system is not a random process, but is
governed by certain objective laws. These laws can be used to consciously develop a
system along its path of technical evolution - by determining and implementing
innovations. One result of Altshuller's theory that inventiveness and creativity can be
learned has fundamentally altered the psychological model of creativity.
2.5.1
History and Background
TRIZ has an interesting history (Terninko, Zusman & Zlotin, 1998) where
after World War II Genrich Saulowitsch Altshuller (1926–1998), at that time a young
and talented inventor in the former Soviet Union and a patent examiner for the
Russian navy, started to develop the „classical‟ TRIZ as a methodology to create
systematic innovations. His research began with comparing patents to analyze how
inventors invent. Because of his criticism of the Soviet system, Stalin banished
Altshuller to Siberia, where he spent a number years in a labor camp. During this
time, he worked intensively on around 200,000 patent analysis and discovered only
around 40,000 of the patents are considered to be really worth it to be considered as
21
inventions. For instance, Altshuller extracted the 40 inventive principles as shown in
Figure 2.5 below, on which many technical solutions are based on. Furthermore, in
the labor camp he met like-minded researchers who were interested in his research,
and with them he discussed his ideas. In the following years, Altshuller developed
successively TRIZ tools and characteristic terms. Later, as a consequence of glasnost
or the openness policy by the Soviet government post 1980s, the TRIZ knowledge
grow enormously especially in the Western Hemisphere when several of
Altshuller‟s students emigrated to the USA, Scandinavia, Israel and Germany.
Figure 2.5: The 40 Inventive principals
2.5.2
Conceptual Basic of TRIZ
TRIZ research began with the hypothesis that there are universal principles of
creativity (Katie, Ellen and Michael, 2006) that are the basis for creative innovations
that advance technology. If these principles could be identified and codified, they
22
could be taught to people to make the process of creativity more predictable. The
short version of this is:
‘’Somebody someplace has already solved this problem (or one very similar
to it.) Creativity is now finding that solution and adapting it to this particular
problem’’.
Throughout the 60 years of research the TRIZ community has accumulated wide
range of experiences and knowledge of former inventors to help support new
inventors when they have to solve primarily technical or technical-economical
problems (Martin Moorle, 2005). The 3 fundamental findings of TRIZ are;
i.
Problems and solutions are repeated across industries and sciences. The
classification of the contradictions in each problem predicts the creative
solutions to that problem
ii.
Patterns of technical evolution are repeated across industries and sciences
iii.
Creative innovations use scientific effects outside the field where they
were developed.
Problem solving within TRIZ can be described using a four-element model as shown
in Figure 2.6:
i.
The problem-solver should analyze his specific problem in detail. This is
similar to many other creative problem-solving approaches
ii.
He should match his specific problem to an abstract problem(or general
problem)
iii.
On an abstract (general) level, the problem-solver should search for an
abstract (general) solution
iv.
If the problem-solver has found an abstract (general) solution, he should
transform this solution into a specific solution for his specific problem.
23
Figure 2.6: The 4 element model of TRIZ
During this process, TRIZ can support the problem-solver by accumulating
innovative experiences and providing access to effective solutions independent of
application area. As there are different TRIZ tools corresponding with different
levels of abstraction, this process may vary in the heights of abstraction, and also in
the number of loops, which the problem-solver is passing through. For each step the
problem-solver can resort to specific TRIZ tools:
i.
The specific problem can be analyzed and transformed into an abstract
solution, for example, through the tool function analysis.
ii.
The tool contradiction can help to formulate the problem on an abstract level.
iii.
Afterwards, the inventor can find abstract solutions for the formulated
abstract problem, for instance using the tools inventive principles, separation
principles or substance- field-modulation.
iv.
Lastly, the problem-solver can use the tool ideal machine for assessing the
found solution concepts and transfer the selected solution concept into a
specific solution
24
2.5.3
Common TRIZ terminologies
Some of the common TRIZ terminologies are as below;
i.
Technical System: Everything that performs a function is a technical
system. Any technical system can consist one or more subsystems. The
hierarchy of technical system spans from the least complex with only 2
elements to a complex with many interacting elements
ii.
Level of Innovation: Analysis of a large number of patents shows not
every invention is equal to its inventive value. There are 5 levels of
inventions beginning from Level#1(simple improvement) to
Level#5(discovery of a new phenomena). Almost 77% of the patents
reviewed showed they belong to either Level#1 and Level#2, and TRIZ
could help innovators to elevate to Level#3 or Level#4
iii.
Laws of Ideality: States that any technical system throughout its lifetime
tends to become more reliable, simple and effective or more ideal.
Systems are considered ideal when mechanisms disappeared while the
function remains.
iv.
Contradiction: are situation where improvement of one characteristic,
cause another characteristic to deteriorate and mostly a compromised
solution are sought. The most effective solutions are solving of technical
problems with contradiction. There are 2 types of contradiction; technical
contradiction are contradiction happening inside a technical system and
the 40 principals are used to resolve them; physical contradiction where 2
opposite properties are required from the same element of a technical
system with multiple methods to solve this type of contradictions
v.
Evolution of Technical System: Altshuller established 8 Patterns or Lines
of technical system evaluation namely Life Cycle, Dynamization,
Multiplication Cycle, Transition from Macro to Micro, Synchronization,
Scaling up and down, Uneven development of parts and Automation
25
2.5.4
Tools in TRIZ field
TRIZ offers a comprehensive set of tools to analyze and solve problems in
different perspectives. Lev Shulyak has summarized TRIZ tools into 3 different
categories as below;
i.
Principals: the tools to overcome contradiction where consist of generic
suggestions for performing an action to, and within a technical system.
The 40 Inventive Principles (Figure 2.5) together with the first 20
Contradiction matrix (Figure 2.7) are example of tools from this category.
ii.
Standards: Are structured rules for the synthesis and reconstruction of
technical systems where it helps to combat complex problems. Standards
provide 2 functions; a) Standards help to improve and existing system or
synthesis a new one and b)Standards are most effective way for providing
graphical model of a problem and is called S-field modeling. Altshuller
offered 72 Standards divided into 5 classes
iii.
ARIZ: is the abbreviation Algorithm to Solve an Inventive Problem is the
central analytical tool of TRIZ. It provides specific sequential steps for
developing a solution for complex problems. The latest version is called
ARIZ-85C which contains 9 main steps with a multiple substeps
On the other hand, Martin G. Moehrle (2005) in his work on „Framework for TRIZ
research‟ summarized that the comprehensive TRIZ toolkit could be structured based
on the five field framework. The tool list in TRIZ field is summarized in the Figure
2.8.
26
Figure 2.7: The contradiction matrix
27
Figure 2.8: Tools in TRIZ field
28
2.5.5
Application and Implementation of TRIZ
TRIZ has made it presence in many large established organizations around
the globe with case real case study has shown the success story. A report from
Fortune Magazine (Peter Lewis, 2005), shows how Semiconductor /Electrical
companies like Samsung is trains the engineers aggressively in TRIZ to develop new
designs and ideas which till date Samsung owns up to 1600 patents in 2005 alone.
Besides them, US companies like GE, Intel, Motorola are fast catching up on the
TRIZ techniques especially in their respective R&D department to invoke
breakthrough ideas.
Research below has shown the evolvement of TRIZ in today‟s real world,
notably the absorption into the corporate world alongside Lean and Six Sigma (Katie,
Ellen and Michael, 2006);
i.
Entertainment: Applications of TRIZ developed eight families of
solutions to solve Singapore‟s automobile traffic on the Sentosa , its
entertainment island (aquarium, bird sanctuary, dolphin show, restaurants,
music, etc)
ii.
Warranty cost reduction: Ford used TRIZ to solve a persistent problem
with squeaky windshields that was costing several million dollars each
year. Previously, they had used TRIZ to reduce idle vibration in a small
car by 165 percent, from one of the worst in its class to 30 percent better
than the best in class.
iii.
IT Product development: A manufacturing company doubled the value to
the customer of their patient interview system for opticians offices by
applying the feedback and self-service principles of TRIZ to the overall
product development, and applying the principles of segmentation, taking
out and composite construction to the training and support.
iv.
School administrators: Creativity has been greatly enhanced in situations
ranging from allocation of the budget for special education to building
29
five schools with funding only for four, to improving racial harmony in
the schools.
v.
Waste processing: Dairy farm operators could no longer dry the cow
manure due to increased cost of energy. TRIZ led the operators to a
method used for the concentration of fruit juice, which requires no heat.
vi.
Dow Chemical Company showed the combined effect of TRIZ with
Design for Six Sigma (DFSS) saw the Breakthrough achieved in
controlling of monomer residuals, handling of raw materials, and reactor
design. The reduction amazed even the project team, when the capital cost
of a plant built to the new standard dropped by more than 25 percent,
from nearly $110 million to < $80 million.
2.5.6
The Shortcomings and Opportunities in TRIZ
As the worldwide appreciation for TRIZ grows, its further development as a
science is hindered by a number of factors (Simon Letvin et al.,2010) mainly due the
fuzzy boundaries of TRIZ but hardly any scientific association, even the most
respected and established one, is able to define with utmost certainty the boundaries
of its “subject” science.
This in response leads to significant differences in the interpretation of some
key concepts, tools and approaches of TRIZ (and in some cases to their corruption).
Unfortunately, there are currently no TRIZ textbooks or training programs that are
universally accepted by the global TRIZ community. These discrepancies make it
difficult to meet an increasingly tangible need for a globally authoritative
certification of TRIZ practitioners.
Presently, there are several new developments based on “classical” TRIZ,
such as I-TRIZ, TRIZplus, TRIZ-OTSM, and some others. Although some of these
developments are used by many TRIZ practitioners, they have not yet attained the
30
status of being universally accepted BUT there is already aggressive steps taken by
especially the International TRIZ Association together with Altshuller Institute to
work on this.
Martin G. Moehrle (2005) in his proposal on TRIZ‟s research framework
pointed out 5 improvement aspects as below:
i.
Core– the various TRIZ tools are located in the centre of the framework. How
effective and efficient are those tools? What combinations of them are
particularly helpful? What theoretical background can be found to enhance
the toolkit? What comprehensive process models (see above) should be used
under what conditions?
ii.
Inspiration- some aspects may inspire the development of TRIZ. New
scientific knowledge domains and new patented inventions may be assigned
to this field of research. What findings, for example, from the field of bionics
can be integrated into TRIZ and its tools? Which new inventive principles
can be found in patents covering newer technologies like photovoltaic or
nanotechnology?
iii.
Adaptation– TRIZ may be applied in different fields. In a narrow sense, one
can think of different technical fields, reaching from classical engineering to
software and gen technology. In a broader sense, other fields of application
like management, sociological or even psychological problems can be
considered (bearing in mind that people are not technical systems)
iv.
Psychological and sociological contingency–the application of TRIZ and its
tools is embedded in psychological and sociological factors. Which personnel
types prefer which TRIZ tools? How is this related to the qualification of the
people? How should group members be assigned to reach acceptable results
with TRIZ tools?
v.
Integration-several creativity techniques, such as lateral thinking, morphology
and CPS have been developed. Furthermore, a lot of quality tools such as
FMEA or QFD are being discussed. How could these techniques and tools be
combined with TRIZ tools
31
2.6
Summary
From the critiques above, it is clear that both the renowned techniques
discussed above have numerous advantages and individual strengths alongside with
its weakness. This shows the possibility of molding both techniques to complement
each other in a changeover process.
CHAPTER 3
METHODOLOGY OF RESEARCH
3.1
Overview
The purpose of this chapter is to give a concise overview of the research
method used throughout this project. This chapter will define the research objectives,
methods used and lastly the limitation faced and some recommendations to overcome
this setback.
3.2
Research Objective
The primary objective of the project is to reduce the overall time taken to
complete high frequency-high variability changeover activity for a handler. The
following are some of the Research Questions that this project to aims to answer:
33
i.
What are the current state and the process flow for the changeover
process?
ii.
Is it possible to reduce the changeover durations using SMED techniques
and how much saving is expected?
iii.
Similarly, is it possible to integrate SMED techniques with TRIZ
principals to enhance the changeover reduction time? How much is
saving is expected with this method?
iv.
How to ensure the improved changeover process does not affect Safety,
Quality and output delivery?
3.3
Analyzing Research Methods
The evolution of research has created an alternative pool of available
Research Methods (RM) to suit individual needs. According to Shariff (2004),
research methods can be divided into 4 types namely experiment, surveys,
observation and secondary data study. In a broader perspective, this could be
categorized as qualitative and quantitative based RMs.
Qualitative research methods are fields or inquiry that crosscuts disciplines
and subject matters. Qualitative researchers aim to gather an in depth understanding
of human behaviour and the reason that govern human behaviour. Qualitative
researchers normally use four methods for gathering information:
i.
Direct observation or ‘’gemba’’
ii.
Participation in the setting or self-experience (experience survey)
iii.
In depth interviews, surveys and questionnaires etc
iv.
Analysis of documents, papers, info or called secondary data study
34
Quantitative research is much more scientific that qualitative approach, and
aims to determine a relationship between variables. The objective of quantitative
research is to develop and employ mathematical models, theories and/or hypotheses
pertaining to natural phenomena. The process of measurement is central to
quantitative research because it provides the fundamental connection between
empirical observation and mathematical expression of quantitative relationships.
Quantitative research is often an iterative process by which evidence is
evaluated, theories and hypotheses and refined, technical advances are made.
Quantitative research is generally approached using scientific methods such as:
i.
The generation of models, theories and hypotheses
ii.
The development of instruments and methods for measurement
iii.
Experimental control and manipulation of variables
iv.
Collection of empirical data
v.
Modelling and analysis of data
vi.
Evaluation of results
Researchers have long debated the relative value of qualitative and
quantitative inquiry (Patton, 1990). Russek and Weinberg (1993) claim that by using
both quantitative and qualitative data, their study gave insights that neither type of
analysis could provide alone. The form of research that combines multiply research
methods
is
usually
described
as
one
of
convergent
methodology,
multimethod/multitrait convergent validation or what has been called “triangulation”.
Based on the research questions posed earlier, it is best to combine both
qualitative and quantitative research methods where the former will be used in at the
early stage especially on understanding human behaviour and the interaction with the
changeover process. Later, this will be coupled with quantitative techniques to
35
analyse the info and data collected to be presented in a more scientific and empirical
way.
3.4
Description of Research Methods used
Below are some of the crisp details of the RM that will be used throughout
this paper. Figure 3.1 then summarizes the whole process flow.
Figure 3.1: Research Methodology flow
36
3.4.1
Secondary Data Study
Some of different categories of secondary data study used throughout this
project are as below;
i.
Historical Data: Once the problem statement and focus area has been
indentified, the 1st step is to collect, compile, and review and thoroughly
analyze the historical data i.e. the time trend taken to complete a changeover
process for the equipment in focus. The raw historical data can be collected
from the automated database of the company (i.e. MyATM) that stores each
event happened on each equipment for a tenure for 26 weeks. The automated
system is designed to capture events based on the SEMI E10 standard where
each time and operator performs changeover (conversions) the event will be
auto-log under the SCHEDULED DOWNTIME bracket. This automated
database are ease for retrieval and are more trustworthy than the manual
tracking since all activities are not easy to be manipulated and minimum
human errors. But nevertheless, it is important to carefully analyze the raw
samples and exclude only any known outliers or out of control (OOC) data to
ensure the data integrity. The extracted raw data can be presented in either in
a periodic line graph or time trend based on week to week and between shift
to shift performances. This information is vital to understand the actual
scenario versus the target goal especially for the changeover process. The
historical trends are important to show obvious gaps and help to estimate the
severity of the problem statement
ii.
Standard Changeover Specification and/or Technician training document:
Normally documents like this are stored in both hardcopy by the engineering
department and also as a softcopy in the organization‟s database. This is
mostly restricted or confidential documents. It is necessary to thoroughly
analyze the standard documents; ensure its updated and valid for use. Also, it
will help to provide an overview of the changeover require and necessary
steps needed currently to ensure a complete conversion. Document the
important steps that will be used at later stage for comparison purpose.
37
iii.
Literature Review: The next best source of information through previous
studies did globally by like-minds that shares interest on similar research area
or problem statement. Through critical analysis of related literatures, scenario
like „reinventing the wheel‟ could be avoided and help to expand the
knowledge gain from the research of others. Good literature review will help
to scope the research better and identify opportunities and gaps from other
research to explore upon.
iv.
Changeover Technician Info: This include the information i.e. Name, ID etc
of the personnel involve during the changeover process (info normally
available thru the operator log book) which will help to understand the
qualification or competency level of the technicians.
3.4.2
Primary Data Study
Some of different categories of primary data study used throughout this
project are as below;
i.
Direct Observation: Also known as ‘’Gemba’’ by the Japanese and
commonly use in the Lean environment. It simply means of going directly
to the point of activity and perform a direct observation personally as
oppose to the secondary data above. In direct observations, the key point
is to „actively‟ observe with no participation in the overall observed
activity (or passive involvement). Key objective of any direct observation
is to understand the activity, connections and flow besides identifying
opportunities in the process to eliminate waste or „muda‟ based on the 7
deadly waste abbreviated as „‟TIMWOOD‟‟. The end-information of a
Direct Observation is very important because only then the actual
scenario or the current state of i.e. the changeover process could be
38
spelled out correctly before any improvements are implemented. An
effective Direct Observation should involve group of individuals from
related expertise area i.e. engineering, IE, manufacturing, supervisors and
the technicians themselves. Each member will be assigned a role where
the team will observe based on the 3 main categories as below in Table
3.1 and through answering the related question with respect to the
changeover process. The Direct Observation will be an iterative process
where it will be conducted at all 4 shifts during changeover process to
understand the impact of variables like time, techniques and competency
levels. Also as continuity, the direct observation will be done pre and post
improvement implementation. To ease analyzing and to map the direct
observation, some of the tools as below will be utilized;
a) Top Down Flow Chart

Provides an overview of the activities and process flow

Helps to focus attention on specific aspects of the
process
b) Spaghetti Diagram

Capturing the flow- of either people or products
between stations

Ease to identify waste in motion, transportation and
inventory
c) Process Map

Depicts what is happening to the process

Useful to identify overproduction, defects,
overprocessing and inventory waste
Also to ease the process, equipments like sketch board, stop watches and video
camera will be used.
39
Table 3.1: Categories of observations
Activities
Connection
Flows

What is the expected result of the activity?

Will the results be clearly understood?

How is the activity structured?

Is the sequence determined?

Is the timing specific?

Is it clear if the expected outcome is not achieved?

Is this clear what the upstream activity/process delivers?

Can it be easily determined if the need it not met?

How the request is made?

Is there one way and one way only to make a request?

Are the flow paths specific? ( is there only one way?)

Are the flow paths simple as possible without excess
travel?
ii.

Are there loops or forks in the process?

Does the flow path minimize waiting in queue?

Is there any other waste in the flow path?
Self-Experience: To ensure a more holistic approach to understand
the changeover process better, the method of self experience will be used
where the observer will now involve directly in the changeover process or
„getting the hands dirty‟ by performing the actual task. This is a more active
approach where personal experience on performing changeover i.e. pre
improvement and post improvement will allow more understanding of each
steps, the time taken and the opportunity to identify the gaps and
improvements needed. It is important especially post-improvement
implementation to involve personally in the changeover as it is required to
understand if the improvement is in place, practical and sustainable before it
can be downloaded or done by other floor technicians. Key note for an
effective and fruitful self experience is to document all findings, revise and
continuously improvise before training others.
40
iii.
Informal Interviews: As it was not possible to involve all the
„changeover‟ technicians in the direct observations, quick fix to that will be
through conducting some informal interview by asking same specific
questions to each individual. Interviews with the personnel whom directly
involve in changeover could help in mining data and information on the
practice, ideas, setbacks and other vital information to help during the
improvement stage.
3.5
Limitation or setback of the Research Method used
Some of the limitation and setbacks that has been identified based on the
research method mentioned above are described as below;
i.
Information and data bias: Possible if the sample size collected is not
sufficient or „skews‟ especially during the direct observation. Data bias is
also possible if no proper scrutiny is done based on the interviews outcome.
ii.
Data Inaccuracy and information misinterpretation: Manual time study,
automated data versus manual data, and insufficient reading of previous
literatures could potentially deviate to inaccuracy of data and
misinterpretation.
iii.
Time consuming and constraint: Performing interviews and observation for
all 4 shifts could be challenging and consume time. The availability of
resources and aligning to personnel to conduct the changeover. Indulging in
self experience or performing changeover personally would be going for
basic training (mostly called Level 1) before being qualified to perform tool
conversion.
41
iv.
Possibility of human error: Many of the data collected are by human i.e.
mapping, time study, process study thus there are high tendency and
possibility of errors
3.6
Summary
This chapter discussed in depth the methodology that has been used in
throughout this project. The research methodology was structured in accordance to
collect both quantitative and qualitative data and information. Various different
approach and techniques will be use to collect as much sample information to ensure
no data integrity or any data bias issue.
CHAPTER 4
PROBLEM IDENTIFICATION
4.1
Overview
This chapter will describe in detail in the case study untaken for this project.
A thorough analysis of the overall changeover process will be covered; including the
historical background, justification, current state and last but not least the problems
or gaps identified.
4.2
Background and Justification
The core objective of this whole project as mentioned earlier is to reduce the
overall changeover time for a test handler from current 4 hours to a target goal of 0.5
hours. The motivation behind this project is driven due to the 3 reasons as below:
43
4.2.1
Low Tester Utilization
From the historical data collected from early Q2’2010 to the end of Q4’2010,
the average equipment (tester) utilization for good production use purpose is only at
70.6% per week where the remaining 30% of the time is lost due to the Non Value
Added activities or simply waste in Lean context. Figure 4.1 shows the breakdown of
the average tester utilization where almost 8.4%/ week or nearly 14 hours/week is
just lost due to changeover activity alone.
Figure 4.1: Breakdown of tester utilization
Also, a point to note is that the high Repair (downtime) around 10.5%/week
and Idling time of 6.5%/week also could be associated due to this ineffective
changeover process. Figure 4.2 on the other hand shows the corporate utilization goal
which highlights a gap of almost 20% in good utilization time itself. Lower
utilization indicates high cost and high cycle time which could be disastrous for the
survival of a High Mix Low Volume (HMLV) industry.
Figure 4.2: The corporate utilization goal
44
4.2.2
Drive for Flexible Manufacturing
Figure 4.3 shows an example of the historical data collected from Q4’2009 to
Q4’2010 on the product’s demand trend. Due to ever volatile customer demand, it is
challenging in both to predict and support the demands timely. Thus the best counter
action is for organizations to be adoptive and also flexible to response quickly to the
ever changing customer demand.
Figure 4.3: Example of demand trend
In actual current scenario, the factory shop floor is non-conducive to practice
lean or to drive flexible manufacturing due to the fact of;
i.
Equipment dedication policy (product family based) at the Test operation
area
ii.
The practice to batch build products from same family
iii.
The WIP linearity issue from upstream operation always hinders
continuous flow of materials to the test operation
Thus, to achieve World Class Manufacturing status, all the above elements need to
be improved and converted to:
i.
Continuous flow of both WIP and information
ii.
Rapid and Quick equipment/process changeover
45
iii.
4.2.3
Lean, flexible and eliminate waste at all levels
Driving Cost Competitive Advantage
The organization’s target cost of per unit chipset is at $5.44 but the actual
effective cost is currently at $6.04. The spike of $0.60 is due to multiple factor but
mainly cost by test operation and manufacturing overhead. Test operation is plagued
mainly by the high capital purchase with below goal utilization are major reasons
behind the higher cost per unit. Figure 4.4 below shows an example of cost
breakdown for a chipset product based on the Q1’10 finance analysis.
Nebula Cost Breakdown
7
6
Cost ($)
5
4
3
2
1
0
Goal ($)
Actual ($)
Core yield loss
0.17
0.18
Mfg OH
0.58
0.83
Test
0.38
0.57
Assembly
0.45
0.51
Package
1.44
1.46
Die
2.42
2.49
Figure 4.4: Cost breakdown of Nebula
With the incoming of higher production volume and the introduction of new
product series (NPIs), the need to purchase new capitals and collaterals arise much
quickly. Figure 4.5 shows an example with current scenario of low utilization and
46
high cycle time practice; this could result in additional capital purchase of 1 tester
quarter to quarter.
Figure 4.5: Inventory impact analysis
In summary based on the justification above, an optimized and flexible
manufacturing process flow will ensure improvement in productivity and cost which
is significant for continuous sustaining and leading the market segments compared to
the competitor.
4.3
Case Study Company
This case study will be based in Intel Kulim Microprocessor and Chipset
Operation (KMCO or KM5) located in Kulim High Tech Park, Kedah DA. Some of
the generic information of the case study company is summarized as below:
47
4.3.1
Background of Company
When Kulim High Tech Park (KHTP) was launched in 1996, Intel
Microelectronics Sdn. Bhd. was the 1st company to setup its facilities beginning with
KM1. Till date, Intel has invested more than RM3.7bil with land size of 75 acres in
KHTP alone. Figure 4.6 shows the topography of the Intel Kulim layout in KHTP.
Today with its significant growth, Intel Kulim has 3 major operating factories
(KM1,KM2 and KM5), 1 R&D facility (KM3) and a multi storey office suites
(KM6) with a grossing of 3500 employees.
Figure 4.6: Intel Kulim’s topography
4.3.2 Kulim Microprocessor and Chipset Operation (KMCO)
KMCO is a 250ku square feet factory building erected in year 2006 with the
initial plan to be the largest single platform chipset operation factory was mainly to
boost the production of some of the already in operation Intel factories in Asia.
KMCO as shown in Figure 4.7 started its full operation mid of year 2007 with 1st
chipset product code named ‘Cantiga’. Later in year 2008, with changes in business
strategy changes and priority rescheduling, KMCO started to support the high
volume and high value microprocessor platform when the mobile microprocessor
codenamed ‘Penryn’ was ramped up in KMCO to support Intel factory in Chengdu.
48
Today, KMCO is the 2nd largest factory in Intel’s Assembly-Test-Manufacturing
(ATM) ring of factories after Vietnam. KMCO is considered as a mix platform
factory supporting High Mix Low Volume (HMLV) products as breakdown below;
i.
High Volume = Microprocessor, demand of 500ku-2Mu/ week
ii.
Mix Volume = I/O or graphical Chipsets, demand of 200ku-1Mu/
week
iii.
Low Volume = Others chipsets, demand of <200ku/week
Figure 4.7: KMCO’s factory view
Due to the mix platform and volume structure, KMCO is considered a pivotal
resource for the organization. Nevertheless, with current structure it is herculean
challenge for KMCO to meet and deliver quality products with low cost/unit, high
tool utilization and lower cycle time.
4.4
Product Background
Today though KMCO supports production of 9 different major products with
hundreds of sub-line items and with 5 new NPI products in coming, but the focus of
this case study will be on 2 of the major chipset products called Nebula Peak and
Nexus Peak.
49
Nebula and Nexus as shown on Figure 4.8 are integrated chipset units build
to support the main stream microprocessors especially to function for the I/O and
graphics. Nebula and Nexus are first of its kind where both the silicon technology
and the processes are apparently new to Intel with no prior successor product to refer
on. Nebula and Nexus are chipset units built to serve a similar function but to serve
a different market segment. Nebula is introduced for the desktop and server market
segment while Nexus is for the mobile market segment. Both the products have
significant different of market demand and this trend swing quarter to quarter.
Nebula Peak
Nexus Peak
Figure 4.8: Nebula and Nexus sample units
From the analysis of the physical attribute, both this product are significantly
same in many ways but differ mainly on the package size where Nebula is slightly
bigger compared to Nexus due to the nature of use for this product. Table 4.1 shows
a summary of the physical attribute difference between these products.
Table 4.1: Physical attributes comparison
50
Figure 4.9 shows a graphical illustration of the product roadmap where Nexus
and Nebula will be in production till the mid of year 2011 until it is replaced by a
more enhanced next generation chipsets codenamed ‘Orio’ and ‘Wicky’ respectively.
This products are the followed on chipset units with main difference on the 3D
graphics function accelerator. Table 4.2 shows the Orio and Wicky physical attribute
compared to Nebula and Nexus.
4.5
Process Background
The process that this case study will focus is the Test Operation area. Figure
4.10 shows a full blown process in respect to Nebula Peak and Nexus Peak. Test
operation area is considered as one of the most complex and costly operation in a
semiconductor process flow.
Orio
Wicky
Figure 4.9: Product roadmap illustration
51
Table 4.2: Physical attributes comparison to NPI
Test operation is the area where units undergo ‘Functional’ testing, where the
Die Under Test (DUT) are exposed to multiple electrical test instructed by the test
programs under extreme temperatures, typically from the range of -5’C (Cold Test)
to +115’C (Hot Test). In Lean perspective, the test operation is considered as a Non
Value Added phase as the operation does not significantly change the physical
attribute of a device but only test to screen out reject. But nevertheless, this is still
considered necessary operation as it is needed to screen out silicon failures due to
wafer fabrication and mechanical defects due to process and equipments.
A test operation normally consist sets of different test sub modules depending
on the product it supports. Test modules are a combination of different test
equipments like Tester unit, Test handler, Tester Interface unit (TIU) , Station
Controller (SC) and product centric Test Programs(TP).
52
Figure 4.10: The process flow
4.6
Equipment Background
The equipment in focus for the changeover is called the M4542AD Dynamic
Test Handler or also known as the Extreme Test Handler as shown in Figure 4.11.
The test handler is part of a test module and an illustration of a test module setup is
as shown in Figure 4.12. The test handler together with the tester is operated for
electrical testing using device specific test programs and TIU.
Developed by Advantest in year 2003 with an average market price around
USD $700k to USD$1.5Million. The equipment became an instant hit in
semiconductor industry especially for those with Ball Grid Array (BGA), Quad Flip
Chip (QSC) and Land Grid Array (LGA) type of packages. The tester handler is able
to perform testing from single unit to a maximum of 8 units per execution. The
handler has capability of 6000 devices/hour throughput with index time averaging
1.0sec/unit. This tester handler is able to operate at extreme temperature ranging
53
from -60’C to +150’C including room temperature. The Extreme Test Handler is
known mainly for 2 reasons; i) motorized pickup arm that automatically adjust its
stroke and speed to prevent device from cracking and chipping, ii) the pressure
control mechanism that enables constant and uniform pressure to devices.
Figure 4.11: The M4542AD Dynamic Handler
Figure 4.12: Test module setup
In a test module setup, the most complex equipment is the test handler since
most operation, process and mechanical movements happen either in or at the
54
handler portion. 95% of a changeover activity is due to the test handler alone with
the highest constraint of the hardware part setups.
Besides the test module, the other subcomponents are which is indirectly
related to this changeover is the Automated Equipment Performance Tracking
(AEPT) which is the core tracking system built in the Station Controller. AEPT
system collects, analyze and stores every activity occurs at the equipments in a
centralize database called the MARS. The AEPT system are designed based on the
SEMI E10 standard where the data collected are categorized by Production, Idling,
Engineering, Schedule Downtime, Unscheduled Downtime and Non Schedule time.
4.7
Changeover Process Historical Study
The corporate goal for allowed changeover duration is 1.1 hour / conversion
while the actual average time taken to complete a test module changeover from
Nebula to Nexus (or vice versa) is around at 4.0 hours/conversion. Figure 4.13
shows the changeover time trend based on the data collected from Q2’10 to Q4’10.
Figure 4.13: Time trend of conversion from Q2’10 to Q4’10
55
The time trend graph can be divided in to 3 phase:
a) Ww13 to ww25 this is considered as the infant stage where changeover
of the tester handler was relatively new concept, engineers and
technicians were slowed trained to be familiar with the equipment. At this
phase, changeover average at 5.0 hours/conversion
b) Ww26-ww39 this is considered as the intermediate stage. More formal
training emphasize for technicians. Minor improvement efforts like
optimizing recipe profiling help to save around 15 minutes of changeover
time. Changeover average at 4.75 hours/conversion
c) Ww40-ww52 better trained and experienced technicians. Also, some
improvements in the TP download time. Changeover average at 4.0
hours/conversion.
4.8
Changeover Process Flow in Detail
The generic changeover flow is shown in Figure 4.14 and the details of each
of the stages are further explained in detail below.
a) Pre Changeover Activities – though not part of an actual changeover
process but the activities are important and ‘need to happen’ to ensure the
equipment ownership is transferred from ‘Production’ to ‘Engineering’. Some
of the significant activities involved in this phase are summarized in Figure
4.15.
56
Pre Changeover
Activities
AEPT Log In
Start of Conversion
Preliminary Soft
Setups
Hardware Parts
Setup
TIU replacement
PnP 'Teaching'
Process
Dry Cycling with
Mechanical Units
8 main
generic
stages
involve in
the
changeover
TP download
Standard Unit
Run
Wrap Up
Activities
AEPT Log Off
End of Conversion
Changeover
Complete
Figure 4.14: The generic changeover flow
57
Supervisor forecast and
decide on the changeover
need
Supervisor identifies the
available technician from
the team
Supervisor and technician
identify the equipment to
be changeover
Supervisor communicates
to technician on the
expectation and other
changeover info
Supervisor ensure no lots
are queued to be
processed by the identied
equipment in the system
(SYSTEM BLOCK)
Supervisor communicates
to operator complete the
last lot and not to load
any new lots
Once last lot completes,
Operator peforms all neccesary
activities called 'End Lot' activity
The operator performs
basic housekeeping at
equipment area
The operator log off the ID
and look for the technician
to handover the
equipment
Figure 4.15: Generic activities in pre changeover phase
i.
Based on the current WIP level, tool status and /or priority, Supervisor
will assess the need for equipment changeover. This is normally done
during the early of the shift and/or mid shift hurdle meeting or when
triggered by the Manufacturing Integrator ( normally on shipment
priority)
ii.
Once supervisor decides on the need of changeover, he/she will
identify an available and qualified technician in the shift to perform
the changeover
iii.
Technician and Supervisor with mutually walk the line and identify an
equipment to be changeover
58
iv.
Once the equipment has been mutually agreed and identified, the
supervisor will communicate with the operator attached to the
equipment to complete the last lot process. Normally, supervisor will
reassign the operator to another machine once the lot complete
v.
In the WIP system, supervisor will block or cancel any new lots in
queue to be processed for the identified equipment
vi.
Once the last lot has completed processing at the identified
equipment, the operator will proceed with the following end lot
process (by sequence);
 Physical move out last lot
 End lot process in the system
 Physical segregation of good and reject units
 Counting and visual verification
 Reject unit process include use red tray, bundle tie with bin
card
 Good unit process include preparing partial tray, bundle tray
and bin card
 Documentation on shift tracking sheet
 Place good units into shipment trolley and push to next station
 Send the reject units to RIMS room for disposition action
vii.
Operator will perform basic housekeeping especially on the work
bench before handover the equipment to the technician
viii.
Operator will log off the ID as user and will go look for the technician
to inform on the equipment availability for changeover
b) Preliminary Soft Setup – are all the activities performed or occurred once
the Technician logs in the ID and the AEPT state changes from ‘Production’
to ‘Schedule Downtime- Conversion’. It is considered as the first official
changeover phase and is summarized in Figure 4.16.
59
Completion Pre
Changeover
Activities
Technician logs-in ID
and change AEPT
from 'PROD' to 'SDTConversion'
Change 'sticky' tag
and/or Barricade
work area
Perform equipment
temperature setup
Collect changekit
from maintanence
room
Collect toolset from
toolshop
Back to equipment area
and standby to start
hardware changes
Figure 4.16: Generic activities in preliminary soft setup phase
i.
The technician will log in the ID and subsequently change the
‘Automated Equipment Performance Tracking’ or AEPT state from
‘Production’ to ‘Schedule Downtime– Hard Conversion’ as shown in
Figure 4.17.
TO
Figure 4.17: The change of AEPT state
ii.
As for visual indicator, technician will change or place the ‘sticky’ tag
on the equipment as shown in Figure 4.18 and also once the AEPT
state is change to Schedule Downtime, the ‘Andon’ lights towered on
top of the equipment will automatically change from ‘GREEN’ to
’RED’ light as shown in Figure 4.19. To ensure safety and minimize
disturbance, some technician prefers to use barricades as shown in
Figure 4.20 around the equipment area.
60
T
O
Figure 4.18: The change of ‘sticky’ tag indicators
Figure 4.19: The ‘Andon’ Light
61
Figure 4.20: The use of barricades
iii.
Normally, equipments used prior for production will be under extreme
temperature which could range from hot temperature (+115’C) or cold
temperature of (-5’C). Thus, the next step in this phase is called the
equipment cool down/ warm up activity where the equipment
temperature will be set to be in the range between 60’C. The
technician will need to reset the temperature in the GUI monitor at the
front end of the handler as shown in Figure 4.21 and also to manually
turn off the ‘TEMP’ button as shown in Figure 4.22. This is an
important step to ensure safety of both the technician and the
equipment itself.
Figure 4.21: Reset temperature in the GUI
62
Figure 4.22: Manual turn off the by pressing ‘TEMP’ button
iv.
Next, technician will walk to the maintenance room to collect the
changeover kit box which contains all the hardware parts that will
need to be replaced as shown in Figure 4.23.
Figure 4.23: Example of change kit box with hardware parts
63
v.
Lastly, the technician will need go to the tool shop to collect the
necessary toolset needed for the changeover. Normally toolbox will
contains set of hex keys and screw drivers as shown in Figure 4.24.
Figure 4.24: Complete tool sets
c) Hardware Part Setups- is the 2nd phase of the changeover and considered as
the most complex activity that causes the overall bottleneck in the changeover
process. There are 12 major hardware parts involve in this changeover and
are summarized in Table 4.3.
Table 4.3: Hardware parts involve in the changeover
No.
Hardware
Hardware Image
Function
Changeover detail
Name / Qty
LD: To pick the
Loader/
1
Unloader pick
up heads
(4pcs/set)
units from input
tray to heat plate
ULD: To pick
units from exit
plate move them
unloader tray
Each pick up head
has 1 screw need to
be loosened in
order to be
removed. A total of
4 screws are
involved. To
remove and replace
64
use hex keys.
A total of 8 screws
2
3
4
Loader/
To ensure units
are involved before
Unloader
are pick up in
the 4 blocks can be
Y-pitch blocks
correct Y-axis
removed and
(4pcs/set)
position
replaced.
Loader/
To ensure units
A total of 2 screws
Unloader
are pick up in
are involved to
X-pitch blocks
correct X-axis
remove and replace
(2pcs/set)
position
the blocks.
Plate to place and
Changeover of the
Heat Plate
to heat/soak units
heat plate involve
(1 pc/set)
before they are
push and lock of 2
moved to the real
clamps
functional testing
To pick up and
Loader/
Unloader
5
buffer/exit
pick up heads
(4pcs/set)
transfer unit from
This pick up heads
heat place pocket
can only removed
to buffer stage
and replaced using
and the unloader
a special device
pick up from the
supplied by the
buffer to exit
supplier
plate
6
Loader /
Unloader
Only 1 screw for
Units from the
each buffer plates
65
Buffer stage
heat plate are
need to be loosen
for Front &
transferred into
before it can
Rear
the front buffer
removed and
(4pcs/set)
stage and transfer
replaced
them into
chamber (test
head)To transfer
units to the
contact arm unit
Act as a stopper
to ensure the
7
assembly
Are removed and
Stopper for
mechanism of the
replaced by
input/exit
loader and
loosen/tightening of
assembly
unloader of buffer
1 screw for each
(2pcs/set)
stage to be in
stopper
correct location
and avoid pick up
issues
Each chucks has 4
clips and in total of
16 clips to be
removed. And, this
includes removal
Contactor
8
Chucks with
nest
(4pcs/set)
To pick up units
and replacement of
at buffer stage
2 screws for each
loader and test the
chuck. Before a
units at TIU for
new chuck is
functional test
replaced, a
thorough cleaning
of the pedestal and
nest is required.
Also to ensure no
temperature issues,
a temperature
66
calibration is
needed. a full
greasing is needed
especially on the
pedestal
The exit plate is
Is use to store
9
Exit Plate
and position
(1pc/set)
tested units from
buffer stage
locked by 6 clamps
and technician
needs to unlock the
clips to remove and
replace with new
plate
Prevent
condensation
from forming on
devices when
10
Unsoak Plate
performing low
(1pc/set)
temperature tests,
prior to
Need to remove 4
screws from each
end to replace a
new plate
transferring the
devices to
customer trays
11
Transfer pick
To pick up empty
To replace, need to
up assembly
trays from buffer
loosen 2 screws and
(1pc/set)
stoker to
remove 1air tube
input/output
stoker
Socket
12
To replace,
cleaning plate
Plate that place
technician need to
(1pc/set)
the surrogate
remove 2 clamps
cleaning device
and replace with
67
(SCD) coupons
new plate.
which are used to
clean TIU pogo
pin. This is a run
based activity
where cleaning is
auto performed
for every
5000units.
d) Replacement Test Interface Unit (TIU) – changing of the interface
component that links the test handler to the tester unit. TIUs as shown in
Figure 4.25 are usually kept in a separate maintenance area called TIU room
with trained technicians dedicated to handle and maintain them separately as
they are prone to pin defects and require special ESD-care. Each TIU are
uniquely designed based on product characteristic especially on the package
size and the pin counts. Figure 4.26 shows a summary of the generic steps
involve in a TIU replacement with detail explanations below.
Figure 4.25: Example of a TIU
68
Complete hardware
part setups
Walk from equipment
area to TIU room
Request and collect
TIU
Perform pre-physical
setup
Perform pre-system
setup
Walk back to
equipment area
Undock the old TIU
Replace with new TIU
Perform post-physical
setup
Collect old TIU and
send back to TIU
room
Perform temperature
setting
Perform post-system
setup
Return to equipment
area standby for PnP
Teaching
Figure 4.26: Generic flow of TIU replacement phase
i.
Technician will need to walk from the tagged equipment area to the
TIU maintenance room to request for a product-specific TIU i.e.
Nebula Peak
ii.
TIU maintenance room administrator will review the request item in
the auto-tracking system to locate the exact rack the TIU is stored
iii.
Once the requested TIU is found, administrator will log out the item
from the inventory and handover it to the technician
iv.
Technician will take the TIU and walk back to the equipment area.
The technician will place the TIU on the table and walk to the rear of
the handler where the current TIU is docked on the test head as shown
in Figure 4.27.
69
Figure 4.27: Rear of the handler
v.
By pressing the ‘Clamper’ button as shown in Figure 4.28, the current
TIU will be released from the test head. The pilot lamp will indicate
once the TIU has completely released
vi.
Technician will next rotate the position wheel anti-clockwise to move
the TIU and test head in an downward position as shown in Figure
4.29
vii.
Technician will next rotate the position wheel anti-clockwise to move
the TIU and test head in an outward position
viii.
Once the test head reaches to the maximum position, the stopper will
automatically halt the movement
Figure 4.28: The ‘Clamper’ button
70
Figure 4.29 The position wheels to move the TIU and test head
ix.
Next, the technician will need to press the ‘PB Free’ button located
below on the handler to ensure the hydraulic releases the TIU from
the test head
x.
Then the technician will undock the TIU in an upright and by holding
the handle
xi.
Technician will bring this TIU to the front of the handler area where
the table is to place this TIU and pick the new TIU
xii.
Technician will move back again to the rear of the handler, place the
new TIU in the correct orientation on the test head and lock it as
shown in Figure 4.30
xiii.
Next, the technician will need to press the ‘PB Lock’ button located
below on the handler to ensure the hydraulic lock the TIU from the
test head
71
Figure 4.30: Docked TIU on test head
xiv.
Once technician validated the TIU is fully secured, then need to rotate
the position wheel in clock wise motion until the test head hits the
automatic stopper.
xv.
Then once the ‘Clamper’ button is triggered again, the TIU will be
position on the test head and will be auto-locked
xvi.
Once completed, the technician will turn on the temperature by
changing the setting in the GUI monitor from ‘No Temp Control’ to
‘Temp Control’ and press the ‘TEMP’ button to turn on temperature
fully. With this step, the equipment heater will ramp the temperature
from the current +60’C to a temperature of +115’C ( +/- 5’C)
xvii.
Technician will next return the undocked TIU back to the
maintenance room for maintenance and storekeeping
e) Handler Pick and Place (PnP) ‘Teaching’ - is the process step performed
by the technician to calibrate settings of the newly replaced hardware parts.
Teaching process is required to ensure the contact between the test handlers
and the device under test is fully optimized to avoid intermittent temperature
fallouts and PnP errors which could cause both quality and reliability issues
during real production runs. Figure 4.31 shows the generic flow of this phase
with detail explanations as below;
72
Complete TIU
replacement
activity
Walk to EIMS
room to collect
Standard Units
Pre physical
setups
Actual
'Teaching'
Process
Temperature
Stabilization
Pre system
setup and
configuration
Post system
setup and
configuration
Post physical
setups
Standby to start
the Mechanical
Unit Validation
Figure 4.31: Generic flow of PnP teaching phase
i.
Technician will go to the EIMS room to collect a tray full of Standard
units (only 4 units are needed for the PnP teaching while other for
later validation
ii.
Once back to the equipment area, technician will need to ensure that
all mechanical movements of the handler has fully stopped
iii.
At the front end of the handler’s GUI monitor, from the main
operating screen choose to access into the Command menu.
iv.
In Command Menu option, choose Teaching Mode Setting screen and
access into the Teaching Execution Condition files where commands
for teaching are set
v.
Based on the Teaching documented available (prepared by engineer),
manual settings are done in the Teaching Stroke Limit screen. The
stroke limit refers to the stroke of the contact press to the test socket
vi.
Next, return to the Command Menu screen and from the dropdown
list, choose the Teaching Sequence button as shown in Figure 4.32
73
Figure 4.32: The command menu
vii.
The technician then places 2 Standard units on the front and rear
buffer stage loader respectively
viii.
On the Teaching Sequence Start screen, technician will need to
specify a target teaching unit mode where the units are placed either
Contact Front or Contact Rear and once confirmed technician will
choose ‘Start Teaching Sequence’ and activate the ‘Teaching Data
Registration’ screen as shown in Figure 4.33
ix.
Normally before the 1st teaching sequence can be executed, the
equipment will wait for the temperature to stabilize before it proceed
with the PnP sequence
x.
In the Teaching Data Registration’ screen, the motor contact point
setting is performed and once complete, the ‘Register Teaching Data’
button is selected
xi.
Technician will repeat the above steps for the remaining units in the
buffer stage
74
Figure 4.33: Starting the teach sequence
75
Figure 4.34: Completing the teaching sequence
xii.
Once all teaching has been completed with sufficient data and high
confidence level,
the technician will select the ‘Teaching Has Been
Completed’ to exit the Teaching Data Registration and the Teaching
Sequence screen as shown in Figure 4.34 above
xiii.
Lastly the technician will select the ‘Operating State’ button to exit the
PnP Teaching GUI and return to normal mode to end this process as
shown in Figure 4.35
xiv.
Technician will then remove the 4 units from the buffer stage
xv.
Place the units on the tray and send them back to the EIMS room
76
Figure 4.35: Exiting the PnP teaching phase
f) Mechanical Unit Validation – is also known as ‘Dry cycling’ phase where
the changeover success will be validated using samples of mechanical units
or unit’s without any die attached to it. This is a 1st level validation process
where the mechanical units will be cycled as per normal process but without
the use of the actual test program. The importance of this validation phase is
to ensure all the hardware components fixed are operating as per optimum
requirement and no intermittent issues or other mechanical or process related
issues like stains on die issues. Figure 4.36 shows a generic flow of this
phase with detail explanation as below;
77
Completed the PnP
Teaching
Walk to Tool Shop
to collect 5 trays of
Mechanical Units
Perform prevalidation visual
inspection
Temperature
Stabilization
Perform prevalidation physical
setup
Perform prevalidation system
setup
Actual Dry Cycling
Process
Perform postvalidation system
setup
Perform postvalidation physical
setup
Standby for TP
download
Return units back
to tool shop
Perform postvalidation visual
inspection
Figure 4.36: Generic flow of dry cycling phase
i.
Before starting the ‘Dry’ cycling phase, technician will go to the tool
shop to request for 5 trays of Mechanical Units
ii.
Once acquired, technician will return to the equipment area and will
begin will the pre-validation visual inspection using the ‘naked’ eyes.
This pre-validation inspection is important to ensure all the units are
screened through and only defect-free units are cycled in this phase. If
any defect units found in this step, technician will need to return the
units back to tool shop and collect the replacement units (Note: if prevalidation visual inspection is skipped, any damage found later during
post-inspection will be questionable as the root cause will be
unknown and delay the overall process)
iii.
Once visual inspection completed, the technician will proceed with
the system setup where from the Operating State in the handler’s GUI,
78
the technician will select the Mechanical Unit Validation option and
Click ‘OK’
iv.
While waiting for temperature to stabilize, the technician will load in
the trays into the input stage and empty trays in the buffer stage and
ensure no other units or trays inside the handler
v.
Once the temperature stabilization done, the handler will start to
process the mechanical unit validation where each and every unit will
be process as per normal but without the actual testing enabled
vi.
The quick dry cycling using the 135 units will test the new hardware
fixtures and trigger alarm if any parts are malfunctioning
vii.
During the dry cycling , the handler auto system will capture all the
mechanical part functionality in its data log for technician to review if
any issue
viii.
Once all the 135 units have completed the dry cycling, the completion
alarm will trigger for technician to end process
ix.
Technician will remove the 5 trays from handler’s output stage and
place them on the work desk
x.
Then later the technician will screen through the auto log for any
failure report ( normally this is done during the cycling itself)
xi.
Technician will continue to perform the post-validation visual
inspection to screen for gross defects like solder ball damage,
scratches on die cap or substrate or any foreign residue or stain on the
units
xii.
If any defects are seen through this visual inspection, the technician
will relook into the system log file to understand the failure root cause
and will have to re-run the whole dry cycling process
xiii.
Once a 100% pass is obtained, then technician will end the Dry
Cycling process by logging off the ‘Mechanical Unit’ validation
option and back to Operating State GUI
xiv.
Technician will collect the units into the tray and send them back to
the tool shop
79
g) Test Program (TP) download – is a set of electrical instructions for the
equipment to perform required testing to the device under test (DUT). TP are
software coding developed by product content experts during the R&D stage
and the complexity of a TP depends on the number of test lines. All of the TP
developed are confidential as they contain information on product technology
and characteristic which are classified as Intellectual Property (IP). A
product can have one or many different TPs which depend heavily on the
type of product, the testing requirements or purpose. All authorized TPs are
stored securely in the centralized database system and TP are retrieve from
this database to equipment through a process called TP downloading. In this
phase, since Nebula and Nexus have a different TP version, thus each time a
changeover is performed; the TP will need to be downloaded before the
equipment is handover back to the Production. Figure 4.37 shows a generic
flow of this phase with detail explanation as below;
Completed
Dry Cycling
Process
Perform TP
reset
Technician
input info into
ELI window
System will
Initialize SC
Actual TP
download
process
TIU
initialization
Standby to
start Standard
Unit Validation
TP download
complete
Figure 4.37: Generic flow of TP download phase
80
i.
Technician will reset the ASTRA framework to basic Station
Controller version to enable the TSS nullify the current TP in the
equipment database as shown in Figure 4.38
ii.
Once complete of current TP reset, turn back CTSC to ASTRA
framework
iii.
Then click on the ‘Initialize Station Controller’ option which will
purge all previous TP related info and prepare the a new CTSC screen
without any TP downloaded
Change
Figure 4.38: Change of CTSC environment
iv.
Once Initialize SC is complete, the system will prompt with
‘Engineering Lot Info’ window which will require technician to input
information of TP name, TP version, product/device name etc. Once
81
the required info is correctly input, the system will auto search the TP
and start the downloading process
v.
A window will prompt showing the duration and status of the TP
download as shown in Figure 4.39
vi.
Once TP download complete, another window will prompt technician
to select the Summary Type; options are STD1 or 2A. For changeover
or engineering, STD1 is used as shown in Figure 4.40 while 2A for
production retest runs
Figure 4.39: TP download status
Figure 4.40: Choosing STD1 summary
82
vii.
Once STD1 summary has been selected, the downloaded TP will be
transferred to the Handler system to initialize the TIU ( this is to
ensure the TIU is able to communicate with the tester module )
viii.
Once the Initialization has complete, CTSC will prompt message that
the system is ready for testing and awaiting technician to click ‘Start
Testing’ at handler
ix.
This complete the TP download process
h) Standard Unit Validation- is the final validation process of the changeover
where samples of good engineering units called Standards (equivalent to
good production units) are tested under production environment with actual
TP and temperature settings. 1 full tray containing 27 units will be cycled
and tested with requirement of 100% yield with no electrical failure rejects.
Standard Units are categorized as High Value Inventory (HVI) thus the
availability are limited and stored securely in Engineering Inventory
Management System (EIMS) locker room with controlled access. The generic
flow of this phase is shown in Figure 4.41 with detail explanations as below;
i.
Technician will use the Standard unit collected earlier during the PnP
teaching phase
ii.
Technician will start off by performing some basic inspection like
units orientated correctly inside the tray pocket and no other
abnormalities on the STD units
iii.
Once complete, technician will load this tray into the handler’s input
source and on the GUI system, technician will choose Standard Unit
Validation option
iv.
On the CTSC screen, technician will click ‘OK’ and press the ‘Test
Now’ button on the GUI screen to begin the validation process
v.
Equipment will initialize and the temperature will need to stabilize
before the validation testing can proceed
83
Completed TP
download
Perform prevalidation basic
inspection
Perform prevalidation system
setups
Actual validation
and testing
process
Temperature
Stabilization
Perform prevalidation physical
setups
Perform postvalidation system
setups
Perform postvalidation physical
setups
Complete process,
return units back
to EIMS
Standby to Wrap
Up changeover
activities
Figure 4.41: Generic flow of the standard unit validation
vi.
Once initialization and temperature stabilize, the actual testing will
proceed for all the 27 units
vii.
Units will be tested with good units place into Bin1 and rejects units
into Reject tray at the handler
viii.
Once all units has complete testing, an alarm triggers to indicate all
units are in output trays
ix.
Technician will collect the units back from the output trays
x.
For STD unit validation, all 27 units should be in the Bin1 output tray,
else technician will need to troubleshoot to understand the issue
xi.
Once validated, technician will clear log and reset the GUI to return to
Operating state window
xii.
Technician will place all units into tray and send back to EIMS room
84
i) Wrap Up Activities - Is the final set of activities to complete the changeover
process before the equipment is handover back to production. Figure 4.42
shows a generic flow of this phase and detail explanation as per below;
Completed Std Unit
validation
Perform basic
housekeeping
Inform
supervisor/operator
of completion
Return change kits
and toolsets
Change 'sticky' tags
to 'UP'
Change AEPT state
from 'SDTConversion to
'PROD'
Completion of
changeever
Figure 4.42: Generic Flow in Wrap up Activities
i.
In this last phase, technician will start by performing basic
housekeeping activities like cleaning the work desk from the dirt,
debris and other waste items from the previous activities. The
housekeeping is needed to ensure the work desk is clean and clear
during normal production operations
ii.
Next, technician will rearrange the removed hardware parts back into
the change kit box and the toolsets into the tool box. Technician will
85
return the changekits and toolsets to the maintenance room and
toolshop respectively
iii.
On the way, technician will inform the supervisor who sits in the
supervisor room or else any operator available to start loading
production lots
iv.
Once back to equipment area, technician will change the sticky tags
from ‘Down’ to ‘Up’
v.
Once the operator is available or the supervisor has acknowledge the
changeover completion, the technician will change the AEPT state
from ‘Schedule Downtime-Conversion’ To ‘Production’ and
technician will log off his id as user
vi.
Changeover process is complete and equipment is successfully
handover back for Production use
4.9
Problems and Gaps Identification
Table 4.4 below shows a general summary of the current changeover process
discussed previously and the average duration to complete each of the activity. Thus,
there is a changeover gap of 4 hours between the last good run lot to the next good
run lot. The impact of this time-consuming changeover has been assessed earlier
from the point of utilization, cycle time and cost, thus this justifies the need to
improve the current process.
The ultimate objective of this project is to reduce the changeover duration by
optimizing the processes and activities involved. This is possible by identifying and
systematically eliminating waste or non value added activities in each of the current
changeover process phase.
In the next following discussion, problems and gaps identified at each stage
will be highlighted and analyzed in detail.
86
Table 4.4: Summary of the changeover process
Sequence
No.
Activity and
Milestone
Pre Changeover Activities
Details
Average
Time
(minutes)
Supervisor communication and alignment with
technician
End lot process for the last production run
Not included
in changeover
time
Official start of changeover process with the
change in AEPT
1
Preliminary
Soft Setup
Activities
2
Hardware Part
Setups
Tagging of equipment i.e. sticky pad or
barricading area
Preparation of change kit and toolsets
12 major hardware part setups
Bottleneck of the overall changeover process
TIU
replacement
Interface unit that need to be replaced
3
4
PnP 'Teaching'
Process
5
Dry Cycling
with
Mechanical
Units
6
TP download
7
Standard Unit
Run
8
Wrap Up
Activities
6
160
11
Required calibration process each time
hardware parts are replaced
9
1st validation process on the hardware part
setup
Using 5 trays of mechanical units
19
Ensure end of cycle, 100% pass with no
mechanical defects
Software coding to instruct tester to perform
electrical testing
2nd validation performed using good
production samples of 1 full tray
Validation under real production atmosphere
Ensure all units pass with 100% yield
Housekeeping and cleaning up work area
Official end of the changeover activity by
change in AEPT
TOTAL
14
15
6
240
Internal
Time
87
4.9.1
Pre-Changeover Activities
As mentioned earlier, though not an actual changeover phase but the pass
over coordination and communication activities that takes place at this phase are
equally important. Figure 4.43 shows the generic flow of this phase with the
estimated time taken to complete each of the activity. Some of the common problems
and gaps observed during this phase are listed in the Table 4.5.
7
minutes
5
minutes
5
minutes
10
minutes
3
minutes
30
minutes
•Supervisor finds and identify the technician
•Communicate on changeover request
•Supervisor and Technican decide and identify of the equipment
for changeover
•Supervisor communicates with operator to complete the last lot
•Assign operator to other equipment upon completion
• Ensure no new lots or WIP in queue to be process for the
identified equipment
•Operator performs end lot process
•Include move out lot, reject preparation, documentation etc
•Operator performs basic house keeping activities
•Operator log off ID as user and passover the equipment to
technican
•Total EXTERNAL time spent for the Pre Changeover Activites
Figure 4.43: Time study of pre changeover activities
88
Table 4.5: Problems identified at the Pre-Changeover phase
No #
Problems or Gaps Identified
Many non standard practices observed especially:

Changeover request done on ad-hoc basis across all
shifts
1

No proper process to define or identify the equipment
for changeover
Potential delay of overall changeover process due to;

misunderstanding and miscommunication
waste in motion, walking and transportation
The 13 minutes operator spends for end lot process and
2
housekeeping are actually opportunities for technicians to
capitalize in preparing upfront for the changeover. Today
technician idle during this period
4.9.2
Preliminary Soft Setups
The preliminary soft setup activities are the 1st phase of actual changeover
activity that involves 4 sub activities. Figure 4.44 shows a generic flow of this phase
and the time taken to complete each of the activity. Based on the multiple
observations done, most of the problems identified at this phase are related to NVA
practices and are summarized in Table 4.6 below:
89
0.5
minutes
0.5
minute
3.0
minutes
2.0
minute
6 .0
minutes
•Technician log in ID as user
•Change AEPT state from 'PROD' to 'SDTConversion'
• Changing Sticky pad from 'Up' to 'Down'
•Barricading work area
• Walk to maintanence bay and toolshop to
collect changekit and toolsets
•Setup to cool down/warm up equipment
•Temperature stabilize
•Total INTERNAL time spent in Preliminary
Soft Setup activities
Figure 4.44: Time study of preliminary soft setups
Table 4.6: Problems identified at the preliminary soft setup phase
Problem
Problems or Gaps Identified
No #
Existence of many non standard practices especially:

1
Different technician /shifts practices different work
flow or sequences

Some technician practice on deploying barricades
around work area while some don’t
Potential delay overall phase > 6 minutes
2
Many of the INTERNAL activities practiced today does not
VALUE ADD directly to the changeover process especially;
90

Practice of placing barricades/sticky pads

Walking to and fro between tool shop/maintenance
room and equipment area to collect toolset/change kits
If able to remove or performed as EXTERNAL activity,
potential able to salvage 3.5 minutes
Temperature setting (cool down / warm up) practiced as
EXTERNAL activity which delays ~2 minute. Technician
idle during this period
3

If able to perform this as a EXTERNAL activity
potential reduction of 2.0 minutes
4.9.3
Hardware Part Setups
Hardware part setups contribute to the major bottleneck concern in this
changeover process. Figure 4.45 shows a summary on the generic hardware parts
changed as per practiced today with the average time taken to complete each of the
activity. From the table, the major bottleneck part change is the replacement of the
contactor chucks and Table 4.7 shows the current changeover flow for this part.
Also, the other problems identified at this phase are summarized in Table 4.8:
91
Loader & Unloader Pickup Heads
•4 pcs/set
• 4.0 minutes/set to replace
Loader & Unloader Y-pitch Blocks
•4 pcs/set
•4.0 minutes/set to replace
Loader & Unloader X-pitch Blocks
•2 pcs/set
•4.0 minute/set to replace
Heat Plate
•1 pc/set
•0.5 minute/set to replace
Loader & Unloader Buffer/Exit Pick Up Heads
•4pcs/set
•8.0 minutes/set to replace
Loader & Unloader for Front & Rear Buffer Stage
•4pcs/set
•4.0 minutes/set to replace
Stopper for Input/Exit Assembly
•2 pcs/set
•2.0 minutes/set to replace
Contactor Chucks
•4 pcs/set
•120.0 minutes/set
Exit Plate
•1 pc/set
•0.5 minute/ set
Unsoak Plate
•1 pc/set
•0.5 minute/set
Transfer Pick Up Assembly
•1 pc/set
•13 minutes/set
Soak Cleaning Plate
•1 pc/set
•0.5 minutes/set
Total time for Hardware Changeover = 160 minutes
Figure 4.45: Time study of hardware part changes
92
Table 4.7: Sequence of the Contactor Chuck Replacement
No
Actual Photo(s)
Description
Time
taken(min)
Remove all
current contactor
1
chucks from the
8
handler
Replace the old
2
chucks with the
new chucks from
the change kit
box
3
93
Collect new
chucks and
3
arrange
1
accordingly on
work desk
Clean each chuck
with cotton swap
4
dipped with
16
Isopropyl
alcohol(IPA)
Remove nests
5
from all the
8
chucks
Remove all the 6
6
pins from each of
the nest
3
94
Again, use cotton
7
swap and IPA to
clean each of the
12
nest
Use the air gun to
8
dry each of the
4
nest
Place all 6 pins
9
back into each of
4
the nest
Clean the nests
10
again with cotton
swap and IPA
8
95
Remove the
contactor
components (
11
4pcs) from each
4
of the chuck and
clean them with
IPA as well
Use the air gun to
12
dry each of the
4
components
Fix the contactor
13
components back
to the chucks
6
respectively
Fix the nest back
14
to respective
chucks
8
96
Use clean cloth to
15
wipe residue on
the contactor and
4
the chuck
Walk to the rear
on the handler
and place all
16
newly setup
1
chucks at the
bottom of the
handler
Use clean and dry
cloth to wipe and
17
clean the internal
4
part of the
handler’s chuck
With cotton
swap, perform
18
greasing to the
internal chuck
parts
2
97
Clean and grease
19
the contactor
2
chuck calibrator
Insert calibrator
into each chuck
20
socket to re-
6
calibrate the
chuck heads
Once calibration
done, fix
respective
21
contactor chucks
12
to the chuck
socket and ensure
all clip are locked
Total
120
minutes
98
Table 4.8: Problems identified during the hardware part changes
Problem
Problems and Gaps Identified
No#
Changeover of the hardware parts are performed without any
1
standardized sequence i.e. replacement of parts in wrong order
delaying overall process
Some of the hardware components today have similar physical
2
attribute and function, thus might be fungible and does not need
to replace
Changeover involve many small, delicate and hard to be handled
3
4
parts
Changeover involve many screws, bolts and/or fasteners i.e the X
and Y pitch blocks requires up to 7 turn to tighten or loosen
The overall bottleneck in the hardware part changeover is during
5
the replacement of the contactor chuck which involve many NVA
activities like greasing, alignment and calibrations
6
The generic hardware parts design are not fully optimized and
need only minor modification to enable as a fungible part
99
4.9.4
TIU replacement
Once all the hardware parts have been successfully replaced, the technician
will move to the next phase called the TIU replacement. Figure 4.46 shows a generic
flow of the TIU replacement that is practiced currently with the time taken to
complete each of the activity.
2.5
minutes
3.0
minutes
4.0
minutes
1.5
minutes
11
minutes
•Walk from equipment area to TIU room
•Request & Collect TIU
•Walk back to equipment area
•System setup
•Undock old TIU activities
•Dock new TIU activities
• Temperature setting
increase from +/-60'c to +/-115'c
•Walk from equipment area to TIU room
•Return old TIU
• Walk back to equipment area
•Total Internal time spent at the TIU replacement phase
Figure 4.46: Time Study of TIU change phase
In the time study flow above, only 3 minutes of the overall 11 minutes is
actually spent on the actual TIU change. A list of problems or gaps observed at this
phase is summarized in the below Table 4.9 together with its forecasted impacts.
100
Table 4.9: Problems identified during the TIU replacement phase
Problem
Problems and Gaps Identified
No#
1
Only 3 minutes of the 11 minutes spent are used for the TIU
replacement while others are NVA activities
2
Almost 4 minutes spent in just walking from equipment
to/from TIU room which is a waste in motion practiced
Internally
3
Performing temperature setting ( increase from current temp
+/- 60’C to +/- 115’C ) INTERNALLY during the TIU
replacement are delaying the overall changeover process
4.9.5
PnP Teaching Process
This is a requisite step that needs to be performed each time hardware
components are replaced. PnP teaching is required to ensure proper contact between
the chuck and the device under test. Figure 4.47 shows a generic flow of this phase
together with the time taken to complete each of the steps. Table 4.10 summarized
the list of problems and gaps identified at this phase.
101
2.5
minutes
1.35
minutes
1.25
minutes
2.2
minutes
0.5
minutes
1.15
minute
8.95 minutes
• Collect 1 full tray of Standard units
from EIMS room
• Pre Physical Setups and validation
• Pre system setup and configuration
• Include temperature stabilization
• Actual Teaching Process
• Post system setup & configuration
• Post physical setup and validation
• Total Internal time spent for PnP
Teaching Phase
Figure 4.47: Time study for PnP Teaching Phase
102
Table 4.10: Problems Identified during the PnP Teaching Phase
Problem
Problems and Gaps Identified
No#
1
The collecting and returning Standard Units performed as
INTERNAL time delays the overall process
2
The temperature turned ON during the TIU replacement step
does not serve purpose or required in this PnP teaching phase
4.9.6
Dry Cycling Validation
This is the 1st phase of validation involving the mechanical units or units
without die. Dry cycling is required to ensure the mechanical parts or movements are
functioning as per requirement and are reliable during real production use. Figure
4.48 shows a generic flow of the Dry Cycling and the time taken to complete each of
its steps. Table 4.11 summarized the list of problems and gaps identified at this
phase.
103
2.5
• Walk to tool shop to collect the mechanical
units
minutes
2.75
• Pre dry cycling visual validation
minutes
0.35
• Pre dry cycling physical setups
minutes
1.5
• Pre dry cycling system setups +
Temperature stabilization
minutes
7.0
• Actual dry cycling process
minutes
0.25
• Post dry cycling physical setups
minutes
1.0
• Post dry cycling system setup
minute
2.75
• Post dry cycling visual validation
minutes
2.5
• Walk to tool shop to return the units
minutes
20 minutes
• Total Internal time spent for Dry Cycling
Phase
Figure 4.48 Time study of dry cycling phase
104
Table 4.11: Problems Identified during the Dry Cycling Phase
Problem
Problems or Gaps Identified
No#
Waste in motion/transportation observed when the technician
walks to/from the tool shop to collect the mechanical units.
1
This is practiced INTERNALLY currently and delaying the
start of this phase
2
The visual verification using bare eyes are actually causing
delay because technician need to perform verification a few
times for unclear reject signatures
Potential delay in starting the Dry Cycling process due to
mechanical units is collected from tool shop fails the pre3
validation inspection i.e. existence of marks, stains etc.
Technician need to walk back to tool shop to returns these units
and collect replacement unit which are free from defects
4
5
The temperature turned ON during the TIU replacement step
does not serve purpose or required in this dry cycling phase
The dry cycling process today uses 5 full trays of units and
consumes almost 7 minutes to complete. This is a NVA activity
but needed only due to validation reasons
105
4.9.7
TP Download Phase
The next phase of the changeover is the Test Program (TP) download step.
Again the generic steps involve in this phase together with the time taken to
complete each step is as shown in Figure 4.49. Table 4.12 summarized the list of
problems and gaps identified at this phase.
2.85
minutes
7.0
minutes
4.15
minutes
14.0
minutes
• Pre TP download setup
• Includes choosing recipe, CTSC reboot etc
• Actual TP download duration from server
to equipment CTSC
• Post TP download Setup
• Include system initialization, TIU check etc
• Total INTERNAL time spent in the TP
download phase
Figure 4.49: Time Study of TP Download Phase
Table 4.12: Problems Identified during the TP Download Phase
Problem
Problems and Gaps Identified
No#
1
During the Actual TP download duration and TIU
initialization, the technician was just idling
2
TP download error due to TIU initialization fail could prolong
the process
106
4.9.8
Standard Unit Validation phase
This is the final validation run using samples of good production units. This
phase is important to ensure the whole changeover of both software and hardware is
fully validated before the equipment is released back for the production use. Figure
4.50 shows the generic flow in this phase together with the time taken to complete
each of the steps. Table 4.13 summarizes the list of problems identified in this phase.
3.5
minute
3.95
minutes
3.2
minutes
0.95
minutes
1.0
minutes
2.5
minutes
15.0
minutes
•Pre-Validation Phsysical setup
•Include quick validation and tray load in
•Pre-Validation System setup
•Include Machine Initialization & Temp Stabilization
•Actual Validation Processing
•Post-Validation System setup
•Post-Validation Physical setup
•Walk back to EIMS to return the units
•Total Internal time spent in Std Unit run
Figure 4.50: Time study of standard unit validation
107
Table 4.13: Problems identified during the standard unit validation
Problem
No#
Problems and Gaps Identified
1
The Std unit collected from EIMS room has physical defects i.e.
solder ball defect, stains which cause need for technician to
walk back to EIMS room to seek for replacement units which
are defect-free
2
Because the temperature was turned ON earlier, it took almost
1.5 minute to re-stabilize the temperature back to 115’C which
is needed before the Standard Unit validation can proceed
3
The overall process of Standard Unit validation can be
considered as a NVA step but necessary for validation. It takes
almost 15 minutes to complete a 1-tray STD unit validation
4.9.9
Wrap Up Activities
This is the final phase before the equipment is fully handover back to the
production. Some of the generic activities involve this is phase is shown in Figure
4.51. The problems and gaps identified in this phase are listed in Table 4.14.
108
3.5
•Housekeeping activities
•Include clean up of the work desk, placing tools and
hardware into respective box
•Change sticky pad from 'Down' to 'Up'
minutes
2.0
•Communication activites
•Includes technician looks for supervisor/operator to
inform on completion
minutes
0.5 minutes
•Technician change AEPT state from 'SDT-Conversion' to
'PROD'
•Technician log off ID
6.0 minutes
•Total Internal time spent during the Wrap Up activities
Figure 4.51: Time study of wrap up phase
Table 4.14: Problems identified during the wrap up activity phase
Problem
No#
Problems or Gaps Identified
1
The changeover duration is prolonged 4.5 minutes because the
technician often changes AEPT state after completing
housekeeping and the communication process
2
All of the activities performed as NVA and does not need to be
practiced as Internal activity, need to find way to move this
activities towards the end of the changeover
109
4.10
Summary
In summary, this chapter illustrated 2 points which is the case study and also
the problem identification. The data gathered at this chapter will be use for counter
proposals and outcome analysis in the next discussions.
CHAPTER 5
PROPOSED FRAMEWORK
5.1
Overview
This chapter is important as it directly relates to the objective of this project
on reducing the changeover process duration by integrating the SMED and TRIZ
techniques. The following discussions will focus directly on counter measure
proposals for the problems and gaps that were identified previously.
5.2
Strategy and Execution for Counter Measures
The root cause of the problems identified earlier can be summarized into 3
main factors namely;
i.
Human (technician)
ii.
Equipment (handler)
iii.
Process Flow
The changeover duration today depends on the interaction of these factors as shown
in Figure 5.1. In order to systematically counter the problems, an integrated proposal
framework consisting both SMED and TRIZ technique will be introduced.
111
As mentioned earlier, both techniques are well renowned with individual strengths
and integration of both will compliment the shortcomings to achieve a better
optimized process.
HUMAN
EQUIPMENT
PROCESS
Figure 5.1: The changeover duration factors
Figure 5.2: The generic proposal model
Figure 5.2 shows a generic proposal model which will be utilized throughout this
chapter. Each identified problem will go through the 3 ‘conceptual’ stages as shown
driven by the SMED and TRIZ techniques. A problem will be proposed with 1 or
more techniques with to achieve one of the conceptual stages. The techniques
proposed can be a generic technique from Shingo’s SMED technique or Altshuller’s
TRIZ. Nevertheless, some of the technique proposed will be a combination of both in
order to achieve a more enhanced solution. Figure 5.3 shows a flow of the technique
extraction from both methodologies.
112
Figure 5.3: Extraction of the proposal techniques
In order to systematically execute the proposals, the identified problems will
be truncated and categorized into 3 main groups which are;
i.
Process Flow Optimization – will focus on process gaps which involves
identifying and separating elements, eliminating Non Value Added
elements, improving parallel activities and streaming some of the tasks
ii.
Hardware Part Setup Optimization – identifying hardware parts
significant to the changeover and improving hardware designs
iii.
Improving the human dynamic and procurements – are generic elements
related to standardization in human practice and documentation
113
5.3
Process Flow Optimization
This is the first and foremost counter measure proposal step where the current
generic changeover process flow as shown in Figure 5.4 is considered as nonoptimized due to:
i.
The ‘successive serial’ flow of activity as shown in Figure 5.5
ii.
The practice of Non Value Added (NVA) or waste activities
iii.
The non standardize and /and complex practices
All the above 3 elements are induces the process to delay further and also the
possibility of both safety and quality issues occurrences.
Figure 5.4: The generic ‘non optimized’ changeover flow
114
Figure 5.5: The ‘serial’ flow of activities
In order to convert a non optimized process to an optimized and efficient
process, the functionality of each phase of changeover will need to be reassessed
especially by:
i.
Identifying individual/group task type
ii.
Segregating task by value to changeover
iii.
Minimize, substitute (streamline) or eliminate task
iv.
Introduce parallel activity execution
Each task or activity in the changeover stage can be divided into either
External or Internal elements. External activities are those executed without
impacting the changeover duration. These types of activities are normally performed
offline or during the process of other activities. Internal activities on the other hand
are those activities that can be executed or performed with machine/process
interrupted and impacting the changeover duration. Internal activities are the major
driver of pro-longed changeover duration.
Once each of task type has been identified, the next step is to define the
functionality of these activities. Each of the External and Internal task can be
categorized into 3 main groups; the Value Added (VA) activities, the Non Value
Added but required activities and lastly the Non Value Added waste activities. It is
115
important to segregate definitely the Value Added activities which has significant
role the changeover as compared to the Non Value Added activities. It is important to
note that although most NVA activities are insignificant and deemed as waste to the
process but a handful of them are directly part of the requirements, routine or
procedures which makes it tougher to be eliminated.
Next step is considered important as it will vet the above mentioned
categorized activities to achieve a more lean process with lesser serial sequences. In
this phase, the activities will be considered to go through either one of more of the
following vetting process;
i.
Moving the easy, simple and low risk Internal task to External Task
ii.
Eliminate the NVA waste activities
iii.
Substitute or use engineering solutions to streamline the NVA but
required activities
iv.
For Internal activity which are value added, find ways to minimize or
stagger the task involve
The last step is actually the end result of the above 3 steps that is the
introduction of parallel activity execution. Parallel activities enhance the opportunity
to pull-in and perform multiple activities at a single phase. Enabling parallel activity
execution will help to truncate the overall changeover and help to restructure the
activity sequencing.
116
5.3.1
Upfront Setup Improvement Proposals
The first and ‘obvious’ improvement proposal is to introduce a better and
standardized upfront setups activities before the actual changeover commences. A
proper early preparedness guarantees in the changeover activities reduction and
minimizing the overall NVA activities occurrences.
In SMED, Shingo’s introduces a Stage 2 technique called the ‘Preparing
Operating Conditions in Advance’ where the technique requires certain tasks which
are directly related to the changeover to be performed and completed earlier than the
current practice. Similarly, in TRIZ this technique is part of Principle No. 10 called
‘Preliminary Actions’ but it expanded the scope by suggesting not only performing
an action before it is needed, but it can be perform either fully or partially depending
on the situation. Also this TRIZ technique also suggest that to optimize this
technique, ensure the objects involve in this action are pre-arranged that they can
come into action from the most convenient place without losing time for delivery.
The goal of this proposal is to identify all activities that are directly related to
the changeover and can be prepared or performed upfront. The advantage of this
SMED’s technique is that it will help to convert many of the current Internal
elements to External element. From direct observation analysis, the 5 pre requisite
kits for this handler changeover are:
i.
Hardware Change kits
ii.
1 set of toolbox
iii.
TIU
iv.
5 trays of Mechanical Units
v.
1 tray of Standard Units
117
Table 5.1 shows the current practice of acquiring the 5 pre requisite items during the
respective stage therefore increasing the Internal time. The opportunity here is to
follow as per SMED and TRIZ technique which is to move this to be part of a
standardized upfront setup activity which is compulsory to be done before a
changeover commences. To minimize further the NVA waste activities like multiple
walking or repeated request submission, per TRIZ’s ‘Beforehand Cushioning’
technique proposal is that technician acquire or prepare more than required part
before the start of the changeover. For instance is that technician can collect and
prepare more than the required 5 trays of mechanical units as if there is any defect
found in the units during the pre validation visual inspection, the technician can
quickly replace with the buffer units rather than the need to go back to the tool shop
and request for a new set of units. Table 5.2 shows a list of ‘Beforehand Cushioning’
proposal that can be performed during the early setup.
Table 5.1: The 5 pre requisite items and current practice
Phase Items
Task
Acquired
Element
Hardware Setup
Soft Setup
Internal
Tool sets
Hardware Setup
Soft Setup
Internal
3
TIU
TIU replacement
TIU replacement
Internal
4
Mechanical Units
Dry Cycling
Dry Cycling
Internal
5
Standard Units
PnP Teaching
Internal
No.
Pre Requisite Item
Phase Items required
1
Change kits
2
PnP Teaching and
Standard Unit Validation
118
Table 5.2: The ‘Beforehand Cushioning’ proposal
Since all the 5 items are located at different areas in the shop floor (except
change kits and tool sets are placed together in the tool shop), it will be a challenge
for technician to move these items to the equipment area. In order to eliminate
multiple walking or waste in motion, the next proposal is to use a trolley or cart that
will be able to store and move all items together. This not only minimizes the
walking time but also reduce the possibility of mishaps like dropping of units or
damaging TIU etc. The use of trolley is inspired by the SMED’s ‘Improving
Part(Die) Transportation’ where Shingo reckon this as part of SMED’s stage 2
technique in improving Internal element to External element. To improvise further,
as per TRIZ’s ‘Prior Action’ technique suggestion, arranging the 5 pre requisite
items in order of usage i.e. change kit box and toolbox on top shelf followed by TIU
and last the validation units will help to improve the efficiency.
As earlier mentioned, the upfront setup phase can consist of all activities that
are related to the changeover and can be performed earlier without jeopardizing the
activity flow. Identifying the potential upfront setup activities can be done using the
activity dependency mapping as shown in Table 5.3. From the table below, only Soft
Setup and TIU replacement stage are potential stand alone activities which can be
performed as part of the upfront setup activities. With this assumption, the first
streamlining of activity will be combining both this phases and performing them
together as ‘External’ activities before the changeover process officially commences.
119
Table 5.3: The activity dependency table
Soft
Hardware
TIU
PnP
Dry
TP
Standard
Wrap
Setup
Setup
replacement
Teaching
Phase
download
Unit run
Up
Activity
Soft Setup
x
Hardware Setup
TIU replacement
PnP Teaching
X
x
Dry Phase
X
x
x
x
TP download
Standard Unit run
Wrap Up Activity
x
X
x
x
x
x
X
x
x
x
x
x
The biggest available opportunity to perform the Soft Setups and TIU change
is during the Pre Changeover phase which currently involves all communication and
coordination activities performed as External element. As highlighted earlier, the
technician practically idles from the moment the supervisor confirm the changeover
till the completion of the end lot process by operator before the equipment is
handover to the technician. This interim idling period will serve as the execution
opportunity for the new proposal of performing Soft Setups and TIU change as
External task. Before this can be executed, the current Soft Setup and TIU change
activities need to be re-evaluated again to ensure only the Value Added activities are
practiced per the new upfront setup proposal.
The 3 significant activities from current Soft Setup phase that need evaluation
are:
i.
The practice of placing ‘DOWN sticky’ tags before the changeover starts.
This though considered as a NVA activity but is required as part of visual
factory indicator. Nevertheless, this can be practiced as External element
rather than Internal if performed during the Pre Changeover phase
120
ii.
Placing barricades around the equipment area is not a common practice
but still performed by some as measure of safety. Propose to eliminate
this activity with justification that changeover are low safety risk with no
major hardware overhaul
iii.
Technician performing equipment temperature setup (i.e. cool down
temperature to ~60’C) before hardware setup can commence; is also
another opportunity to convert from current Internal to External. Optimize
the operator’s end lot period to parallel perform this.
Using TRIZ’s ‘Segmentation’ technique, the TIU change phase can be
fragmented as shown in Figure 5.6. The technique proposed is useful for separating
Internal and External activities where in this case, the segmentation will also help to
identify the NVA activities inside the TIU change phase. The current activities can
be segmented into 4 sections are per the color coding below:
Figure 5.6: Illustration of ‘segmented’ TIU phase
i.
Segment 1: Collecting TIU as an internal activity can be rescheduled and
be made part of the compulsory upfront setup activities as earlier
mentioned. Thus, this can be eliminated.
ii.
Segment 2: Consist of 4 activities which require 3 minutes to complete.
These are the only VA activity of this phase and thus cannot be
eliminated. Nevertheless, all these 4 activities are Independent activities
where they can be performed without the concern of its predecessors
outcome. In other words, these activities can be eliminated from this
phase and performed together in parallel during the upfront setup.
iii.
Segment 3: Technician normally performs equipment temperature setting
(increase from room temperature to hot temperature of 115’C) post
121
hardware replacement. From observation, the temperature setup as early
after TIU replacement delays all the subsequent activities (i.e.
Temperature Stabilization) as the correct temperature setting of 115’C is
only needed during the Standard Unit validation before the equipment is
handover back to Production. In other words, performing temperature
setup is a NVA activity in this phase and the subsequent phases thus
proposal is to move this activity to be part of the Standard Unit
validation’s system setup activity.
iv.
Segment 4: Returning the old TIU back to TIU room for store keeping is
a NVA activity and does not need to perform during the changeover time.
This is logical and thus proposal is to keep the old TIU and all items
returning practice can be accumulated and performed towards the end of
changeover process as External time.
With the above proposals, it is clear that the current TIU change phase not only could
be practiced as a standalone activity during the upfront setup phase but also truncate
the whole activities as the NVA steps can be eliminated.
Other minor improvement proposals which will be part of the TRIZ’s ‘Prior
Action’ technique are the upfront setup of empty JEDEC trays inside the handler.
The pre physical setup during validation phases are delayed mostly due to the fact of
needing technician to insert new trays into the input and buffer stage of the handler
before the actual process can commence. With new proposal, technician will load in
empty trays into the buffer stocker as part of the upfront setup activity.
Figure 5.7 shows a generic summary of the changes in the changeover
process with the above counter measure proposals. The most significant changes will
be reduction of changeover phases and duration with more activities performed in
parallel during the upfront setup phase.
122
Figure 5.7: The generic changeover process flow with upfront setup proposals
5.3.2
TP Download Improvement Proposals
The TP download phase is a dependent Internal activity that can only be
performed once the new TIU has been properly replaced and setup. It is a
requirement today that technician will need to ensure the correct Test Program has
been downloaded into the test modules before returning the equipment back to the
Production ( in order to minimize Production Downtime). Due to this requirement,
improving and optimizing this phase is equally important.
Using TRIZ’s ‘Segmentation’ technique again, the major activities that make
up the TP Download phase can be shown as in Figure 5.8. The 5 major activities can
be divided into 2 segments
i.
Dependent Tasks
ii.
Independent Tasks
123
Figure 5.8: Segmented TP download phase
The dependent task of the TP download phase can be described as activities
that require manual interaction from the technician especially on instructions and
parameter input to the system in selecting and downloading a current TP. During this
period, the equipment or the handler will be idling with no other activities allowed to
take place as machine is awaiting or analyzing the instruction. Thus, dependent
activities will remain as ‘Internal’ elements with no significant changes.
The independent task of the TP download phase is utmost unique compared
to other activities as per the observation. The 2 activities listed as Independent tasks
are actual activities that run on the ‘background’ or known as passive activities. As
observed, once the correct parameter and setting has been done by the Technician,
the system will auto-start the TP download from the centralized server into the
equipment database and subsequently continue with the system initialization process
with the new TIU docked. From the data analyzed, the 11 minutes of Independent
activities execution is an opportunity to perform other setup in parallel. From the
current changeover process, the only activity that can be performed in parallel during
the TP Download phase is the Hardware Part Setup activities which is mutually
exclusive and does not interrupt the download or the initialization process. In SMED,
Shingo suggested the ‘Parallel Execution’ technique as the basis to streamline
activities to achieve process optimization.
124
For better enhancement, the actual TP download duration of current 7
minutes can be further reduced if the current 100MB public network cables are
substituted with the new broader and efficient 1GB network cables. Besides to
expedite TP reset and SC switch, if the current Station Controller’s Memory (RAM)
and processor (CPU) are upgraded to a better and higher speed units. This system
and network upgrade will require high cost but will able to salvage 50% of the
current duration.
Though the new proposals above did not directly minimize the Internal time
of the TP Download stage but the proposal of executing Hardware Part Setup will
help to reduce the overall Internal time. Figure 5.9 shows an illustration of the
changeover process with TP download now reschedule to be the first Internal time
activity of the changeover with Hardware Part Setups performed in parallel when the
actual download process takes place.
Figure 5.9: The changeover flow with TP Download improvement proposal
125
5.3.3
Validation/Calibration Improvement Proposals
As highlighted during the problem identification phase, the validation
practices in the current changeover process is a major setback in achieving a quicker
and leaner conversion. Almost 18% or 43 minutes is spent on performing the PnP
Teaching, mechanical and standard unit validation.
The most significant similarity between these 3 stages are that there are inter
dependent on each other where PnP Teaching is dependent of the Hardware Setup
and all previous setup while Mechanical unit validation is dependent on PnP
Teaching and all previous setups and lastly Standard Unit Validation is all fully
dependent stage. This ultimately shows that all these 3 stages can only be executed in
a typical serial show as currently practiced. This condition limits much of the
improvement proposals but nevertheless, the real optimization opportunity is actually
the activities embedded inside this phases.
Both Mechanical unit and Standard Unit validation have very similar steps
especially on the requirement of performing Pre Validation Visual Inspection. This
step is required to ensure only good and reject free units are tested during the process
to eliminate ‘noise’ factor during actual validation. The practice of performing Pre
Validation VI as Internal element currently can be opt to be performed upfront
especially during the period technician idle or when machine is busy processing.
From observation and data collected, an average technician idle time during each
phase is summarized in Table 5.4. To reduce the Internal time spent for the Pre
Validation VI, proposal is for Technician to perform the Pre VI for all 5 trays of
mechanical units during the PnP Teaching phase while capitalized the idling time
during Mechanical Unit validation to perform the VI for the 1 tray of Standard Units.
This can be linked back to both Preparing Operation Condition in Advance and Prior
Actions techniques from SMED and TRIZ respectively. To help improve the Visual
Inspection’s efficiency and reduce the duration, the proposal is to introduce
magnifier scope at work desk for technician to perform the Visual Inspection. The
magnifier scope will replace the current bare eyes inspection which is both time
126
consuming and error prone. Introducing a scope is part of the SMED’s
‘Mechanization’ technique where Shingo suggest improvising manual work with the
help of better efficient tools.
Table 5.4: Technician idle time
The validation phases are also ‘plagued’ by NVA or waste activities like the
act of returning the units back to respective room once the validation complete and
this is performed as Internal time. As suggested earlier, all practices of retuning kits
and items will be standardized as a post setup activities which will be performed only
when the changeover complete as External elements. In TRIZ, this is called the
‘Taking Out’ technique where elements that are not needed are vetted out from the
process.
The validation of using 5 full trays or 135 mechanical units contributes to the
long validation period with each tray requires 1.5 minutes to complete. Standard unit
validation using 1 tray takes around 3.2 minute to complete. As validations are NVA,
the proposal will be to streamline both activities by either:
i.
Reduce the Mechanical Unit samples but remain Standard Units
samples
ii.
Eliminate Mechanical Unit validation and combine together with the
Standard Unit validation with higher sample quantity
127
iii.
Eliminate Mechanical Unit validation and combine together with the
Standard Unit validation with same sample quantity
iv.
5.3.4
Eliminate both validation phases
Post Setup Activities Alignment Proposals
As highlighted during the early phases, those activities categorized as NVA
and waste which cannot be directly eliminated or embed part of other phase will be
proposed to be aligned and performed towards the end of the changeover process.
Using TRIZ ‘Merging’ technique, all the NVA activities identified throughout the
changeover process is listed in Table 5.5 below.
Table 5.5: ‘Merging’ technique on identifying NVA activities
No.
Post Setup Activity
Phase Activity practiced
Task
Element
1
Returning TIU back to the TIU room
TIU replacement
Internal
2
Returning Mechanical Units to the toolshop
Dry Cycling
Internal
3
Returning Standard Units
Standard Unit validation
Internal
4
Return changekit and toolsets
Wrap Up Activity
Internal
5
Housekeeping and changing 'sticky' tag
Wrap Up Activity
Internal
6
Changeover completion communication
Wrap Up Activity
Internal
From earlier discussion, the reason behind the pro-long duration of the Wrap
Up activity phase is because all the NVA activities are performed as Internal element
with AEPT state change performed towards the end of the flow. To ensure all these
128
NVA activities are captured as External element, the proposal is to first complete the
changeover and log off the AEPT system before performing all the activities listed
above. In other words, applying the TRIZ’s ‘Do it the other way’ technique, the
current flow of the Wrap Up activity phase is inverted or mirrored with AEPT
system setup performed first followed by the other activities. Below is the proposed
sequence for the new post setup phase:
i.
AEPT change state from ‘SDT-Conversion’ to ‘PROD’ ( log as official
completion of changeover process)
ii.
Housekeeping activities i.e. cleaning of the work desk, area around the
equipment
iii.
Inform supervisor/operator on the changeover completion and equipment
readiness for process
iv.
Change the ‘sticky’ tag from ‘DOWN’ to ‘UP’
v.
Re-arrange all hardware parts into the change kit accordingly
vi.
Re-arrange all tools back into the toolbox
vii.
Place the change kit box and toolbox on the trolley
viii.
Ensure the TIU and the units ( standard and mechanical) are also on the
trolley
ix.
Push trolley to respective room to returns the parts accordingly
The new proposed flow is illustrated as shown in Figure 5.10 while the new
changeover process with the proposed elimination of Wrap Up phase and the new
NVA activities aligned towards the end of the changeover as External elements as
shown in Figure 5.11.
129
Current Wrap Up Phase
Proposed ‘Invert’ flow of Post Setup
Change AEPT state from
'SDT-Conversion to
'PROD'
Perform basic
housekeeping
Perform basic
housekeeping
Return change kits and
toolsets
Inform supervisor/operator
of completion
Inform
supervisor/operator of
completion
Change 'sticky' tag to UP
Change 'sticky' tags to 'UP'
Rearrange changekit and
tools into respective
boxes and on trolley
Change AEPT state from
'SDT-Conversion to 'PROD'
Rearrange all items
including TIU and units
on trolley
Figure 5.10: The new aligned Post Setup phase
130
Figure 5.11: The changeover process flow with wrap up phase elimination proposal
5.4
Optimizing Hardware Setups phase
The next important proposal will fully focus on the Hardware Setup phase.
This phase as mentioned earlier has the longest duration to complete and causes the
bottleneck to the changeover process. Unlike other phases, the Hardware Part Setup
activities are most directly related and significant to this changeover.
Currently there are 12 hardware parts replaced during the changeover and it
takes almost 160 minutes to complete a whole set of hardware setup. Table 5.6
illustrates the current hardware setup flow and the time taken to complete each of the
part. The current non standardized and non optimized hardware setup delays the
overall changeover process.
Based on the problems identified earlier, the current hardware parts can be
divided into 2 different categories as below;
131
i.
Generic ‘fungible’ parts
ii.
Generic ‘non fungible’ parts
In the following discussions, the counter measures proposed will target mainly on the
items below;
i.
Identify the ‘fungible’ and non ‘fungible’ parts
ii.
Identify the non ‘fungible’ parts that are multi function
iii.
Identify the non ‘fungible’ parts that can be redesign
Table 5.6: The current hardware setup phase
132
5.4.1
Proposal to Identify Fungible Parts
In this part of the proposal, the plan is to identify the ‘fungible’ or generic
component that is common for both the Nebula and Nexus configuration. By using
SMED’s ‘Function Check’ and TRIZ’s ‘Universality’ technique, the current 12
hardware parts from both configuration can be compared to analyze the difference in
physical and function. Generic ‘fungible’ hardware parts from both configurations
will have the exact same physical appearance and form function. Then by using
TRIZ’s ‘Taking Out’ technique, the identified fungible parts can be segregated from
the other non fungible hardware. This will be quickest method to optimize the
number of hardware parts needed to be setup.
5.4.2
Proposal to Identify Non Fungible Parts
A non fungible hardware part from both configurations will have similar
function but significant physical difference thus a challenge in optimizing the
hardware setup further. For the non fungible hardware parts, the proposals will focus
2 main goals;
i.
Identify the non fungible hardware which is multifunction
ii.
Identify opportunity to redesign hardware parts
5.4.2.1 Non Fungible Parts With Multi Function
Using SMED of ‘Function Standardization’ technique with TRIZ’s ‘Local
Quality’ technique, the proposal is to identify the non fungible hardware part with
different physical appearance but has the capability to perform dual function for both
133
configuration. Normally the hardware parts that fall under this category are those
involve in low risk mechanical movements with no significant impact to quality and
safety.
5.4.2.2 Non Fungible Parts with Hardware Redesign
A hardware part will be considered for redesign if the current setup involves
many complex steps or consist of NVA activities. In the problems highlighted the 2
hardware parts which above concern is the Contactor Chuck and the X-pitch blocks.
The contactor chucks are discussed earlier is the main bottleneck of the
hardware part setup involving 21 steps which included many NVA activities like
cleaning, greasing, alignment and calibration which consume 120 minute for 4
chucks setup. The current contactor chuck setup is shown in Figure 5.12 below
where the only difference between both configuration’s chuck is the ‘nest’ used
which is unique to Nebula and Nexus width size. The function of the ‘nest’ is to
clamp the DUT when testing is being performed. The current ‘nest’ design is as
shown on Figure 5.13. Thus, using the TRIZ’s ‘Dynamization’ technique, the
proposal will be to consider a redesign on the current rigid ‘nest’ to be more adaptive
as inspired by a screw wrench or a flexible bed. The new design will allow the ‘nest’
to be flexible and adjustable based on the different width size. The intention of this
whole design is to allow technician to perform a rapid setup without the need to
perform all the complex activities with reduce duration. Also to consider is that the
new design will fit into the current production environment with no safety and
quality concern.
134
Figure 5.12: Current contactor chuck setup
Figure 5.13: Current ‘nest’ design
The X-pitch blocks is one of the 5 hardware parts which are fastener -based
Today it takes almost 4 minute to complete a setup with each block requires up to 7
turns to either tighten or loosen. The current snapshot of the X-pitch block is shown
in Figure 5.14 with it’s current screw design shown in Figure 5.15. The X-pitch
block is located deep inside the handler, thus to remove and replace this part under a
confine area is challenging and time consuming. The proposal is per SMED’s
technique of ‘One Turn Screw’ method where new design will focus on established a
more efficient and simple screw with 1 turn design.
This together with other techniques will be explored to perform both simple
and complex design changes to the current hardware setup if there is a need for
improvement and with boundary conditions like safety, quality and cost.
135
Figure 5.14: X-pitch block
5.5
Figure 5.15: The current screw design
Human Dynamic and Procurement Improvement Proposal
The final proposal focus will dive into the improvements on the human
dynamics and the process documentation. This though considered as basic
requirements but are often neglected or overlooked. Though not part of the initial
scope, this will be covered to ensure a full blown improvement implementation takes
place to ensure sustainability.
Table 5.7 shows the current technician ratio and status per shift. The table
clearly shows though the technician ratio per shift meets goal but the competency is
questionable. To ensure align and standardization, all technician both certified and
non certified will need to go through the refresher training and those non certified
will need to assess immediately. The training will include both theoretical and
practical as technician will need to be familiar with both parts. A qualified senior
staff or engineer will need evaluate each and every of the technician to ensure
standardization in assessment and only the qualified are allowed to perform the
136
changeover. SMED puts high focus on human development and thus a
comprehensive training and assessment will be the core for a better changeover
sustainability.
Table 5.7: Technician headcount per shift
Shift
A
B
C
D
Technician
Non
In
Actual headcount Certified
headcount goal
Certified progress
6
6
2
3
1
6
6
3
1
2
6
7
4
2
1
6
8
4
3
1
All the above mentioned proposals once validated will need to be
documented in either control specification or as training package. The standardized
and organized document will ensure proper flow of information without any
discrepancies. A checklist will be useful for technician to quickly refer during a
changeover and ensure no steps are skipped or missed. Also, better business
processes need to be enforce during the changeover request from supervisor to
technician. An efficient business process will eliminate the ad hoc and unstructured
request which is prone to miscommunications.
5.6
Summary
This chapter has highlighted all the counter measures per SMED and TRIZ
techniques in response to the problem identified earlier. The proposal focused on
process improvement, hardware setup optimization and lastly on the human
dynamics and procurements. The outcome of these proposals will be analyzed in the
next chapter.
CHAPTER 6
RESULTS AND DISCUSSIONS
6.1
Overview
In this chapter, the outcome of the counter measures proposed earlier will be
thoroughly analyzed and discussed. The results or outcome are purely based on
empirical data analysis and engineering decision like Design of Experiments (DOEs),
real time experience, structured brain storming session and cycle runs among others
but not limited to. All the below information below are based on the 13 weeks
(Dec’2010 to Feb’2011) data collection. This chapter will also include the Return of
investment (ROI) analysis based on the implemented outcome.
6.2
Implementation of Pre Setup phase
The whole initial intention of introducing this new Pre Setup phase was to
enable the elimination of the Preliminary Soft Setup phase and to enhance further the
Pre Changeover activities. In other words, the new Pre Setup phase streamlines
activities from both phases by improving the upfront pre-setup activities which
automatically converts the ‘Internal’ tasks to become ‘External’ tasks. Besides that,
the introduction of this phase also helps to move the whole TIU replacement phase
138
from an ‘Internal’ stand alone phase to an ‘External’ activity performed in parallel
with other Pre Setup activities.
Table 6.1 shows the list of current, proposed and the actual implementation
status for this Pre Setup phase. All the proposal were accepted and were able to be
implemented successfully expect 1 proposal of acquiring 2 TIU board instead one the
typical 1 board as a precautionary measure incase the TIU fail during the
initialization session (during TP download). This proposal was rejected because as
TIU are controlled and high value collateral items, the engineering team suggest to
limit 1 TIU per transaction to avoid any misuse or damage to the part.
Table 6.1: Pre setup improvement result
No
1
Current
Proposal
Implementation Status
All the changeover
All these items will be
Accepted- Is the parallel
required items i.e.
part of the upfront setup
activity where while waiting
Change kits, toolbox,
activity requirement.
the last lot to complete, the
TIU, Mechanical
Upon identifying the
technician will go to get the
Units and Standard
equipment for
necessary items. This
Units are collected or
changeover, the
reduce the technician idle
prepared during the
technician will start to
time during the early stage
changeover duration
prepare this items
and help to convert the
as ‘Internal’ tasks
‘Internal’ tasks to ‘External’
tasks
2
Technician needs to
Introduce the practice of Accepted – The already
walk multiple times
using a trolley to place
available 3-tier ergo trolley
between the
and move all the items
in the shop floor can easily
equipment to
together. This improves
place all the 5 major items
respective rooms each
the waste in motion and
and can be moved easily
time to collect the
transportation. Also
required items
help to improve safety
139
of the technician and the
items
3
Some technician
Remove this activity as
Accepted- There is no
practice placing
it is a NVA waste
safety concern by removing
barricades before start
activity
this task, changeover
of changeover and
activity is allowed to be
this is practice as an
performed without the need
‘Internal’ task during
to barricade the work area
the Preliminary Soft
Setup phase
4
Changing ‘sticky’ tags These are NVA but
Accepted- The temperature
and performing
required activities,
setup will now be part of
temperature setup
proposed to change this
the end lot process activity
(warmup/cooldown)
activity ownership to
while changing the ‘sticky’
are part of the
operator before the
tag (from ‘UP’ to ‘DOWN’)
technician’s role
equipment is handover
will be part of pre
during Preliminary
to the technician.
changeover housekeeping
Soft Setup phase
activity done by operator.
which is practiced as
Though this activities
‘Internal’ activities
prolong the operator’s
activity but it reduce the
overall ‘Internal’ time
5
Before performing the
To minimize the pre-
Accepted- When operator is
validation run,
validation physical
busy performing the final
technician will need
setup during each
housekeeping activities,
perform pre-
validation run, propose
technician can perform this
validation physical
to ensure enough empty
clear/load in empty trays
setup that include
trays are pre-loaded into into buffer stocker. This
clear old trays from
the buffer stocker
will help to minimize
reject stocker and load before hand
‘Internal’ time during
trays to buffer stocker
Standard Unit and Dry
140
Cycling phase
6
7
TIU replacement is
TIU replacement
Accepted- During the ‘end
performed as an
activities are
lot’ process once operator
‘Internal’ phase once
independent activities
has clicked ‘End Summary’,
all the hardware parts
thus propose to perform
the technician in parallel
have been replaced
is during the Pre-Setup
will start the TIU
activity
replacement activities
Technician collect 1
To minimize walking
Rejected – TIU boards are
TIU board each time
time and delaying the
High Value and controlled
during a changeover
overall process, suggest
collaterals thus to avoid any
and if any failure with
technician to collect 2
misuse and potential
TIU board need to
TIU boards instead of 1
damage, only 1 TIU
walk back to TIU
during the Pre Setup
transaction is allowed per
room for replacement
preparation
time. Furthermore, placing
2 TIUs on the trolley are not
encouraged
8
The Pre Changeover
Eliminate these phases
Accepted – New phase
and Preliminary Soft
and streamline all the
called the Pre Setup will be
Setup phase are 2 non
activities into 1 single
introduce where all the
standardized activities
phase
activities from the current 2
practiced today
phases will be performed as
‘External’ task and
subsequently eliminate both
phases
Table 6.2 shows the new ‘end lot’ process and its activity sequence that
operator will need to comply before the equipment is handover to the technician for
changeover. The new flow included an extra step where operator will perform the
equipment temperature cool down (or warm up) before starting ‘end summary’
action. The equipment temperature will cool down (warm up) by itself while
operator will continue to perform the units segregation thus the temperature setup is a
parallel activity.
141
Table 6.2: The new ‘End Lot’ process sequence
No.
1
New ‘End lot’ Process
Move out lot from the
Move out lot from the handler to
handler to the table
the table
Counting and visual
2
verification
3
4
Current ‘End Lot’ Process
Counting and visual verification
End lot process in system
End lot process in system
Physical segregation of good
Equipment temperature setup
and bad units
(cool down/ warm up)
5
Good unit shipment process
6
Bad units reject process
Tracking sheet
7
documentation
Physical segregation of good and
bad units
Good unit shipment process
Bad units reject process
Send good units to next
8
station and reject to RIMS
Tracking sheet documentation
room
9
Nil
Send good units to next station
and reject to RIMS room
Figure 6.1 shows the new Pre Setup activity flow which combines all
activities from 3 different phases namely; Pre Changeover, Preliminary Soft Setup
and TIU replacement phase into one single standardized phase. In this new phase,
parallel activity execution occurs at least 3 times especially during:
i.
Operator prepare to complete lot while supervisor ensures no lot’s
queue in the system awaiting to be processed by the identified
equipment and in parallel technician is already preparing the
changeover kits and required items
ii.
When operator is busy performing end lot process, the equipment
temperature is being automatically setup (cool down/warm up)
and the technician will proceed to replace the TIU
142
iii.
When Operator is busy performing housekeeping , the technician
will clear all old trays inside the handler and replace with new
trays in the buffer stocker
1. Supervisor forecast and
decide on the changeover need
3. Supervisor and technician
identify the equipment to be
changeover
5.The operator
ensure the last
lot complete
processing
succesfully
2.Supervisor identifies the
available technician from the
team and passdown on
requirement/expectation
4a.Supervisor ensure no lots are queued
to be processed by the identied
equipment
4b. Technician start to i.e changekits, TIU,
tools, mechanical and std units
6a. The operator perform all neccesary
end lot process include temperature
setup (cool down/ warm up)
6b. Once end summary complete,
technician will proceed with the TIU
replacement phase
7a. The operator performs basic
housekeeping include change
'sticky' tag
7b. Technician will clear the old
trays and replace with new trays
inside the handler
8. The operator log off his/her
ID and handover the
equipment to technician
9. Technician log in the ID
and AEPT change from
'PROD' to 'SDT- Conversion'
Figure 6.1: The new pre setup activity flow
Figure 6.2 shows the development of the Pre Setup phase and the scheduling
of parallel activities. Figure 6.3 shows a more clear breakdown and time study of the
newly implemented Pre Setup phase.
143
Figure 6.2: Illustration of the pre setup phase development
144
•Supervisor finds and identify the technician
•Communicate on changeover request and
expectation
•Supervisor and Technican decide and identify
of the equipment for changeover
•Supervisor communicates with operator to
complete the last lot
• Ensure no new lots or WIP in queue to be
process for the identified equipment
•Technician proceed to prepare the requires
kits, tool, and units for later changeover and
place near equipment area
•Operator performs end lot process including
temperature setup ( external step)
• Technican will proceed with TIU replacement
activity
•Operator performs basic house keeping
activities including changing 'sticky' tag
• Technician will clear old trays and replace
new trays in buffer stocker
•Operator log off ID as user and passover the
equipment to technican
•Total EXTERNAL time spent in the new Pre-Setup
phase
7
minutes
5
minutes
7
minutes
10.5
minutes
3
minutes
32.5
minutes
Figure 6.3: Actual time study of the new pre setup phase
Figure 6.4 illustrates the actual change in the changeover process flow with
the introduction of the new Pre Setup phase and the elimination of the Preliminary
Soft Setup and the TIU replacement phase.
145
Figure 6.4: Illustration of changeover process change with pre setup phase
In summary, the introduction of this Pre Setup phase has been an advantage
to the overall changeover process by not only reducing the duration but also helped
to streamline the initial 3 phases into one single phase. The new phase is lean as all
the activities are performed as ‘External’ task while those NVA waste activities are
eliminated.
146
6.3
Improvement of the TP Download phase
There is no much significant improvement on duration reduction of the TP
Download phase and all activities 5 major activities will remain as current ‘Internal’
element. The major proposal on system and network upgrade was rejected by the
management due to budget and resource constraint but these improvements are on
the long strategic plan for year 2012.
Nevertheless, with the ‘trimming’ of the TIU change phase and moving it to
be part of the Pre Setup phase, the TP Download stage will now move up the
sequence to be first changeover process. The unique characteristic of this TP
download phase is that though all 5 major activities need to be performed as
‘Internally’ but almost ¾ of the activities duration are background process or autorun in the background. This characteristic identified as an opportunity earlier will be
used to perform Hardware Setups which will be illustrated in detail later.
Table 6.3 shows the current practice, proposal and the implementation status
for this phase.
Table 6.3: The TP download improvement result
No
1
Current
Proposal
Today TP download
Implementation Status
TP download phase is
Accepted- Is the most
only dependent on the
logical solution that once TP
activity of the
TIU replacement phase
logs in the ID as user, will
changeover after the
thus proposal is to move start off by performing the
completion of Dry
this phase to be the new
TP download activity.
Cycling phase.
1st activity of the
Technician will use all the
Technician idle
changeover process.
info provided by the
is practiced as the 6
th
147
almost 10 minutes
supervisor before choosing
during this phase
the correct ‘recipe’ and
downloading the TP
2
During the 2nd half of
Since the actual
Accepted – In average the
the TP download
download and TIU
TP download and TIU
phase which is during
initialization are
initialization takes around
the actual TP
background process,
11 minutes to complete, thus
download and TIU
suggestion is for
to minimize idle time, the
initialization step, the
technician to start
technician can start to
technician idles
performing hardware
perform hardware part setup
during this period as
part setup
without interrupting the
most activity as
download/initialization
system setup or
process
machine time
3
The actual TP
Faster and quicker TP
Rejected – This change
download takes in
download is possible if
consume cost and resources
average 7 min to
the current network
beyond the budget. But the
complete from server
cable of 100MB can be
network cable upgrade will
database to the CTSC
switch and replaced
be performed during the
with the 1G
whole factory migration in
Q4’11
Figure 6.5 shows the activity breakdown of the current TP download phase
with the latest completion duration. With the TP download phase moved to be the 1st
changeover activity, the process flow of the changeover changed significantly and is
illustrated in Figure 6.6.
148
Figure 6.5: TP download activity breakdown
Figure 6.6: Illustration of changeover process change with TP download
phase improvement
149
6.4
Improving of Hardware Setup phase
The next important improvement proposal to analyze is the optimization of
the hardware part setup. The success of this improvement implementation will
determine the overall success of reduction the changeover duration. Today, in total
12 hardware parts are replaced and setup during a changeover with duration of 160
minutes.
The outcome of the Hardware Setup improvement can categorized into 3
categories as below:
6.4.1
i.
Identifying the ‘fungible’ hardware parts
ii.
Identifying the ‘non fungible’ hardware parts
iii.
Redesigned or modified hardware parts
Identified ‘fungible’ parts
Based on the experimentation and observation, there are 4 components per
today’s hardware parts which are naturally fungible or common to both
configuration. These hardware parts are similar in both physical appearance and in
functionality thus they are not required to be replaced. The 4 hardware components
identified are;
i.
Loader/Unloader Pickup Head
ii.
Loader/Unloader Y-pitch Conversion Blocks
iii.
Loader/Unloader Buffer/Exit pick up heads
iv.
Stopper for Input/Exit assembly
150
Besides these 4 components, another hardware part that is not naturally
‘fungible’ but is able to perform dual function is the Transfer Pick Up Assembly
component which was identified through the TRIZ’s ‘Local Quality’ technique. This
hardware part is used to pick up the empty trays from buffer stocker to the exit
stocker. An illustration of the Nebula and Nexus Transfer Pick Up Assembly
component is shown in Figure 6.7 below. The most significant physical difference
between both configurations is the number of suction cups placed on this component
where Nebula will have 2 cups and Nexus with 1 cup. The suction cups count
corresponds directly to the empty pocket in the respective configuration trays as
shown in Figure 6.8 below.
Nebula
Nexus
Figure 6.7: The transfer pick up assembly
Figure 6.8: The Nebula and Nexus tray difference
151
The DOE performed with 3000 units for 20 cycles showed that the Transfer Pick Up
Assembly designed for Nexus is able to function for both configuration as the single
suction cup is able to lift trays from both configuration without any major concern
and the risk assessed is low as the hardware functions only to transfer empty trays
without units. With this new implementation, the technician can salvage 13 minutes
and exclude the complex step to setup this part. Table 6.4 summarizes the list of
identified ‘fungible’ parts that can be excluded from the hardware setup list.
Table 6.4: The identified ‘fungible’ hardware parts
No.
Hardware
Name
Hardware Snapshot
Function
Status
LD: To pick
the units from
input tray to
heat plate
1
LD and UD
pickup head
Generically
fungible thus
ULD: To pick
not required
units from exit
to be changed
plate to move
to unloader tray
Qty: 4 pcs/set
To ensure units
2
LD and UD
are pick up in
Generically
Y-pitch
correct position
fungible thus
conversions
in Y-axis
not required
blocks
Qty: 4 pcs/set
to be changed
of 8 screws
3
LD and UD
To pick up and
Generically
152
buffer/exit
transfer unit
fungible thus
pick up
from heat place
not required
head
pocket
to be changed
exchanger to
buffer stage
and from buffer
to exit plate
Qty: 4 pcs/set
To ensure the
assembly
Stopper for
4
Input/exit
assembly
mechanism
Generically
buffer and exit
fungible thus
to be in correct
not required
location and
to be changed
avoid pick up
issues
Qty: 2 pcs/set
Generically
Transfer
5
Pick Up
Assembly
To pick up
non fungible
empty trays
but propose
from buffer
to use the
stoker to
Nexus design
input/output
as standard
stoker
part for both
configuration
Qty: 1 pc/set
and need not
to be replaced
153
6.4.2
Identified ‘non fungible’ parts
All the remaining hardware that does not fit into the above 2 category will be
listed as non ‘fungible’ as these parts though similar in function but significant
physical difference that requires it to be replaced each time. Table 6.5 shows a
summarized non ‘fungible’ list of the remaining hardware parts.
Table 6.5: The identified non ’fungible’ hardware parts
No.
Hardware
Name
LD and ULD
1
X-pitch
conversion
Hardware Snapshot
Function
To ensure units pick
up units in correct
X-axis Qty: 2
pcs/set
To place the DUT
and also to
2
Heat Plate
heat/soak units
which improve the
cycle time
Qty: 1pc/set
To place units from
3
LD and UD
heat plate and
front/rear
transfer them into
buffer stage
chamber ( test head)
Qty: 4 pcs/set
154
To pick up units at
4
Contactor
Chucks
buffer stage loader
and test the units at
TIU for functional
test
Qty: 4 pcs/set
Only used for cold
test to remove
5
Unsoak plate
moisture from the
units before moving
out
Qty: 1 pc/set
To place the tested
units from Buffer
stage
6
7
Exit Plate
Qty: 1 pcs/set
Socket
Plate that place the
Cleaning
SCD surrogate
155
Plate
cleanin device
coupon . SCD
coupon to clean TIU
pogo pin. Every
5000units run, need
to perform cleaning.
Different due to
coupon size
Qty: 1 pc/set
6.4.3
Redesigning Hardware Parts
From the proposals highlighted earlier, the 2 hardware parts that was
considered for redesign or modification is the Contactor Chuck and the X-Pitch
Conversion blocks. The details of the hardware redesign are explained as below;
6.4.3.1 Nest redesign
The current contactor chuck setup involves 21 steps and takes 120 minutes to
complete. As highlighted earlier, the only difference between Nebula and Nexus
contactor chuck is the ‘nest’ which again differs only by the width size.
Thus, out of 21 steps performed, the only value added step in this setup is the
replacement of the ‘nest’ on the chucks but when each time a nest is replaced, other
NVA activities like cleaning, greasing and calibration is needed to be performed as
well which delays the overall process. Besides this, the fixing and replacement of the
chucks and the nest involve many small and delicate parts which is hard to be
handled in addition to many screws and fasteners as well.
156
To counter this, the new proposal was to use engineering solution to design a
new nest which will have the ability to adjust it’s width size depending on the
configuration. Also, the new design nest will not require technician to dismantle the
whole chuck as currently practice which will help reduce overall setup of the time.
Using TRIZ’s ‘Dynamization’ technique, a new nest was designed and is as shown in
Figure 6.9 below where new ‘adjuster clips or clampers’ are introduced to allow the
width size adjustment. The new design model is adjustable and flexible to support
both configuration with technician needs only to move the clips either inwards
(Nexus) or outwards (Nebula) by adjusting the screw position. Table 6.6 shows the
steps involve in changing the ‘nest’ size using a standard hex key. The clips are
fabricated only to move in this 2 pre-defined positions with each screw can only be
loosen/tighten in 3 turns. Figure 6.10 shows a side to side comparison between the
old and new nest while Figure 6.11 shows a comparison of the chucks. The new
‘nest’ design was put under exhaustive and extreme testing for 20 cycles of 3000 unit
to fully validate its reliability and durability. Since the core design and material use
to build the hardware are similar, thus no concern from safety, quality and reliability.
With this new ‘nest’ design, the whole 21 steps involving multiple complex and
NVA activities can be truncated to a generic 6 quick steps as illustrated in Table 6.7.
This new design demonstrated the advantage of using the TRIZ technique in
identifying specific solutions from general problem.
157
Figure 6.9: The new ‘nest’ design
Figure 6.10: Old versus new ‘nest’ design comparison
158
Figure 6.11: New chuck (L) and old chuck (R ) comparison
Table 6.6: The steps in changing the ‘nest’ size
Step
1
Illustration
Description
The Nebula contactor with
default ‘nest’ size setting ( the
clampers are ‘outward)
159
Use the standard hex key to
‘loosen’ the screws
2
3
Once loosen, use fingers to
push the clips inwards
4
Tighten the screws back into
position using the hex keys
The Nexus contactor chuck
5
and ‘nest’ ready to be used
160
Table 6.7: The actual changeover steps with the new ‘nest’ design
Step
1
Illustration
Description
Technician picks the
hex key
161
2
Technician adjust the
contactor chuck’s
position from bottom
opening of the handler
3
Technician loosen the
screws for each of the
individual nest from
bottom opening of the
handler
4
Manually push the
clips to either inwards
or outwards
162
5
Technician tighten the
screw back into
position
6
Technician completes
task and return the
contactor chucks to
position
6.4.3.2 One turn screw design
The next proposal will focus on performing a simple design change on the X
pitch converter part which is a non ‘fungible’ part that need to be replaced. It takes
currently 4 minutes to remove and replace a pair of this part during a changeover.
The long duration is mainly caused by the motion of tightening and loosening of the
screw that need to be performed using both the fingers and the screw driver. In
average, a technician needs to turn 7 complete cycles to either loosen or tighten a
screw for to remove or replace this part. Figure 6.12 shows an example of the current
X-pitch component together with its screw thread. In SMED, Shingo suggest to
minimize the need for adjusts and the use of fasteners altogether.
163
Thus, the new proposal is to redesign the screw-type used for this part using
the Shingo’s model of 1 turn screw. A low cost design of a 1 turn screw was
fabricated for the X pitch conversion as shown in Figure 6.13 where with this new
design it is forecasted the hardware replacement time can be reduced to half at it
takes only 1 turn for each of the screw to tighten or loosen without the need for any
tools i.e. screwdrivers. Table 6.8 shows an illustration of removing and replacing a
X-pitch conversion using this 1-turn screw proposal.
Figure 6.12: The current X-pitch conversion block and screw design
Figure 6.13: The new X-pitch conversion block with the new screw’s thread
164
Table 6.8: The functionality of the one turn screw
Step
Illustration
Description
Current X-pitch
1
conversion in place in
the handler
Technician use 2
2
fingers to loosen the 1
turn screw
The 1 turn screw has
loosen, remove the X
3
pitch conversion and
replace a new one
with the same method
With the above implementations, Table 6.9 summarizes the Hardware Part
setup’s current, proposal and implementation status. Table 6.10 summarizes the
average time study of the new optimized hardware part setup. The new result shows
the hardware part setup now can be completed below 11 minutes compared to the
160 minutes previously. Figure 6.14 illustrates the parallel execution of the
hardware setup when the TP download and TIU initialization is in progress. Figure
165
6.15 shows the actual changeover process flow and duration with the new optimized
hardware part setup changes.
Table 6.9: The result of hardware setup proposals
No
1
Current
Proposal
Technician changes
4 of the hardware are
all the12 hardware
identified to be generic
parts/components
in design thus are
during a changeover
applicable for both
Implementation Status
Accepted –The 4 parts;

Pickup Head

handler configuration
and does not need to be
Loader/Unloader
Loader/Unloader Y
pitch conversions

replaced
Loader/Unloader
Buffer/Exit plate
pick up heads

Stopper for
Input/Exit assembly
Have been validated to be
generic for both Nebula and
Nexus thus does not need to
be replaced
2
Technician changes
The ‘Transfer Pick Up
Accepted – DOE performed
the Transfer Pick Up
Assembly’ part
with 30x cycle run showed
Assembly component
designed for Nexus
the Nexus ‘Transfer Pick Up
as part of the
could actually serve
Assembly’ used in Nebula
hardware replacement
purpose for both
configuration able to
activity
configuration thus
transfer empty trays without
remain this as the
any issue especially
standard part that does
dropping of trays or
not need to be replaced
intermittent pick up issues.
Thus this will now be part of
166
the fungible hardware part
that does not need to be
replaced
3
The longest duration
The new ‘nest’ design
Accepted – DOE performed
activity is the
will allow technician to
with 30x cycle run showed
replacement of the
do quick adjustment
the new ‘nest’ design able to
contactor chucks
without the need to
perform equivalent to the
which also involve
remove the whole
current design without any
many NVA task like
chuck and perform the
safety, quality or reliability
cleaning, wiping,
whole NVA activities
issues. Also, the technicians
greasing and
are more pleased with this
alignment calibration
new design as it saves time
and resources
4
The X-pitch block
Use new design of 1-
Accepted- New 1-turn screw
used today use screw
turn screw to minimize
design able to perform
methodology where
the effort to
similar to current design but
technician need to
loosen/tighten the part
able to reduce the time taken
turn at least 7x to
to loosen/tighten the part
loosen/tighten the part
5
Hardware part setup
Propose to perform
Accepted- With the new
is performed as the
Hardware setup as a
optimized hardware
2nd activity in the
parallel activity during
changeover plan and since
changeover flow once
the TP download phase
this phase is independent of
the preliminary soft
as both activities are
the TP download, then it can
setup and equipment
independent of each
be executed in parallel
temperature setting is
other
without any impact to the
done
changeover
167
Table 6.10: The time study result of new improved hardware setup
No
Pre-
Hardware Part / Snapshot
Post-
Improvement/
Improvement/
Time(Min)
Time(Min)
LD/ULD pick up head
1
Fungible thus not
Replaced
replaced
4.0 minutes/set
LD and UD Y-pitch conversions
blocks
2
Fungible thus not
Replaced
replaced
4.0 minutes/set
LD / ULD X-Pitch Blocks
Replaced
1.0 minutes/set
3
but reduced
Replaced
duration due
4.0 minutes/set
to 1-turn
screw
4
Replaced
Heat Plate
Replaced
168
0.5 minutes/set
0.5 minutes/set
LD and UD buffer/exit pick up head
5
Fungible thus not
Replaced
replaced
8.0 minutes/set
LD and UD front/rear buffer stage
6
Replaced
Replaced
4.0 minutes/set
4.0 minutes/set
Stopper for Input/Exit Assembly
7
Replaced
Fungible thus not
2.0 minutes/set
replaced
169
8
Replaced
Contactor Chucks
120.0 minutes/set
Fungible but need
manual
adjustment
4 minutes/set
Exit Plate
9
Replaced
Replaced
0.5 minutes/set
0.5 minutes/set
Unsoak Plate
10
Replaced
Replaced
0.5 minutes/set
0.5 minutes/set
Fungible if use
11
Replaced
13.0 minutes/set
Transfer Pick Up Assembly
the Nexus’s
design head
170
Soak Cleaning Plate
12
Replaced
Replaced
0.5 minutes/set
0.5 minutes/set
160
TOTAL TIME (minutes)
11
Figure 6.14: Illustration of the hardware setup executed as parallel activity
171
Figure 6.15: Illustration of changeover process change with the optimized hardware
setups
6.5
Improvement of PnP ‘Teaching’ Phase
This as earlier mentioned is a required step to adhere whenever hardware
especially when pick up heads or contactor chucks are replaced where the ‘Teaching’
process will calibrate the hardware settings to perform at optimum condition during
actual production runs.
Though not many significant improvements were proposed in this phase, but
with the new improved changes to other phases, it will indirectly impact this process
phase from its current state. As both Dry Cycling phase and Standard Unit validation
are dependent on the completion of PnP ‘Teaching’, thus there will be no sequence
change from the current flow. As earlier identified, the 2 opportunity in the PnP
172
phase to be analyzed are the reduction in setup time and also the opportunity to
perform parallel activity during the actual PnP machine process time.
Table 6.11 shows the current, proposal and the implementation status in the
PnP ‘Teaching’ phase. Figure 6.16 shows the actual time study of the new PnP
Teaching phase with improved duration compared to the current practice. Figure 6.17
shows the breakdown of the PnP activity and its respective duration. The figure also
shows the implementation Pre VI activity from Dry Cycling phase as a parallel
activity during the actual Teaching process. Figure 6.18 shows the latest change in
the changeover flow with the improvement in the PnP Teaching phase.
Table 6.11: The improvement result of PnP teaching phase
No
1
Current
Proposal
Proposal Status
Before the start of the
As part of the early pre
Accepted – This proposal
activity, the
setup activity, the
has been implemented as
technician need to
technician has clear old
part of the new Pre Setup
clear the trays from
trays and ensure buffer
phase activity which is
the input and reject
stoker is filled with new
performed in parallel during
stoker and ensure
trays
operator’s housekeeping
correct trays are
activity. With this new
placed inside the
practice, the Pre Validation
buffer stoker
physical setup duration will
be reduced
2
Only 4 Standard Units Suggest Technician to
Accepted with condition-
are used for the PnP
collect >1 tray units
Technician will acquire 2
‘Teaching’ phase and
from EIMS so that the
full trays instead of 1 tray
the setup is slightly
extra units can be
since partial trays or units
delayed because
directly used during the
are not allowed in the EIMS
technician need to
PnP ‘Teaching’, the
process but since Standard
173
transfer the 4 units
Technician does not
Unit are controlled items
from the current full
need to perform
thus acquiring 2 trays are
tray to another empty
multiple transfer
based on availability
The system needs to
If Temperature Setup
Accepted – PnP Teaching
stabilize the
step in the TIU
phase does not require
temperature before
replacement phase can
temperature thus
the actual PnP
been move to be part of
eliminating the temperature
Teaching can begin
the Standard Unit
setup will help to minimize
Validation activity, thus
lost time during the actual
no temperature setting
process
trays before
transferring the units
to the handler’s PnP
3
or stabilization occur
during PnP Teaching
phase
4
Technician idle
Propose for technician
Accepted – To minimize
during the actual PnP
to start performing the
Technician idle time and
Teaching process
pre validation visual
Internal time during Dry
awaiting it to
inspection of the 5 tray
Cycling phase, this proposal
complete
mechanical units during
will be accepted. The pre-
this period
validation visual inspection
will be performed during
this idle period. Also, the
proposal to use the 3x
magnifier to perform
inspection is also approved
to ease the visual inspection
5
Technician perform
Find quicker way or
Rejected – Since there are
PnP ‘Teaching’ as
eliminate the need to
still hardware parts replaced
part of the
perform PnP Teaching
and also the ‘nest’ manually
requirement each time as they new optimized
adjusted, it is still necessary
174
Hardware parts are
hardware part setup
to perform the PnP
replaced
minimizes the required
Teaching. And no
number of parts to be
engineering data to proof
replaced, potentially
the risk is low
low risk
• Pre Physical setups and validation
• Pre system setup and configuration
• Actual Teaching Process
• Pre Validation VI of Mechanical Units
• Post system setup and configuration
• Post physical setup and validation
• Total 'Internal' time spent for the new
improvized PnP Teaching Phase
0.75
minutes
0.75
minutes
2.2
minutes
0.5
minutes
0.75
minute
4.95
minutes
Figure 6.16: Actual time study of the improved PnP teaching phase
175
Figure 6.17: Illustration of the PnP teaching phase activity breakdown
Figure 6.18: Illustration of the changeover process flow change with the improved
PnP teaching phase
176
6.6
Improvement on the Mechanical Unit Validation Phase
This is the first validation phase on ensuring a correct and successful
changeover before the equipment is handover to the production. Today technician
needs about 20 minutes to complete this dry cycling phase and it’s been highlighted
that this is a NVA activity but required per process today.
The proposal to eliminate and streamline this phase has been rejected as there
the data collected for changeover the last one quarter is still insufficient to show that
validation process can be eliminated. This proposal will be put on hold for now and
pursued when enough empirical data is available to assess the risk.
Nevertheless, the proposal to reduce the mechanical unit sample in this phase
has been agreed and the new process is to perform with 1 full tray of mechanical
units instead of 5 full trays. This has significantly help to reduce the overall duration
of the dry cycling phase. On the other hand, to minimize the technician idling time,
the Standard Unit pre validation post inspection will be performed in parallel the
actual dry cycling process is in progress. The introduction of the 3x magnifier to
technician to perform visual inspection has helped to improve efficiency and speed
up the process as well. All the other improvements proposed earlier were accepted
and successfully implemented and this is summarized in Table 6.12 below.
Figure 6.19 shows the new optimized flow for this Dry Cycling phase per the
improvement implementation. Figure 6.20 illustrates the buildup of new optimized
Dry Cycling phase while Figure 6.21 shows the actual time study collected based on
these improvements. Figure 6.22 illustrates the changeover process flow and duration
change with the optimized Dry Cycling phase.
177
Table 6.12: The result of dry cycling phase proposals
No
1
Current
Proposal
Proposal Status
Before the start of
The 5 trays of units are
Accepted – As the proposal
Dry Cycling phase,
collected earlier
allow the task to be
Technician will walk
together with the tool
converted from ‘Internal’ to
back to the tool shop
sets during the Pre
‘External’ and save the
to collect 5 trays of
Setup phase
duration of 2.5 minutes
Incase of any pre
Suggest technician to
Accepted – Mechanical
visual validation fail,
collect >5 trays of units
units are not controlled or
Technician will need
( or atleast an extra 1
limited collaterals, thus
go back to tool shop
tray of units) as buffer
Technician are allowed to
to get replacement
inhand that technician
collect 6 trays of units per
units
can replace immediately transaction during the Pre
mechanical units
2
Setup phase
3
Technician will need
Eliminate the need for
Rejected – No changes
to perform pre
technician to perform
allowed to the current
validation visual
pre validation visual
business process and tool
inspection before the
inspection by requesting shop admin R&Rs.
start of the Dry
the tool shop’s admin to
Cycling phase
inspect, check and store
only good mechanical
units for use
Eliminate the need to
Rejected - The Pre VI are
perform pre-validation
required to ensure only
VI steps as this is NVA
defect free units are tested
during the dry cycling
Move the ‘Internal’
Accepted- This task can be
practiced Pre VI during
practiced as an ‘External’
the Dry Cycle phase to
activity when it is
an ‘External’ task
performed in parallel during
178
the actual PnP Teaching
process
4
Technician perform
Suggest to use the
Accepted – The use of the
the both pre and post
readily available 3x
3x magnifier to perform
VI using bare or
diopter or magnifier
visual inspection are much
‘naked’ eye
placed at each work
quicker and efficient. Able
inspection
desk
to reduce the inspection
duration and improve the
reject identification
5
Technician need to
This activity can be
Accepted – This has been
clear all the trays
performed part of the
agreed to be part of the Pre
from input/exit
Pre Setup activity
Setup steps and reduce the
stocker and place new
pre validation physical
trays into the buffer
setup
stocker
6
Before Dry Cycling
Temperature is not
Accepted- Dry Cycling
process can
needed for Dry Cycling
process does not need
commence, the
thus temperature setting
temperature and with this
equipment
can be eliminated.
activity eliminated, able to
temperature need to
Move the temperature
save the Internal time spent
stabilize
setting in TIU
during temperature
replacement phase to
stabilization
Standard Unit
Validation
8
Technician idle
Opportunity for
Accepted- Opportunity to
during the actual Dry
Technician to perform
reduce the ‘Internal’ during
Cycling process
parallel activity during
the Standard Unit validation
this period and
phase
recommended to
perform the Pre VI for
the 1 tray of Standard
Units
179
9
10
Technician will
Suggest to perform the
Rejected – Standard Unit
perform Post VI and
Post VI during the
validation can only
ensure all units are
Standard Unit validation commence upon
100% defect free
period which can save
confirmation that all units
before officially
‘Internal’ tine
are 100% defect free with
ending the Dry Cycle
all mechanical units
Phase
working in order
Upon completion of
Keep the units aside on
Accepted – Eliminate the
the dry cycling phase,
the trolley and return
‘Internal’ task and waste in
Technician will
together towards the end motion
arrange units into tray
of the changeover
and return them back
to the tool shop
11
Current changeover
Eliminate the need for
Rejected - Dry Cycling is
process requires the
Dry Cycling validation
needed as the ‘gate’ process
validation data from
as this is a NVA but
to identify any abnormality
Dry Cycling with
required task
with the changeover setup
100% defect free
Streamline the Dry
Rejected – Mechanical
units from 5 trays.
Cycling phase function
damage to Standard Units
And this validation
with the Standard Unit
are too risky and costly.
process delays the
validation phase so can
Also both phase have
overall changeover
eliminate the Dry
significant different
Cycling phase
function
Minimize the no. of
Accepted with condition –
units or trays used for
Dry Cycling phase reduced
the Dry Cycle phase
from 5 full trays to 1 full
from current 5 trays to 1
tray with condition if any
tray
failure found first time,
once issue fix, the
revalidation will need to be
performed with 5 full trays
180
Perform prevalidation system
setup
Perform prevalidation
physical setup
Perform postvalidation system
setup
Actual Dry
Cycling Process
Perform postvalidation
physical setup
Perform postvalidation visual
inspection
Figure 6.19: The improved dry cycling activity flow
Figure 6.20: Illustration of the improved dry cycling phase buildup
181
•Pre dry cycling system setups
•Pre dry cycling physical setups
•Actual dry cycling process
•Post dry cycling physical setups
•Post dry cycling system setup
•Post dry cycling visual validation
•Total 'Internal' time spent in the new
improvised Dry Cycling Phase
1.0
minutes
0.2
minutes
1.5
minutes
0.15
minutes
1.0
minute
0.5
minutes
4.35
minutes
Figure 6.21: Actual time study of the improvised dry cycling phase
182
Figure 6.22: Illustration of changeover process change with the improved dry
cycling phase
6.7
Improvement of the Standard Unit Validation
This is the last phase of the validation before the equipment can be officially
handover back to the production. This phase is also considered critical as the
outcome of the standard unit run determines the full completion of the changeover.
As highlighted earlier, the proposal to eliminate or streamline the validation
phases were rejected due to insufficient data support. Also, as Standard Unit
validation uses only 1 full tray, no further sampling reduction is allowed.
Nevertheless, some of the earlier implemented upfront setup and parallel activities
has contributed in reducing the overall duration of this validation phase. The
summarized outcome of this phase is listed in Table 6.13 below;
183
Table 6.13: The result outcome of standard unit validation
No
1
2
Current
Proposal
Proposal Status
Standard units are
Standard Units are
Accepted – This has been
collected before the
prepared upfront before
part of the Pre Setup
PnP Teaching phase
the changeover starts
activity to convert the
begins
during the Pre Setup
Internal time to External
phase
task
In case of any pre
Suggest Technician to
Accepted with condition-
visual validation fail,
collect >1 tray units
Technician will acquire 2
Technician will need
from EIMS so that if
full trays instead of 1 tray
go back to EIMS
there is any Pre VI
since partial trays or units
room to get
failed units, they can be
are not allowed in the EIMS
replacement units
directly swapped and
process but since Standard
used for the Standard
Unit are controlled items
Unit run
thus acquiring 2 trays are
based on availability
3
Technician will need
Eliminate the need for
Rejected – No changes
to perform pre
technician to perform
allowed to the current
validation visual
pre validation visual
business process and EIMS
inspection before the
inspection by requesting admin R&Rs.
start of the Standard
the EIMS’s admin to
Unit phase
inspect, check and store
only Standard units for
use
Eliminate the need to
Rejected - The Pre VI are
perform pre-validation
required to ensure only
VI steps as this is NVA
defect free units are tested
during the Standard Unit
Move the ‘Internal’
Accepted- As this Pre VI
practiced Pre VI during
validation is only for 1 tray,
184
the Standard Unit phase
thus will be done in parallel
to an ‘External’ task
during the actual Dry
Cycling phase
4
Technician perform
Suggest to use the
Accepted – The use of the
the both pre and post
readily available 3x
3x magnifier to perform
VI using bare or
diopter or magnifier
visual inspection are much
‘naked’ eye
placed at each work
quicker and efficient. Able
inspection
desk
to reduce the inspection
duration and improve the
reject identification
5
Technician need to
This activity can be
Accepted – This has been
clear all the trays
performed part of the
agreed to be part of the Pre
from input/exit
Pre Setup activity
Setup steps and reduce the
stocker and place new
pre validation physical
trays into the buffer
setup
stocker
6
Temperature setting is Move this activity to be
Accepted- Correct
performed part of the
part of Standard Unit
temperature (115’C) is
TIU replacement
run validation
needed during the Standard
activities
Unit run but not needed for
PnP Teaching and Dry
Cycling phase. Temperature
setup in Standard Unit run
can be performed as a
parallel activity when they
pre physical setup is
performed
10
Upon completion of
Keep the units aside on
Accepted – Eliminate the
the dry cycling phase,
the trolley and return
‘Internal’ task and waste in
Technician will
together towards the
motion
arrange units into tray
end of the changeover
and return them back
185
to the EIMS room
11
AEPT state change is
AEPT state changed is
Accepted – As long all the
from ‘SDT-
performed once
units passed with no rejects
Conversion’ to
Standard unit validation
and the supervisor agrees.
‘PROD’ is performed
complete as part of the
All other NVA like
towards the end of the
post system setup
housekeeping can be
Wrap Up activity
performed upon completion
phase
12
Current changeover
Eliminate the need for
process requires the
Standard Unit validation validation is needed as the
validation data from
as this is a NVA but
‘gate’ process to identify
Standard Unit with
required task
any abnormality with the
100% pass of all the
Rejected - Standard Unit
changeover setup
units tested And this
Streamline the Dry
Rejected – Mechanical
validation process
Cycling phase function
damage to Standard Units
delays the overall
with the Standard Unit
are too risky and costly.
changeover
validation phase so can
Also both phase have
eliminate the Dry
significant different
Cycling phase
function
Use a different TP for
Rejected – to avoid many
Standard Unit validation noise or variables in the
which is shorter in
shop floor and also to
download time and
standardize the TP use.
quicker to execute
during the actual
validation process
.
Figure 6.23 shows the new activity flow for the Standard Unit Validation
phase while Figure 6. 24 show the actual time study taken to complete this validation
phase based on the new improvements while Figure 6.25 shows changeover flow and
duration change with the new optimized Standard Unit validation phase.
186
Perform prevalidation system
setups including
Temperature Setup
Perform prevalidation physical
setups
Actual validation and
testing process
Perform postvalidation system
setups
Perform postvalidation physical
setups
Figure 6.23: Activity flow of the improvised standard unit validation phase
•Pre-Validation System setup
•Machine init and Temperature setting
•Pre-Validation Physical Setup
•Actual Validation Processing
•Post-Validation System setup
•Post-Validation Physical setup
•Total Internal time spent in the improved
Standard Unit run
3.0
minute
1.0
minutes
3.2
minutes
0.95
minutes
1.0
minutes
9.1
minutes
Figure 6.24: Actual time study of the improvised validation phase
187
Figure 6.25: Illustration of changeover process change with new improvised
standard unit validation phase
6.8
Elimination of Wrap Up phase
The last phase of the current changeover process is the elimination of Wrap
Up activities phase which consists mostly NVA activities to the actual changeover.
As indicated in the proposal phase, the major change to occur at this phase is the
reverse flow of the activities where technician will log off the AEPT state change
upon completion the Standard Unit validations and thus officially complete the
changeover process. All the other NVA activities removed from previous stages will
be accumulated with the other Wrap Up phase activities and performed together as
188
External elements. The new phase will be known as Post Changeover activities with
no impact to the changeover duration.
Table 6.14 summarizes the new changes to the changeover process by the
elimination of the Wrap Up phase. Figure 6.26 shows the new Post Changeover
activities ‘mirrored’ from the Wrap Up Activity phase while Figure 6.27 shows the
changeover process flow with the elimination of Wrap Up activity phase.
Table 6.14: The result outcome of the wrap up phase elimination
No
1
Current
Proposal
Proposal Status
The AEPT state
Technician to perform
Accepted – AEPT state
change is performed
AEPT state change as
change will be move to be
st
towards the end of the
the 1 activity of the
part of the Standard Unit
Wrap Up Activity
phase
validation phase and
performed part of the Post
System Setup activity
2
The Wrap Up phase
Eliminate this phase and Accepted – Changeover
of the changeover
perform all this
official end process will be
consist activities like
activities together with
the Standard Unit validation
housekeeping,
the returning of the
and once it complete, the
cleaning, arranging
parts once AEPT state
equipment will be ready to
and post
has change.
be handover back to
communication
production. All the other
NVA activities will be
performed in this new phase
called the Post Changeover
phase
189
Change AEPT state
from 'SDT-Conv' to
'PROD'
Perform basic
housekeeping
Inform
supervisor/operator
of completion
Rearrange all items
including TIU and
units on trolley
Rearrange
changekit and tools
into respective
boxes and on trolley
Change 'sticky' tag
to UP
Send all items back
to respective rooms
Figure 6.26: The new flow of post changeover activities replacing wrap up phase
6.9
The New Optimized Test Handler Changeover Process
With all the above improvement and implementation, the whole changeover
process structure has changed significantly to be more optimized and lean with lesser
duration to complete. Table 6.15 shows the actual time study based on the average
data collected for the whole of Q1’2011. Figure 6.27 shows the new optimized
changeover process flow.
Thus from the data above, it is clear that the new changeover process duration
is 32 minutes compared to the 240 minutes previously. Besides that, the number of
stages and phases has significantly reduce with more parallel activities are executed.
190
In summary, the techniques proposed has enabled to reduce the changeover
duration to 32 minutes and achieving the objective. The benefits and other ROIs will
be analyzed in the following discussions.
Table 6.15: Time study of the overall new optimized changeover process
191
Figure 6.27: The new optimized changeover flow
6.10
Return on Investment (ROI) analysis
With the new improved and optimized changeover process, the benefits and
advantages of achieving this can be summarized through the ROI analysis:
6.10.1 Capital and cost savings
The ability to convert the equipment rapidly has helped the organization to
response to the volatile market demand more systematically. With this new
changeover process, the factories will eliminate the tool dedication policy and also
192
maximize the equipment utilization due to low conversion downtime. The impact of
this can be analyzed through the cost benefit analysis as shown in Figure 6.28.
The ROI analysis shows the capital tool saving or eliminating the need to
purchase new equipments with this improvement. It is estimated at least 1 tool
purchase can be eliminated in Q3’2011 for the case study factory which is about
USD $ 1.5M in savings as shown on Table 6.16. Besides, the other minor savings
can be materialized by the equipment collaterals savings especially in kits, TIU, pogo
pins among other with an estimated sum of USD $50,0000.
With the reduction in capital and collateral purchase together with a better
overall utilization, the cost per unit at Test operation was able to be reduced to a
grossing USD $ 0.25 and exceeding the goal target. With this the overall cost per
unit was able to be reduced to almost USD $0.30 as shown in Figure 6.29.
Though there was some minor expenditure on the R&D and re-designing of
the hardware parts, but since the cost is below USD$1k, thus it is negligible in the
overall finance analysis.
In summary, the optimized changeover has brought in positive cost and
capital ROI for the overall organization.
193
Figure 6.28: The capital purchase ROI analysis
Table 6.16: The actual tool saving by factories for Q3’11
Q3'11
Available
Inventory
Pre
Post
factory(KMCO)
9
10
9
Factory B
7
8
6
Factory C
3
3
3
Total
19
21
18
Case Study
194
Figure 6.29: The cost breakdown analysis (post improvement)
6.10.2 Total Utilization Indicator Improvement
The impact to the overall utilization can be analyzed from the total good
production utilization and from the idling duration of the technician itself. Figure
6.30 shows an example of an actual utilization study based on before and after
improvement of the changeover. The data collected post improvement though
slightly below the corporate goal but it showed significant improvement compared to
previous actual data in Q4’11.
195
Figure 6.30: The post improvement utilization analysis
A huge reduction in both downtime and also idling can be demonstrated due to the
improvements in changeover process. With further enhance and fine tuning, the
utilization and exceed the expectation of the corporation’s goal target.
On the other hand, the significant reduction in the technician’s idling time can
be shown in Figure 6.31. The new improved and compress changeover able to
minimize a technician idle time below 1 minute in order to achieve a changeover in
196
30 minutes. There are still a few minor idling instances in the new changeover and
more analysis will be done to reduce this further to an ideal goal of ‘zero’ idling.
T
O
Figure 6.31: The post improvement technician idling time reduction
6.10.3 Leaner and Efficient Training
The new improved and implemented process is now well documented into
both official specification (spec) called the ‘Standard Work Instruction’ module
where technician will be trained and assessed based on this latest changes. All the
current certified technicians will also undergo refresher training to ensure all shift
technicians are aligned and standardized on the practice. The new and simplified
process has truncated the duration of the training from 8 hours class to a 3 hours
197
class including of both theory and practical. With a more lean and efficient training,
more technicians are to be trained and disseminate quickly to the production shifts.
6.11
Critical Appraisal
The SMED and TRIZ techniques used to optimize the changeover in this case
study helped to significantly reduce the duration from an average of 4 hours to ½
hour. If understanding the actual contribution of TRIZ in this changeover
optimization, the Table 6.17 shows the forecasted changeover duration changes of
using a single changeover technique like SMED alone compared with an integrated
technique of SMED and TRIZ. The table clearly shows the major contribution of
TRIZ especially in hardware designing and also process segmentation.
Table 6.17: The contribution of TRIZ
Duration
No.
Changeover Status
(minutes)
1
Pre Improved changeover
240
Improvement with SMED
105
2
3
techniques only
Improvement with SMED + TRIZ
32
techniques
Nevertheless, as per Shingo’s SMED goal, a changeover is considered
successful only if the whole setup can be completed below 10 minutes thus the new
improved changeover in this case study can’t be considered as fully optimized and
successful. Though, the 3 value added process in this changeover can be completed
in 14 minutes but the remaining 18 minutes due to NVA like calibration and
validation gates the whole process optimization.
198
Also, the available TRIZ principals and SMED techniques are aligned mostly
to hardware and process improvement but lesser focus or tool proposed for areas like
software, IT or network computing which is also part of most processes today. Due
to this, the TP download phase cannot be fully optimized.
Though many of the identified problems and proposed /implemented
solutions are common sense and logical at basic but the introduction of SMED and
TRIZ techniques helped solving problems in a standardized and structured manner.
The most important element that was less focused in this case study was the
human factor improvement the continuous motivation and training to the personnel
ensures a better process sustainability and further enhancement.
6.12
Future Recommendation / Studies
Some of the suggested recommendation and studies for the future are:
i.
Continue to pursue improvements to reduce the current test handler
changeover to below 10 minutes by;
a. Eliminating the validation and calibration phases with more
empirical data
b. Reducing the TP download phase
c. Improve the hardware setup further especially the non ‘fungible’
parts by redesign or eliminate the NVA activities
ii.
Apply the TRIZ principles and SMED techniques suggested in this case
study for any other applicable semiconductor based equipment
iii.
Extend the scope of TRIZ by introducing the other advance TRIZ tool’s
like Standards and ARIZ
199
6.13
Summary
This chapter covered the results and outcome based on the earlier proposed
counter measures. Though many was accepted and implemented, but a handful of
proposal were rejected due to current circumstances. Nevertheless, the outcome of
the optimized changeover met and aligned with the initial objective.
CHAPTER 7
CONCLUSION
This project focused on a changeover case study where the non optimized
process affected the organization from both cost and productivity point of view. The
equipment, a test handler consists of 8 major changeover steps with average
completion duration of 4 hours. The hardware part setups and the non standardized
process was the major gating issue of the changeover. Thus, to counter the issues and
gaps identified, the techniques from both SMED and TRIZ was introduced and
integrated into the changeover process. Solutions included redesigning hardware
parts and rescheduling the process sequence help to minimize the ‘Internal’ time and
eliminate the Non Value Added (NVA) activities. One of the major breakthroughs
was redesigning a hardware part which helped to reduce the setup time from an
initial 120 minutes to a staggering 4 minutes using a TRIZ and SMED concept. With
the implementation of the other improvement proposals, the changeover duration was
successfully reduced from the current 4 hours to ½ hour with no safety and quality
concern. The case study showed the integration of these techniques helped to
achieve a better optimized changeover process. The improved changeover resulted in
cost, capital and resource benefit for the organization. Nevertheless, there are still
room for future consideration and improvement that can be explored upon.
In summary, the integration of SMED and TRIZ techniques helped to
successfully improve the changeover process for a test handler in a semiconductor
industry from 240 minutes to 30 minutes.
REFERENCES
Anderton, J. (2009). Quick Changes. Canadian Plastic. January/February 2009,
29-30.
Angeles, Rolly S. (2009). World Class Maintenance Management – The 12
Disciplines. Central Books Supply Inc.
Association for Manufacturing Excellence (2009). Sustaining Lean: Case Studies
in Transforming Culture. New York: Productivity Press.
Farlow, D. (2005). Efficient Line Changeover. SMT magazine, March 2005, 4445
Fey, V., and Rivin, E (2005). Innovation on Demand: New Product
Development using TRIZ. Cambridge University Press.
Karlsson, C., and Ahlsrrom, P.(1996). Assessing changes towards lean
production, International Journal of Operations & Production management.
16 (2), 24-41.
Kearney, W. (1997). A proven receipt for success: the seven elements of work
class manufacturing, National Productivity Review. 16, 67-76.
Lev, S. (2009). Introduction to TRIZ. Altshuller Institute Online.
Levinson, W., & Rerick, R. (2002). Lean Enterprise: A Synergistic Approach to
Minimizing Waste. Milwaukee, WI: ASQ Quality Press
Liker, J.K, and Meier, D (2006). The Toyota Way Fieldwork: A Practical Guide
for Implementing Toyota’s 4Ps. United States: The McGraw- Hill Companies.
202
Lotter B., and Wiendahl H.P.(2009). Changeable and Configurable Assembly
systems. Springer.
McIntosh, R.I., Culley, S.J., Mileham A.R., and Owen, G.W. (2000). Critical
Evaluation of Shingo’s SMED Methodology. International Journal of
Production Research. 38(11), 2377-2395. Taylor and Francis Ltd.
McIntosh, R.I., Culley, S.J., Mileham A.R., and Owen, G.W.( 2001).
Changeover improvement a maintenance perspective. International Journal of
Production Economics. 73( 2), 153 – 168. Taylor and Francis Ltd.
Mann, D (2001). TRIZ: The Theory of Inventive Problem Solving. 10(2), 123125. Blackwell Publishing Ltd.
Martin G. M. (2005). What is TRIZ? From Conceptual Basics to a Framework of
Research. Creativity and Innovation Management, 14, 3-13.
Moxham, C., and Greatbanks, R. (2001). Prerequisites for the implementation of
the SMED methodology: A study in a textile processing environment. IJQRM.
18 (4 ), 404-414. Taylor and Francis Ltd.
Nakajima, S. (1988). Introduction to TPM. Cambridge , MA : Productivity Press.
Orloff. M (2003). Inventive Thinking Through TRIZ: A Practical Introduction.
Berlin: Springer.
Reik, M.P., and McIntosh, R.I. (2006). Formal Design for Changeover
Methodology. A Case Study. Proc. IMechE, Part B. 220(528),1237-1247
Ross & Associates Environmental Consulting, Ltd. (2003). Lean Manufacturing
and the Environment: Research on Advanced Manufacturing Systems and the
Environment and Recommendations for Leveraging Better Environmental
Performance. United States: Environmental Protection Agency.
203
Santos, J., Wysk, A.R., Torres, M.J. (2006). Improving production with lean
thinking. New Jersey: John Wiley & Sons, Inc.
Shingo, S. (1984). A Revolution in Manufacturing: The SMED system.
Cambridge , MA : Productivity Press.
Simon, L., and Vladamir, P (2009). TRIZ Body of Knowledge. International TRIZ
Asscociation (MA TRIZ), Alshuller Institute, 1-9.
Sousa R.M., and Lima R.M (2009). An Industrial Application of Resource
Constrained Scheduling for Quick Changeover. Proc. IEEM, 189-193.
Taylor, B.W. (2006). Introduction to Management Science. (9th ed.). Virginia:
Prentice Hall.
Van, J. V. (2001). Transport Developments Support Fast Changeover In High
Volume SMT production. SMT magazine, May 2004, 66-69.
Van Goubergen, D., and Landeghem, H.(2002). Rules for Intergrating fast
changeover capabilities into the new equipment design. Flexible
Manufacturing. Elsevier Science Ltd.
Vardeman, S. B.(2010). The Impact of Dr. Shingo on Modern Manufacturing
Practices. IE 361. Utah State University.
Whitney D.E., (2004). Mechanical assemblies – their design, manufacture and
role in product development. Oxford University Press.
Womack, P.J., Jones, T.D., Roos, D. (1990). The Machine that Changed the
World: The Story of Lean Production. New York: Harper Perennial.
Womack, P.J., Jones, T.D. (2003). Lean Thinking. New York: Free Press.
Yyes, D.G (2006). SMED reduces changeover time : A case study in a
204
food industry. Food Engineering and Ingredients, November 2006, 32-34.
www.altshuller.com
www.intel.com
www.leanlearningcenter.com
www.ross-assoc.com/lean
www.trizland.com
205