supply chain performance evaluation and improvement
TRANSCRIPT
Ref. code: 25605422300342CMU
SUPPLY CHAIN PERFORMANCE EVALUATION AND
IMPROVEMENT METHODS: APPLICATION OF SCOR
MODEL AND FUZZY QFD
BY
PIYANEE AKKAWUTTIWANICH
A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF
PHILOSOPHY (ENGINEERING AND TECHNOLOGY)
SIRINDHORN INTERNATIONAL INSTITUTE OF TECHNOLOGY
THAMMASAT UNIVERSITY
ACADEMIC YEAR 2017
Ref. code: 25605422300342CMU
SUPPLY CHAIN PERFORMANCE EVALUATION AND
IMPROVEMENT METHODS: APPLICATION OF SCOR
MODEL AND FUZZY QFD
BY
PIYANEE AKKAWUTTIWANICH
A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF
PHILOSOPHY (ENGINEERING AND TECHNOLOGY)
SIRINDHORN INTERNATIONAL INSTITUTE OF TECHNOLOGY
THAMMASAT UNIVERSITY
ACADEMIC YEAR 2017
Ref. code: 25605422300342CMU
ii
ACKNOWLEDGEMENT
With boundless love and appreciation, I would like to extend my heartfelt
gratitude and appreciation to the people who help me to complete this PhD dissertation,
and bring this study into the reality. I would like to extend my profound gratitude to the
following.
First and foremost, I would like to express my sincere appreciation to my
beloved advisor, Assoc. Prof. Dr. Pisal Yenradee, for his expertise, continuous guidance,
plenty of time spent, and your caring to bring this PhD study into success. His
knowledge and advice have helped me to complete this dissertation. Thanks for always
exploring a new theory with me, understanding, and enduring. Throughout the research
years, there were several failures, tears, restart, and depress along the PhD process, but
he never taught me to give up. Instead he supervised me that every problem can be
solved with patience. Along the six years of being your PhD advisee, and four years of
the undergraduate years at SIIT, I can count on him as my “second father”.
Next, I wish to express my sincere thanks to Assoc. Prof. Dr. Ruengsak
Kawtummachai, who is the first who introduce the idea of pursuing the PhD degree to
me. Apart from being my external committee who frequently examine my research
progress from the beginning and suggest the useful advice until this dissertation
complete, he also provides a great support through the ups and downs of my life during
the period of study. Without his kindness and continuous caring, I could not come this
far.
My gratitude goes out as well to all of my committee members; Assoc Prof. Dr.
Navee Chiadamrong, Assoc. Prof. Dr. Jirachai Buddhakulsomsiri, and Asst. Prof. Dr.
Suchada Rianmora. I am extremely grateful for your assistance and suggestion,
especially at the early years of the study to criticize my research progress and provide
the sense of how researcher work to meet with the doctoral standard. My appreciation
Ref. code: 25605422300342CMU
iii
to the Industrial Engineering faculty member, P’Noi, for her generosity and the great
assistance while I am here.
To my family; my mother, my grandmother, my aunt, my grandfather, and all
of Akkawuttiwanich’s family members to the endless love and strong support when I
need it.. Especially thanks to my most beloved woman of my life, my mother, who wish
to see her daughter to be called Dr, who encourage me that PhD must be one of a
destination of my life, who love me unconditionally, and who always provide a spiritual
support to me that my study is going to be complete. Thanks for her love and caring,
and nurturing me to be who I am. I love you to the moon and back. Thank you to my
aunt, who always hearten me with a positive attitude, tell me to be patient, and one day
the success will come.
To my husband, Wynn, you are the only one who see where it all starts. Thank
you for helping me to submit the application since the first day, accompany me through
the laugh and tears in the long PhD journey, support me when I run out of the belief
that I can do it, and for always be there at my side until the day of my achievement.
Thank you for your never-ending love. You are the inspiration to me to make this all
complete. Thank you to Wichitphan family; grandmother, mother, and brother for your
kindliness and best wishes.
Finally, I express my thanks to the sisterhood and brotherhood at SIIT,
especially Dr. Tantikorn Pichpibul, who have listened to me during my nervousness and
the cherish support for all period of the study. To my school colleagues; Oil, Tarn, June,
May, B, Soh, Wan, and the Logistics class of 2014 who have helped my learning an
enjoyable and stimulating experience during my study.
Last but not least, I would like to thank you to myself for having the forbearance,
enthusiasm, and determination to complete this research. I dedicate this dissertation to
my family for their constant support and eternal love. I love you all dearly.
Ref. code: 25605422300342CMU
iv
Abstract
SUPPLY CHAIN PERFORMANCE EVALUATION AND IMPROVEMENT
METHODS: APPLICATION OF SCOR MODEL AND FUZZY QFD
By
PIYANEE AKKAWUTTIWANICH
Master of Science (Supply Engineering and Logistics), University of Warwick, 2008
Doctor of Philosophy (Engineering and Technology),
Sirindhorn International Institute of Technology, Thammasat University, 2017
The effective supply chain design is evaluated by the successful implementation of
a strategy deployed, and the index that determine a successful implementation is known as
a performance measurement system (PMS). Based on the literature reviews, a good
performance evaluation system should be able to anticipate outputs and provide the
mechanism for performance improvement. The aim of this dissertation to pay an attention
on the performance evaluation and improvement in the supply chain system.
This dissertation consists of five chapters. The first chapter deals with the
introduction of the PMS and the SCOR model where the research problems are identified.
In chapter 2, the literature reviews are provided. It consists of the theory of the SCOR
model, the MILP model, uncertainty and the fuzzy set theory (FST), and, the fuzzy QFD
approach. These philosophies are provided as the background to support the establishment
of the proposed methodology. In chapter 3, this dissertation develops a methodology to
evaluate the SCOR KPIs by using the predictive MILP model with fuzzy parameters. The
novelty of this chapter is to relate the manufacturing parameters to the SCOR KPIs, and
Ref. code: 25605422300342CMU
v
use the MILP model with fuzzy parameters to enable the performance prediction process.
The results of this chapter indicate that the proposed methodology can use as a tool to
perform the predictive process when the manufacturing parameters are changed. In chapter
4, the dissertation proposes the fuzzy QFD approach to manage the SCOR KPIs for
improvement. The eight-step QFD approach for managing the SCOR KPIs are proposed
where the SCOR KPIs are identified as “Whats”, and the manufacturing capabilities are
identified as “Hows”. This dissertation is the first to attempt to develop the fuzzy QFD
approach to combine with the SCOR model in performance management issue. In chapter
5, the findings of each chapter are recapped, then the theoretical and practical contributions
in this research are summarized. Finally, the limitations and recommendation are outlined.
Keywords: SCOR, MILP Model, Fuzzy QFD, Performance measurement, Supply Chain
Management, Case study.
Ref. code: 25605422300342CMU
vi
Table of Contents
Chapter Title Page
Signature Page i
Acknowledgement ii
Abstract iv
Table of Contents vi
List of Tables x
List of Figures xii
1 Introduction
1
1.1 Definition of the Performance Measurement System (PMS), and
role of PMS in the supply chain management
1
1.2 Research problem statements 5
1.3 Overview of this dissertation 12
2 Literature reviews
13
2.1 A Supply Chain Operations Reference (SCOR) model 13
2.1.1 SCOR processes 14
2.1.2 SCOR metrics 16
2.2 The SCOR model in performance evaluation 19
2.2.1 Application of the SCOR model by using system simulations 20
2.2.2 Application of the SCOR metrics to other decision support
model and methodologies
21
2.2.3 SCOR model that decompose a problem into a hierarchical
structure using Analytical Hierarchy Programming (AHP)
23
2.2.4 Case studies using SCOR model 24
Ref. code: 25605422300342CMU
vii
2.2.5 The relationship of SCOR model to other external factors 25
2.3 The MILP model and its applications 26
2.3.1 Fundamental of the mixed integer linear programming (MILP)
model
26
2.3.2 Application of the MILP model for supply chain production
planning
28
2.4 Uncertainty in the supply chain system and fuzzy set theory 34
2.4.1 Fuzzy set theory (FST) 36
2.4.2 Defuzzification to crisp sets 38
2.4.3 The fuzzy MILP model for supply chain planning under
uncertainties
39
2.5 Quality function deployment (QFD) 42
2.5.1 Fundamentals of the Quality function deployment (QFD) 42
2.5.2 Further process after the QFD 47
2.5.3 Fuzzy QFD 48
2.6 Concluding remarks 52
3 Evaluation of SCOR KPIs using a predictive MILP model with
fuzzy parameters
55
3.1 The proposed methodology to evaluate the SCOR KPIs 55
3.2 The predictive model 56
3.2.1 The MILP model 56
3.2.2 The MILP model with fuzzy parameters. 61
3.3 The proposed methodology to evaluate the SCOR KPIs 63
3.3.1 Percent of orders delivered in Full (RL2.1) 64
3.3.2 Make cycle time (RS2.2) 64
3.3.3 Upside Make Flexibility (AG2.2) 65
Ref. code: 25605422300342CMU
viii
3.3.4 Upside Make Adaptability (AG2.7) 66
3.3.5 Downsize Make Adaptability (AG2.12) 66
3.3.6 Cost to make (CO2.3) 68
3.3.7 Inventory days of supply (AM2.2) 68
3.3.8 Return on make fixed assets 68
3.3.8 Return on make working capital 69
3.4 Data collection and case study 69
3.4.1 Cost structure and inventory holding policy. 71
3.4.2 Current fixed assets, estimated accounts receivable, and
Accounts payable.
72
3.5 Results and discussion 73
3.5.1 Outputs from the predictive model 73
3.5.2 The SCOR KPIs 74
3.6 Concluding remarks 77
4 Fuzzy QFD approach for managing SCOR performance indicators
79
4.1 The proposed methodology to manage SCOR KPIs using fuzzy
QFD
79
4.1.1 Fuzzy QFD approach for managing SCOR KPIs 81
4.2 Data collection and case study 88
4.2.1 Cost structure and inventory holding policy 89
4.2.2 Options to increase and decrease the production capacity 90
4.2.3 Current fixed assets 91
4.2.4 Opinions of Decision Makers (DMs) 91
4.3 Results and discussions. 93
4.3.1 Current SCOR KPIs of the company. 94
Ref. code: 25605422300342CMU
ix
4.3.2 Selection of high priority TIs to improve SCOR KPIs 96
4.3.3 Relationships between results in chapters 3 and 4 102
4.4 Concluding remarks 102
5 Conclusions
105
5.1 Summary of the research 105
5.2 Key Contributions of the research 109
5.3 Limitations, and recommendation for further study 111
References 113
Appendix 131
Appendix A
132
Ref. code: 25605422300342CMU
x
List of Tables
Tables Page
1.1 PMS in the supply chain system as categorized by Akyuz and Erkan (2010) 2
2.1 The SCOR performance attributes 17
2.2 The SCOR Level-1 metrics 18
2.3 The modeling approach and purpose of the model 30
2.4 The shared information process contained in the supply chain planning
model
32
2.5 Types of QFD methodology and purpose of study 50
3.1 SCOR performance attributes and level 2 KPIs used in this dissertation. 64
3.2 Options to increase production capacity and the estimated lead time. 65
3.3 Options to decrease production capacity and the estimated lead time. 67
3.4 Operating cost information 71
3.5 Inventory holding policy of the company 72
3.6 Estimated company's total fixed assets 72
3.7 The fuzzy parameters used in the MILP model 73
3.8 Outputs from the MILP model, and MILP model with fuzzy parameters. 74
3.9 SCOR KPIs of the company 75
4.1 SCOR KPIs focusing on Make process and definitions used in this
dissertation
82
4.2 List of possible Technical Improvement actions (TIs) in a manufacturing
system
86
4.3 Operating cost information 89
4.4 Inventory holding policy 90
4.5 Options to increase production capacity 90
4.6 Options to decrease production capacity 91
Ref. code: 25605422300342CMU
xi
4.7 Estimated company total fixed assets 91
4.8 DMs’ relative importance on SCOR KPIs 92
4.9 List of corresponding TIs and their implementation lead time 93
4.10 Degree of influence of TIs on SCOR KPIs 93
4.11 Current SCOR KPIs of the company 94
4.12 Current revenue-cost structure of the company 96
4.13 Derivation of the average importance rating, competitive analysis, and
final importance rating
97
4.14 Relative importance (weight) of WHATS ( *~mW ) and the relationship score
( mnr~ )
98
4.15 Final rating and ranking of TIs 98
4.16 The new SCOR KPIs after improvement 99
4.17 The total Revenue-Cost structure obtained from LP model after
performance improvement
101
Ref. code: 25605422300342CMU
xii
List of Figures
Figures Page
1.1 Block diagram of the overall methodology 10
2.1 The SCOR model with six management processes (APICS,2016) 15
2.2 A hierarchical structure of SCOR 16
2.3 Set A (top), and the crisp set A (bottom) 37
2.4 A fuzzy set H 37
2.5 A fuzzy set with λ cut 38
2.6 A House of Quality (HOQ) (Left), and the HOQ with detailed description
(Right)
43
2.7 The HOQ planning matrix (Bozdana, 2007) 45
2.8 The typical 4 phases QFD model 47
2.9 A triangular fuzzy number 49
3.1 Block diagram of the SCOR KPIs evaluation procedure 56
3.2 Structure of manufacturing system 57
3.3 The proposed procedure to evaluate Upside Make Flexibility 65
3.4 The proposed procedure to evaluate Upside Make Adaptability 66
3.5 The proposed procedure to evaluate Downsize Make Adaptability 67
3.6 The manufacturing process of a case study 71
3.7 Graphical representation of the SCOR KPIs 75
4.1 Block diagram of the research methodology 80
4.2 Fuzzy QFD approach for managing SCOR KPIs 81
4.3 Linguistic representation of U 83
4.4 Relationship matrix between “Whats” and “Hows” 86
4.5 Current production process of case study 88
4.6 Graphical representation of current SCOR KPIs 95
Ref. code: 25605422300342CMU
xiii
4.7 Graphical representation of new SCOR KPIs 100
Ref. code: 25605422300342CMU
1
Chapter 1
Introduction
The performance evaluation is defined as the process of quantifying the effectiveness
of action. In the organization, the objective of performance measurement is to identify
success, to recognize process bottleneck, and to communicate the right messages for
the improvements. In the supply chain management, measuring the supply chain
performance can help the company to disclose the gap between planning and execution.
The effective organization management requires the framework, information, and tools
to support a decision-making process and to identify the area for improvement. In this
chapter, a basic definition of performance measurement system (PMS), and a role of
PMS in the supply chain management are presented. Then, the problems of current PMS
are drawn as a dissertation’s problem statements, followed by research objectives, block
diagram of the overall research methodology, and finally the research overview.
1.1 Definition of the Performance Measurement System (PMS), and role of PMS
in supply chain management.
A performance measurement system (PMS) is defined as a process to measure the
effectiveness of action (Neely et al., 1995). The performance measures and metrics are
essential in the business management because they provide the information that is
necessary for organizations to make decision and take action, especially in a
competitive economy. Parker (2000) identified the purpose of measuring organizational
performance as follows;
(1) to measure the business success;
(2) to determine whether the customer needs are satisfied;
(3) to help the organization to understand its process;
(4) to identify the problems and point out the area for improvement where
necessary, and
Ref. code: 25605422300342CMU
2
(5) to ensure that decisions are made based on facts, not the intuition.
In the supply chain management (SCM), a PMS facilitates the inter-understanding
among supply chain members, and provides the outlook to identify success, as well as
the potential activities (Chan and Qi, 2003). Therefore, the PMS makes a considerable
contribution to the field of SCM in a decision-making process, especially in the business
redesign, and reengineering process. Regarding to the literature, the research topic in
this area is not new. According to Akyuz and Erkan (2010), more than hundreds of
articles on the PMS and metrics were published during 1997-2009. Most of the articles
were discussed about the PMS design, and metric selection. The study from Shepherd
and Gunter (2006) revealed that the PMS in the supply chain system can be categorized
according to the following characteristics.
Table 1.1: PMS in the supply chain system as categorized by Akyuz and Erkan (2010).
1. Balanced Scorecard perspective Kaplan and Norton (1997) 2. Component of measures Beamon (1999), Gunasekaran et al., (2001),
De Toni and Tonchia (2001), Chan (2003), and Chan and Qi (2003)
3. Decision levels Gunasekaran et al., (2001) 4. Supply chain process Chan and Qi (2003), Huang et al., (2005), Li
et al., (2005), Lockamy and McCormack
(2004)
Firstly, Kaplan and Norton (1997) proposed a Balanced Scorecard (BSC)
performance system which is built upon 4 perspectives of financial, internal business
process, customers’ satisfaction, and the learning and growth. BSC presents the
performance measurement in a balanced framework of the total business performance;
however, prioritization of these different perspectives for a firm is an issue that needs
to be addressed. Secondly, for the component of measures, it means that the PMS is
classified into groups. Beamon (1999) categorizes performance measures in two distinct
groups, namely; a qualitative and quantitative measures. Beamon (1999) identified three
Ref. code: 25605422300342CMU
3
types of measures which is resource, output, and flexibility. This categorization is
followed by some researchers that try to address the issue in the SCM. For example,
Van Landerghem and Persoons (2001) built a causal model related to the use of best
practices to group the performance under four objectives which are; flexibility, reaction
time, quality, and cost. The PMS in this category is also proposed as a cost and non-
cost measures such as; quality, cost, delivery, and flexibility (De Toni and Tonchia,
2001), and input, output, and composite measures (Chan and Qi, 2003). However, this
method of categorization has received a criticism of not being connected with strategy,
lack of a balance approach to integrate cost and non-cost measures, and most
importantly, it losses of the supply chain context. Thirdly, for the decision levels in
SCM system, Gunasekaran et al., (2001) classified a PMS as strategic, tactical, and
operational focuses, so this performance system can support each other to achieve the
overall business objectives, and to assist the company to make a right decision. The
conceptual design of PMS by Gunasekaran et al., (2001) was widely supported by
several researchers to design the strategic tool that can align the supply chain from the
operational level to the firm’s strategy. For example, Lin et al., (2005) studied the
operational issues and develop a mathematical model to optimize performances through
a supply chain redesign. Various techniques such as a deterministic model (Chen et al.,
2005), a stochastic analytical model (Chiang and Monahn, 2005), and a simulation
model (Huang et al., 2005) were developed in order to link supply chain strategy to
objectives, and to the operations. However, despite the popularity of the framework,
some authors analyzed that the approach lacks of a systematic thinking for the SCM, as
the supply chain system must be viewed as the whole processes, and the PMS should
span to cover all business aspects.
Therefore, a renowned framework, the Supply Chain Operations Reference
model (SCOR), was originally developed by the Supply chain council in 1997. It has
been described as the most systematic approach for identifying, evaluating, and
monitoring the supply chain performance (Stephens, 2001). The proposed metrics allow
Ref. code: 25605422300342CMU
4
the company to manage performance on multiple dimensions in a hierarchical structure
which is defined in the reference model. According to the SCOR model, a company’s
supply chain would be represented by 6 meta-level processes of plan, source, make,
deliver, return, and enabler, in which all processes are managed under a SCOR
performance metrics for a supply chain. A major advantage of this model is the creation
of a common and standardized language among supply chain members, hence, it
enables the companies to compare supply chain performance with others. To
demonstrate the applications of SCOR model, there are a number of publications that
published the works using the SCOR model as a reference framework. For example, in
the exploratory work of the SCOR model, Lockamy and McCormack (2004) were the
first author who studied this model by investigating a relationship between supply chain
management planning, and the supply chain performance based on four decision areas
of SCOR model of plan, source, make, and deliver. The result reveals that the important
planning function such as the importance of collaboration, process measure,
integration, and information technology are the enabler for success implementation.
Afterwards, McCormack et al., (2008) integrated the SCOR model with the business
orientation maturity model using the previous study as a reference. The study provides
a comparison between traditional versus innovative performance measurement system
based on a Brazilian company surveyed. The result puts forward a clear support for the
need of new performance measurement methodologies that emphasis the important of
business maturity. In terms of practicality of the SCOR model in industries, Hwang et
al., (2008) performed a case-based study for the Taiwanese TFT-LCD industry. The work
contains a comprehensive set of SCOR model that only emphasizes on a sourcing
process and then perform a stepwise regression analysis to analyze the dependency of
different performance measure. Li, et al., (2011) also adopted the SCOR model in order
to ensure a supply chain quality performance to help companies develop and maintain
a supply chain process according to the quality standards. However, these are only the
brief applications of the SCOR model, where the full literature reviews are presented
in the next chapter. Despite the discussion of various PMS in the supply chain system,
Ref. code: 25605422300342CMU
5
this research found that whichever the PMS are used, the effective performance
evaluation should be practical, comparable to other organizations, and be able to
provide the feedback for performance improvements. These are the starting points of
discussion in this dissertation, where the issues of performance evaluation and
improvement is explained subsequently in the following chapters.
1.2 Research problem statements
Based on the literature survey, most of the articles have proposed the PMS as a metric
design, or a requirement to compose a good performance measurement system. There
are only few papers that addressed the issue of performance evaluation such as the
method or the underlying mechanism for assessment. Neely et al., (1995) actually
defined the term “performance evaluation” as a definition to measure, depending on
how it will be calculated, and where the data is obtained from. Therefore, a simplest
way to determine performance based on Neely’s definition is to choose a preferred
metrics, collect the related information, and perform the assessment straightforwardly.
However, this method has a drawback as it reveals only the performance from the past,
where the future direction cannot be anticipated. A good performance measurement
system and evaluation method, actually, should encourage the improvement rather than
just monitoring. Therefore, the measurement method should also integrate a feedback
mechanism in order to tell the company, or the manager on the improvement areas and
the decision to move on. Moreover, the measurement mechanism should also be able to
adjust overtime, as a company needs changes (Maskell, 1991), and be able to compare
to similar organizations where the same performance criteria is applied. With this
reason, and based on the evidence from the literature survey that only the few reports
are focusing on this issue, it becomes the interest of this dissertation to pay attention on
the topic of performance evaluation and improvement in the supply chain system.
Firstly, this thesis begin with the discussion and selection of the standard supply chain
framework that is used throughout dissertation.
Ref. code: 25605422300342CMU
6
The Supply Chain Operations Reference (SCOR) Model is one of a well-established
process reference model which is now supported by the APICS Supply Chain Council
(APICS, 2016). It is organized into five main processes. SCOR Model is comprised of
performance attributes and the measurement metrics in a hierarchical structure. These
organized features allow the framework to be widely adopted by the supply chain
research, and practically adapted to various industries. According to the related
publications that work on the SCOR model, researchers and practitioners agreed that
the SCOR model is a good reference model because;
It provides the standard descriptions of each business process along the supply
chain, which consists of “Plan”, “Source”, “Make”, “Deliver”, “Return”, and
“Enabler”.
The key performance indicators (KPIs) are classified by attributes, which are
dependent on each business process, and lastly
There are the best practices, which can be used as a guideline to achieve good
performances.
With the successful implementation of the SCOR model that appears broadly in the
academic literature, so in this dissertation, the SCOR model is employed as a reference
framework to work on the performance evaluation and improvement system. Although,
the model has provided a definition that is ready to be used which is quick and easy,
and it is possible to assess the values of these KPIs directly based on the business
outcomes as agreed by Neely et al., (1995). The underlying disadvantages are, it lacks
of a procedural methodology, and the obtained KPIs cannot be further analyzed.
Moreover, when the SCOR KPIs are used, the indicators can only help to identify
problems in the current situation, but the logical methodology to manage those KPIs
for further improvement is still unclear. Even though in the traditional method,
managers have relied on experience and intuition to determine how to improve KPIs,
which is a swift decision-making process, this method is still non-systematic and
unexplainable.
Ref. code: 25605422300342CMU
7
From the statement above, even this dissertation is interested in exercising the
SCOR model to address such issue, there are still problems that need to be clearly
examined in order to create a reliable performance evaluation and improvement method
in the supply chain system by using the SCOR model. These points are clarified as the
research problems in this dissertation as follows.
Problem statements
1. Currently, the SCOR model is only the reference model, so when their
performance metrics (SCOR KPIs) are applied, there is no relationships between
the values of the SCOR KPIs and the system parameters under studied. Hence,
it is not possible to predict the consequences of the SCOR KPIs when the system
is changed or improved.
2. There are the agility measures in the SCOR KPIs which are difficult to evaluate.
The agility measures determine flexibility of the system when the upside, or
downsize in demand occurs. Without a procedural methodology and a model,
the evaluation of the agility measures is unclear and non-systematic.
3. The SCOR KPIs compose of many aspects, so when the organization need to
improve the performance in the supply chain, they have wide ranges of direction
and possibilities to be managed without a systematic approach.
4. Since the SCOR KPIs compose of many metrics to be managed, so the
improvement of SCOR KPIs needs to be compromised. The company cannot be
best in all metrics, and there must be some reliable method for the company to
trade off among the improved KPIs that can satisfy the need of the organization.
5. Lastly, there is a complexity of interrelationship between the variables in the
SCOR KPIs and the parameters of the system under studied, so the management
of KPIs for performance improvement needs a detailed methodology to
determine the direction of improvement.
Based on the above research problems, it is an initiative of this dissertation to propose
a model and a procedural methodology to assess these SCOR KPIs, and also the
methodology to manage these KPIs for improvement. According to the supply chain
Ref. code: 25605422300342CMU
8
planning problem, and the work from the literature reviews, this research aims to
address the problems (1) and (2) by contributing the new knowledge to evaluate the
SCOR KPIs by using a predictive MILP model with fuzzy parameters. The complete
literature review of the SCOR model, application of the SCOR model in performance
evaluation are provided in Chapter 2 to support how the SCOR KPIs are defined and
used as a fundamental theory in this research. Followed by the reviews of the MILP
model in the supply chain planning problem, how the uncertainties in the supply chain
system are defined and managed, and why it comes up with the fuzzy set theory (FST)
to endorse the establishment of the MILP with fuzzy parameters that is used to evaluate
the current KPIs, and to predict the future performance in many what-if scenarios.
Apart from the proposed method to assess the supply chain KPIs systematically and
from the obtained outputs in Chapter 3, in order to meet the business objective which
is required by the organization, the procedural method that explain the outputs of the
KPIs must also be able to identify the direction for improvement. This is addressed by
the research questions (3)-(5). Another aim of the dissertation is to propose the new
approach to manage the SCOR KPIs for improvement. The literature review of a Quality
function deployment (QFD), which is considered as a successful tool for systematic
planning of the new product development, is described in Chapter 2 to support the
rationale of why the QFD approach is appropriate to use as a tool to guide the managers
for performance improvement. QFD integrates the customer requirements into every
aspect of the design by outlining the need of the customer, and translate it into the
technical requirement, so that the end products meet the customer expectation. (Liu and
Wang, 2010). The central element of QFD planning process that enables the
transformation of customer requirement to design specification is the House of Quality
(HOQ) matrix that contains the information on “What”, “How”, and the interrelationship
between them to determine the output priority level to fulfil the need of customer (Chen
and Ko, 2011). With the successful implementation of QFD methodology that has
contributed to various fields of study, it is also expected that the QFD methodology is
Ref. code: 25605422300342CMU
9
suitable to determine how to improve the SCOR KPIs. From the statement of discussion
above, This research can briefly summarize the objective of this dissertation as follows,
1. To propose a predictive model and a procedural methodology to assess the
SCOR KPIs with fuzzy parameters.
2. To propose a new methodology to manage the SCOR KPIs by using the fuzzy
QFD approach.
3. To demonstrate the effectiveness of the proposed method by using a case
study of bottled drinking water factory.
Research scope
From the research problems and objectives, the overall contributions of this
research are; to predict the SCOR KPIs by using the predictive MILP model with fuzzy
parameters, and to enhance the SCOR performance by using the QFD methodology.
However, this dissertation scope down the supply chain process under consideration.
For the SCOR model that contains six processes of plan, source, make, deliver, return,
and enabler, this research focus only the Make process and develop the proposed
methods. To be precise, the SCOR KPIs in this dissertation are mostly referred to the
level 2 KPIs of the Make process, the MILP model of the system under studied is
developed based on the manufacturing system of the Make-to-stock production, and
lastly the engineering characteristics of the QFD are proposed based on the
manufacturing parameters, and the production methodology that can be improved in
the factory. The demonstrate the application of the method, the case study of a bottle-
water manufacturing factory is applied. The brief idea of the overall research
methodology is presented in Figure 1.1
Ref. code: 25605422300342CMU
10
Figure. 1.1: Block diagram of the overall methodology.
Step 1: Identify the manufacturing system to be studied. Define the production
parameters, and the variables.
Step 2: Develop the MILP model (predictive model) with the objective function
and constraints to represent the manufacturing system
Step 3: After MILP model validation, apply uncertainties into the MILP model
using a TFNs, and solve the model for optimal outputs under uncertainties.
Step 4: Evaluate the SCOR KPIs, based on the outputs from step 3 and the
proposed methodology in Chapter 3.
Step 5: The current SCOR KPIs, and the level 2 SCOR-Make metrics are defined
as the requirement “WHAT” in the QFD matrix.
Step 6: Apply the proposed method of using the fuzzy QFD approach to manage
the SCOR KPIs for performance improvement in Chapter 4.
Step 7: Determine a technical improvement from the set of priorities “HOW”, and
define the performance improvement actions to be implemented.
Step 8: Implement the selected “HOW” action in the predictive model proposed in
Chapter 3 to predict the new SCOR KPIs after performance improvement.
Stop
Are the new SCOR KPIs satisfied ? No
Yes
Ref. code: 25605422300342CMU
11
Firstly, the manufacturing system to be studied is identified. It can be the system of
interest, or the current manufacturing system in the factory. Next, the operations of the
manufacturing system are represented by the formulation of the MILP model. The
objective of the MILP model is to simulate the system that the research are working
with, and to use as a foundation to assess the SCOR KPIs based on the proposed
methodology. Then, to model the system in a realistic way, uncertainties are included
into the MILP model. This research concern about uncertainties in the supply chain
system because the uncertainties are actually unavoidable in the production system, and
it serves as one of the main factor that disturbs the effectiveness of the operations. In
this dissertation, a triangular fuzzy numbers (TFNs) are used to characterize
uncertainties by the fuzzy parameters in the MILP model. A predictive model is used
because the dissertation wants to establish the relationships between the values of the
SCOR KPIs and the manufacturing parameters. The proposed methodology to evaluate
the SCOR KPIs in Chapter 3 relates with the formulation of the mathematical formula,
interpret the definition of the SCOR KPIs, and turn those KPIs into the measurable
equations with some algorithms to assess the agility measures. The list of the SCOR
KPIs are defined in QFD matrix as “WHAT”, or performance requirement of the
company, where the technical improvement actions (TIs) or “HOW” in the QFD matrix
is related with the production parameters that can be improved. The fuzzy QFD
approach to manage the SCOR KPIs for improvement is then presented in Chapter 4,
where a set of priorities “HOW” are the outputs of the proposed methodology. The
company can predict the new SCOR KPIs by implementing such TIs in the predictive
model, and use the methodology in Chapter 3 to anticipate the company’s performance
after improvement. If the organization, or the manager is not satisfying with the
improved KPIs, then they can re-apply the QFD methodology in step 6 to evaluate the
new relative importance and performance settings, and follow the approach until the
new SCOR KPIs are satisfied before implementing the real actions in the factory. The
detail of each step is presented further in subsequent chapters.
Ref. code: 25605422300342CMU
12
1.3 Overview of this dissertation
This dissertation is organized into five chapters. The first chapter deals with
dissertation introduction, the problem statements, research objectives, research scope,
and overview of the dissertation. Chapter 2 presents a literature review which is divided
into subtopics that include the foundation of SCOR model, the previous works of using
the SCOR model in performance evaluation, the introduction to the MILP model, and
its application in supply chain planning. Next is the uncertainty in the supply chain
system, fundamentals of fuzzy set theory (FST), application of the FST in the research
and their use in the MILP model. The last topic on the literature review chapter is the
fundamental concepts of Quality Deployment Function (QFD) philosophy, and the
presentation of the fuzzy QFD that is successfully implemented in other research fields.
Chapter 3 proposes the methodology to evaluate the SCOR KPIs by using a predictive
MILP model with fuzzy parameters. The MILP model, the MILP model with fuzzy
parameters to handle uncertainties in the manufacturing system, and the methodology
to assess the SCOR KPIs based on the SCOR-Make process is presented. The proposed
methodology is demonstrated by the case study, where the obtained results are
discussed at the end of this chapter. Based on the procedural approach to predict the
SCOR KPIs depicted in last chapter, chapter 4 proposes the fuzzy QFD approach to
manage the SCOR performance indicators for improvement. The eight-step fuzzy QFD
approach is discussed, whereby the similar case study is used to prove efficiency of the
proposed method. The results are again discussed at the end of the chapter. Finally, the
conclusion of the dissertation, theoretical and practical contribution, limitations, and
recommendation for further studies are presented.
Ref. code: 25605422300342CMU
13
Chapter 2
Literature Review
The review of the literature that is related to this dissertation is presented in this chapter.
The review contents are divided into four main topics. Firstly, the review will introduce
the Supply Chain Operations Reference (SCOR) Model, in which the basic definitions,
processes, and metrics are presented. Then, it is followed by applications of the SCOR
model in the performance evaluation field to provide the guideline on how the model is
applied. Secondly, the Mixed Integer Linear Programming (MILP) model and its
application on the supply chain planning are described to justify the use of MILP model.
Thirdly, uncertainties in the supply chain system, and the fuzzy set theory (FST) are
discussed to describe how uncertainties work, and why this research includes the
uncertainties in the proposed method. Finally, the Quality Deployment Function (QFD)
philosophy is presented, its successful implementation to transform the customer needs
into requirements to meet customer expectation is described to support the proposed
methodology that applies the QFD with the SCOR model to manage the SCOR KPIs
for the performance improvement.
2.1 A Supply Chain Operations Reference (SCOR) model
A SCOR model is the standard supply chain framework that is originally
proposed by the Supply Chain Council (SCC), a non-profit professional forum founded
in 1996, where many industry experts are gathered to discuss on the emerging issues of
the supply chain management, to consolidate the methodology and analytical
techniques, and to identify the benchmarking standards to help the organizations to
make a dramatic improvement in their supply chain processes. Generally speaking, the
SCOR model is a process reference model that provides a unified framework to manage
a supply chain under the same standard and format that can be applied to any product
and service, in any industry to communicate the supply chain problems under the same
definition. From the model introduction in 1996, the SCOR model has undergone about
thirteenth revision, and its current version is 11. The SCOR model is now supported by
Ref. code: 25605422300342CMU
14
the APICS Supply Chain Council (APICS, 2016). The latest version of the model
consists of four parts, namely performances, processes, practices, and people.
1. Performance composes of the standard metrics to describe the overall performance,
and the specific performance under each process in order to define the strategic goals.
2. Processes consist of the standard descriptions of the supply chain management
processes and the relationship between them.
3. Practices refer to the management practices as a guideline to the SCOR users to
achieve the significant improvement in the business processes.
4. People part contains the standard definitions for skills that the workforce is required
to perform supply chain processes.
The successful implementation of the SCOR model has provided the evidence
that help the company to eliminate the wasteful practices along the supply chain and to
establish a standard terminology to communicate within and across the organizations,
which result in the improvement of the overall processes. The details of the SCOR
applications, especially in the performance management field, will be presented later in
this chapter. (Bolstorff et al, 2003, Harelstad et al, 2004, Kevan, 2005, and Kocaoğlu et
al, 2011).
2.1.1 SCOR processes.
The SCOR model is developed to describe the business activities with all phases
in the supply chain system in order to satisfy customer’s demand. The model is
organized based on six primary management processes of Plan, Source, Make, Deliver,
Return, and Enable as shown in figure 2.1
Ref. code: 25605422300342CMU
15
Figure 2.1: The SCOR model with six management processes (APICS, 2016)
The definition of the SCOR processes according to the SCOR model is described as
follows.
1. Plan: The plan process is to coordinate the supply and resources in supply chain
system in order to meet demand. It includes determining the requirements and
identifying the actions that have been drawn in order to achieve the supply chain
objectives.
2. Source: It is the process related with ordering, delivering, receiving, and transferring
raw materials to produce products and services.
3. Make: It is the process that adds values to the product such as scheduling and
manufacturing in order to transform raw materials to finished products. The make
process can include mixing, separating, forming, and machining activities. The
Make process is the main focus in this dissertation.
4. Deliver: It is the process to perform customer-facing order management and order
fulfilment by means of transportation and distribution of orders to end customers
5. Return: It is associating with the returning of disapproved products, parts,
components from customers back to suppliers to address defects in product or to
perform the maintenance activities.
6. Enable: It is the process that facilitates the management of business rules,
performance, and regulatory requirements to meet the company needs. The enable
process mostly interacts with other department in the organization, such as finance,
Ref. code: 25605422300342CMU
16
HR, IT, and facility management to support the governance of the planning and
execution in the supply chain.
Figure 2.2: A hierarchical structure of SCOR
As shown in Fig.2.2., the model is designed to support the supply chain analysis
in hierarchical levels. The Supply chain council has focused on the top three process
levels. The first level deals with the six process types of the supply chain. The second
level is the process category. For example, the Make process consists of three process
categories of make-to-stock, make-to-order, and engineer-to-order. And lastly, the third
defines the configuration of the individual process that is capable to execute.
2.1.2 SCOR metrics
The performance section of the SCOR model consists of two types of elements,
namely, performance attributes and metrics. The performance attribute is used to
express a strategy and it cannot be measured. The metrics measure the ability of a supply
chain to achieve the strategic attributes. For example, the superior performance for
reliability is expressed by a performance objective of perfect order fulfillment. The
SCOR model consists of 10 performance metrics that are grouped into 5 performance
attributes. The supply chain council suggests that scorecards should contain at least one
Ref. code: 25605422300342CMU
17
metric for each performance attribute to ensure a balanced decision-making process
(Lima-Junior and Carpinetti, 2016). The clarification of these metrics and their causal
relationship make the SCOR metric capable to analyze the performance of a supply
chain for different perspectives. Table 2.1 exhibits the SCOR performance attributes
according to the SCOR definition, where Table 2.2 explain their level 1- strategic
metrics.
Table 2.1: The SCOR performance attributes
Performance Attribute Definition
Reliability The ability to perform tasks as expected. Reliability focuses on the
predictability of the outcome of a process. Typical metrics for the
reliability attribute include: On-time, the right quantity, the right
quality.
Responsiveness The speed at which tasks are performed. The speed at which a supply
chain provides products to the customer. Examples include cycle-time
metrics.
Agility The ability to respond to external influences, the ability to respond to
marketplace changes to gain or maintain competitive advantage. SCOR
Agility metrics include Flexibility and Adaptability
Costs The cost of operating the supply chain processes. This includes labor
costs, material costs, management and transportation costs. A typical
cost metric is Cost of Goods Sold.
Asset Management
Efficiency (Assets)
The ability to efficiently utilize assets. Asset management strategies in
a supply chain include inventory reduction and in-sourcing vs.
outsourcing. Metrics include: Inventory days of supply and capacity
utilization.
Ref. code: 25605422300342CMU
18
The first three attributes are regarded as customer-based, whereby costs and
asset management are considered as the organization-based. Each attribute consists of
one or more level-1 measurable metrics, as shown in Fig.2.2., where the organization
can measure how successful their business is achieved when comparing to the market
space.
Table 2.2 The SCOR Level-1 metics
Performance Attribute Level-1 Strategic Metrics
Reliability Perfect Order Fulfillment (RL.1.1)
Responsiveness Order Fulfillment Cycle Time (RS.1.1)
Agility Upside Supply Chain Flexibility (AG.1.1)
Upside Supply Chain Adaptability (AG.1.2)
Downside Supply Chain Adaptability (AG.1.3)
Overall Value at Risk (AG.1.4)
Costs Total Cost to Serve (CO.1.001)
Asset Management
Efficiency (Assets)
Cash-to-Cash Cycle Time (AM.1.1)
Return on Supply Chain Fixed Assets (AM.1.2)
Return on Supply Chain working capital (AM.1.2)
Organizations that use the SCOR performance metrics can compare their performance
levels against others by using a benchmarking tool which is called SCORmark. The
database of SCORmark contains historical data of over 1,000 companies and 2,000
supply chain systems (APICS, 2016). The benchmarking process using the SCORmark
consists of 5 steps which are (1) to define the supply chains to be compared, (2) to
measure the internal and external performances, (3) to compare the performance to the
relevant industry, (4) to establish the competitive requirements, and (5) to calculate the
opportunity value of improvement. For ease of comparison, the SCORmark categorizes
Ref. code: 25605422300342CMU
19
the process performances according to three positions as follows (Ganga and Carpinetti,
2011).
1. Superior position, indicates the 90th percentile of companies in the database.
2. Advantage position, is the performance level halfway between Parity and
Superior (i.e., at 70th percentile).
3. Parity position, indicates the 50th percentile of performance in the SCORmark
database.
The SCOR model suggests the development of new tool to combine with the
SCOR metrics such as simulation modeling in order to support the management
activities in the supply chain performance measurement, risk assessment, and supplier
evaluation (Agami et al, 2014). However, the SCOR model and applications in this
dissertation are only focused on the supply chain performance measurement aspect. The
scope of dissertation is narrowed down to focus only at the Make process, on the SCOR
KPIs issue of how the Level-1 strategic metrics can be evaluated based on a predictive
approach, and how these KPIs can be further improved to achieve the strategic direction
that is preferred by the company. Since the level-2 metrics serve as a diagnostic tool for
level-1 metric, so we are extending the measurement metric to explore the level-2 metric
of the MAKE process on the method of evaluation. In the next section, the previous
works of SCOR model that is related to the performance measurement issues is
discussed.
2.2 The SCOR model in performance evaluation
The literature review is conducted based on a number of SCOR assessment
criteria to review the selected SCOR model application papers that have been published
around 2004-2017. Stephens (2001) was the first author that presents the first SCOR
publication that describes its development and applications. Since then, the application
of the SCOR model has been reported in several industries. For example, in the lamp
industry (Vanany et al, 2005), the ethanol and petroleum industry (Russel et al, 2009),
geographic information systems (Schmitz, 2008), in service industry (Ellram et al.,
Ref. code: 25605422300342CMU
20
2004), IT and technology consulting (Dong et al., 2006), transistor-LCD industry (Hwang
et al., 2008), the construction industry (Cheng et al., 2010, and Pan et al., 2010), the
automotive industry (Potthast et al., 2010), and in the shipbuilding industry
(Zangoueinezhad et al., 2011). The model is also connecting to many research
methodologies to broaden their applications. By integrating the model with Six Sigma,
it provides a usable strategic toolset for lean management (Malin and Reichardt, 2005).
By combining the SCOR model with AHP, it constructs a framework to evaluate the
performance of the system in the prioritization purposes (Bhagwat and Sharma, 2009,
Elgazzar et al., 2012, Charkha and Jaju 2014, Mendoza, 2014, and Alomar and Pasek.,
2014). Fuzzy theory is combined with the SCOR model to address the issues of
uncertainty (Chan and Qi, 2003, and Lima-Junior and Carpinetti,2016). Discrete event
simulation is introduced to the SCOR model to create a template to use as a decision
support tool (Person, 2003 and Dong et al., 2006). And lastly, the case studies are applied
to the SCOR model to investigate the problems in the particular area such as in
environmental considerations (Bai and Sarkis, 2010, and Xiao et al., 2012), delivery
processes (Soffer and Wand, 2007), inventory management (Gumus et al., 2010), and the
footwear industry (Sellitto et al., 2015). To summarize the applications of the SCOR
model particularly in the performance measurement system, the characteristics of the
study are divided into five subgroups based on the methodological approaches that were
applied by the authors in various performance measurement topics.
2.2.1 Application of the SCOR model by using system simulations.
The authors have embedded the SCOR model with System Dynamics, Discrete
Event Simulation, and Hybrid Simulation techniques to meet their objectives.
Simulation is regarded as a well-known technique to analyze a complex and dynamic
systems. The objective of integrating simulation to the SCOR model is to create a
reusable template that can examine the supply chain model in various configuration
with the what-if scearios (Ellram et al.,2004, and Dong et al., 2006). Persson and Araldi
Ref. code: 25605422300342CMU
21
(2009) focused on the Plan process of SCOR model and work with the enterprise
simulator to understand the impact of e-solutions in terms of the standard operations
and the financial measures. Guruprasad and Herrmann (2006) attempted to improve the
simulation elements that represent the activities in a supply chain by using the Arena
software to create a standardized the supply chain model that can be generally used.
Roder and Tibken (2006) used the Matlab software to create a modular modelling based
on the SCOR agility for the intra- and inter- company process chains in an automotive
industry. Gulledge and Chavusholu (2008) applied Oracle software to automate the
SCOR model as an enabler for process-oriented business intelligence. They prove that
automated KPIs determination is feasible but difficult if the data collection to support
the KPIs is not automated. The simulation of a specific level of operation in a supply
chain system using the SCOR model reveals that all companies can share the same set
of processes, as the ARENA discrete event simulation is used as a tool to understand
their static operations ( Persson and Araldi, 2009). Pan et al., (2010) integrated the model
with dynamic simulation to create a hybrid model that enable the supply chain
participants to define their roles, facilitate communication, and help the management
team to identify the bottleneck. Referring from above SCOR simulation based papers,
it is concluded that the integration of system dynamics and discrete event simulation
provides the effective modeling technique to enhance the overall value chain. The
results based on the papers also indicate the achievement of profits, customer
satisfactions, and supply chain responsiveness.
2.2.2 Application of the SCOR metrics to other decision support models and
methodologies.
The application of SCOR performance metrics is broadened as the multi-criteria
decision making framework that links itself to other operation research methodology.
The characteristics of this type of research is to integrate a decision support tool to
SCOR model such as a fuzzy logic, optimization modeling, and other heuristic
algorithms in order to find the best solution to the problem. For example, Thakkar et al.,
Ref. code: 25605422300342CMU
22
(2009) combined the SCOR model with the Balanced Scorecard to develop an integrated
performance framework for the small and medium enterprise. Reyes and Giachetti
(2010) used the Delphi method to develop a supply chain maturity model that help firms
to evaluate their operations and create the improvement road-maps. Li et al., (2011)
proposed an analytical method using the structural equation modelling (SEM) to extend
the return process and integrate the ISO 9000 series into the process. The authors
observed that reliability and responsiveness are the important quality indicators for the
return process. Fuzzy logic was also applied to the SCOR model to predict the
performance based on a causal relationship between levels 1 and 2 by using a
SCORmark as a reference (Ganga and Carpinetti, 2011). Optimization was also used by
Xiao et al., (2012) to model a closed-loop logistics model based on the profit function
and incorporate the selected SCOR metrics to measure the system’s performance in
optimizing the multi-echelon inventory model by using simulation. Researchers have
developed the complex methodologies such as a rough-set theory, and grey-based
neighborhood rough-set theory (Bai et al., 2012), MACBETH ๖Cliville and Barrah,
2012), fuzzy Choquet intergral operator approach (Ashayeri et al., 2012), and other
multi-criteria decision analysis method to combine with SCOR model to identify and
select proper set of performance indicators that increase value within the chain.
Researchers suggest that performance metrics of SCOR and processes should be
enforced as a common language in a company, so that all metrics are available and can
be benchmarked for improvement to other organizations that works with SCOR. The
example can be seen by the attempts of Giannakis (2011) to explore the utility of SCOR
model, which is claimed to be manufacturing-biased, in a service sector supply chain in
order to create a service-oriented referencing model.
Ref. code: 25605422300342CMU
23
2.2.3 SCOR model that decompose a problem into a hierarchical structure using
Analytical Hierarchy Programming (AHP)
AHP is a decision-making tool that decomposes a problem of operation into a
leveled structure. AHP reflects a thought of human mind that systematically sort the
element of problems according to level of priority and eventually group the importance
of problems into a different level. AHP breaks down a complex problem into multi-level
that consists of objectives, criteria, sub-criteria, and construct the objective function to
solve optimization problems. These works are studied by Wang et al., (2004), Rabelo et
al., (2007), and Han and Chu (2009) that incorporate qualitative factors to guide decision
making using AHP after simulation results. Rabelo et al., (2007) integrated the AHP
technique and SCOR model based on sourcing process to demonstrate how the
manufacturing facilities handle the returned defective product, and the maintenance
repair and operations for sold products from all local warehouses by using discrete
event simulation. Kocaoglu et al., (2010) used the integrated AHP-TOPSIS-SCOR
approach for measuring a benchmarkable supply chain performance. Palma- Mendoza
(2014) applied the SCOR model for the identification of the supply chain process, and
used the AHP as a tool to support the supply chain redesign. Elgazzar et al., (2012)
studied the dempster-AHP model to develop a performance measurement method that
links the supply chain processes to a company’s financial strategy. Wang et al., (2004)
adopted SCOR level 1 metrics as criteria in product-supplier selection using AHP and
goal programing optimization. AHP is a popular technique to deal with the complex
situations specially to give weight to a qualitative decision-making factor without
acquiring the advance mathematical knowledge, and a decision maker can exploit their
experience and expertise as a part of priority in assigning weight so they do not require
a complete information regarding to all aspects to solve particular problems.
Ref. code: 25605422300342CMU
24
2.2.4 Case studies using SCOR model.
Case studies applied the SCOR model to investigate problems on a specific
decision area. For example, Burgess and Singh (2006) employed a case study to a steel
product manufacturing and distribution to develop a framework for analyzing how
social and political factors impact the performance of a supply chain. Yilmaz and Bititci
(2006) applied the value chain method to compare the performance measurement of
manufacturing and tourism industries. Theeranuphattana and Tang (2008) combined the
strength of SCOR model and Chan and Qi’s measurement algorithm to develop an
empirical measurement to resolve the problem in the cement manufacturing. Hwang et
al., (2008) improved the performance of the sourcing process in a Taiwanese’s TFT-LCD
industry. Two years later, he continued to use the SCOR model and applied the
Structural Equation Modelling (SEM) to determine the relationship of green purchasing
behavior among the green label products (Hwang. et al., 2010). Schnetzler et al., (2009)
applied the SCOR model to the forestry industry to describe the second level of wood
supply chain, and map the forest wood to improve its intra-organizational logistic
processes. The case studies by using the SCOR model to solve the supply chain
problems are still appeared continuously in many industries such as in the after-sales
services (Cavalieri et al., 2007), construction industry (Cheng et al., 2010), furniture and
rubber industry (Banomyong and Supatn, 2011). Moreover, Soffer and Wand (2007)
focused on single decision area of SCOR delivering process for the make-to-stock
manufacturing environment and analyzed the key performance indicators that are
important in delivering activities. Jalalvand et al., (2011) proposed the method to
compare supply chain of an industry for benchmarking and supply chain ranking
purposes by using SCOR model as main business stages and employ Data Envelopment
Analysis (DEA) and Promethee II as a tool to compare supply chain in Iranian broiler
industry. The review in this section concludes the fact that the SCOR model is practical
and can be adapted in various contexts.
Ref. code: 25605422300342CMU
25
2.2.5 The relationship of SCOR model to other external factors.
Due to an expansion of supply chain management to the inter-organization
network, a number of research works try to explore the relationship between SCOR
models to other management issues in a broader context. This kind of research usually
involves inferential statistics to determine a connection between factors. The empirical
research that examines the causal relationship between SCOR model and other
organizational issues in a global context are also emerged in recent years. Lockamy and
McCormack (2004) studied the exploratory work by linking SCOR metrics to a supply
chain planning and conclude that the planning process is important to overall SCOR
processes. McCormack et al., (2008) investigated the relationship between supply chain
maturity and performances in many industrial sectors such as manufacturing,
construction, retail, and communication. Li et al., (2010) investigated the relationship
between quality assurance standard and SCOR performances in order to help companies
to maintain supply chain processes that meet certain performance metrics. Wang et al.,
(2010) suggested the possibility of aligning SCOR with Business Process Reengineering
to enhance the multi-national enterprise resource planning and proposed the study to
deal with supply chain performance in overall expression of the whole supply chain.
Röder and Tibken (2006) proposed a methodology for process optimization based on
SCOR model that integrates product and process documentation between enterprises in
order to evaluate the benefits of inter-company supply chain using simulation-based
decision support system.
It can be seen from the literature review that the SCOR model has provided as
a systematic framework to manage the supply chain under the standard processes, and
can combine with many operation research tools to improve the business in many
industries. Applications of the SCOR model in the performance measurement system
also indicate that the model is feasible to determine and compare a performance in the
organizations and against others. With this reason, this dissertation aims to address the
SCOR model as a reference framework, and combine with some models that can help
Ref. code: 25605422300342CMU
26
organizations to establish the relationship between the operations and the SCOR KPIs,
as well as including the model with some diagnostics tool to manage the SCOR KPIs
for improvement. From the next section, the model and tools that are used to manage
the SCOR KPIs will be introduced.
2.3 The MILP model and its applications
This section presents an introduction to the field of mixed integer linear
programming (MILP), the basic notions, the formulation processes, and the MILP
model that have appeared in the supply chain planning problems.
2.3.1 Fundamental of the mixed integer linear programming (MILP) model
The mixed integer linear programming (MILP) is a technique for optimizing the
decision that take place in a complex system in various research fields such as chemical
engineering, biology, medicine, transportation, telecommunications, sports, and
national security (Papoulias and Grossmann, 1983, Floudas and Anastasiadis, 1988, and
Shah and Pantelides, 1991). The mixed integer linear programming (MILP) model for
the supply chain production planning is originally proposed by McDonald and Karimi
(1997). The aim of this tactical model is to optimally allocate limited resources of a
company to satisfy the market demands at a minimum cost. The MILP model uses the
basis of the classical linear programming model that includes a set of variables, which
represent actions that can be taken in the system being modelled. When we optimize a
function of these variables which is to find the minimum or maximum values of the
objective function, the mechanism of the MILP maps each possible sets of decisions
that satisfy with a set of constraints. Supposing that x1,…,xn is denoted as a set of decision
variables, the general form of the linear model is expressed as follows. (Klir and Yuan,
1995).
Minimize or maximize nxncxcxc ...2211 (1a)
Subject to nxnaxaxa 1...212111 1or , b (1b)
nxnaxaxa 2...222121 2or , b (1c)
Ref. code: 25605422300342CMU
27
…
nxmnaxmaxma ...2211 mb or , (1d)
.,...,1,0 nx jj
In the context of linear programing (LP) and the MILP problems, the function that
assess the quality of the solution, or the “objective function” (Eq. 1a), is a linear function
of decision variables. An LP will either minimize or maximize the value of the objective
function. The decisions that must be made are restricted under a set of “constraints” in
the model (Eq. 1b-1d). Each constraint requires that a linear function of decision
variables is either equal to, not less than, and not more than a particular scalar value,
and each decision variable must be non-negative. The value cj , .,...,1 nj is referred as
objective coefficients, which are often associating with the costs in the minimization
problems. The values b1,…, bm are the right-hand-side values of the constraints, which is
mostly related to the amounts of the available resources for constraints, or
requirements for constraints. The aij denotes how much of resources or requirement i
is consumed or satisfied by decision j. The problem in the above form is called the
“linear program” because the objective function and constraints are all linear. A mixed
integer LP program, is a linear program with the added restriction that some of the
variables must be the integer values. The exposure to the integer linear programming
with respect to more approaches can be referred to the books of Schrijver (1986),
Nemhauser and Wolsey (1988), and Parker and Rardin (1988).
Modelling the MILP problems usually involve three steps. The first step is to define
a set of decision variables that represent the choices to be optimized. The second step is
to construct a statement of constraints, and the last step requires the statement of the
objective function. Then, to efficiently solve the MILP problem, it requires an
understanding of how the MILP solvers work. The MILP solver such as Lingo, CPLEX,
and Solver use a combination of branch-and-bound and cutting plane techniques ( Land
and Doig, 1960, Beale and Forrest, 1976, Crowder et al., 1983, and Van Roy and
Ref. code: 25605422300342CMU
28
Wolsey, 1986) while a tutorial of how these techniques work is not stated in this
dissertation. A solution that satisfies all constraints is called a feasible solution. The
feasible solutions that achieve the best objective function value is called optimal
solutions. If the solution for the MILP model does not exist, the MILP model itself is
called infeasible. However, some feasible MILP models have no optimal solution
because it achieves the unbounded objective function values with the feasible solutions.
Such problems are called unbounded. The numerical example of how the MILP model
is formulated and solved is not illustrated in this dissertation, instead the application of
the MILP model in the supply chain production planning is reviewed in the next section.
2.3.2 Application of the MILP model for supply chain production planning
A wide range of applications can be modeled as the MILP problems. These
applications have attracted a lot of attention in the field of operations research such as
in the allocation problems, in facility planning, in scheduling problems, and in network
transportation problems. The literature review in this section put forward the LP and
the MILP problems that have been worked by different authors.
The popularity of the linear programing model is originally proposed in the field of
production and distribution planning (Martin et al., 1993, Chen and Wang, 1997, Ryu
et al., 2004, and Kanyalkar and Adil, 2005), and it was applied in several industries such
as in glass and steel factories. The LP is also capable of modeling the cross-organization
planning in a multi-plant, multi-period, and multi-product environment (Oh and Karimi,
2006, Peidro et al., 2010, and Diabat and Theodorou 2015)), to work with the tax and
financial data that related with the firm’s business activity, to use in the centralized and
decentralized production planning process (Jung et al., 2008, and Amirtaheri et al., 2017
), and to adapt in the supplier selection problems (Shaw et al., 2012, and Shakourloo et
al., 2016 ). With the development of the LP model, several authors continue to use the
MILP technique to better model the supply chain management. A lot of authors
employed the MILP model to design the production planning, inventory planning,
Ref. code: 25605422300342CMU
29
national transport planning, sales planning, and product-marketing match under
different aspects (Mcdonald and Karimi, 1997, Dogan and Goetschalckx, 1999, Timpe
and Kallrath, 2000, Jayaraman and Pirkul, 2001, Kabak and Ulengin, 2011, Sazvar et
al., 2014, Syam and Bhatnagar, 2015, and Mogale et al., 2017 ) . For example,
Goetschalckx et al., (2002) presented the extension of the model with the seasonal
demand management. Jang et al., (2002) elaborated the MILP model to include four
modules for supply chain management, consisting of the supply chain design,
production-distribution planning, the management module, and the data processing
module. Wu (2010) used the LP model to examine the production loading problem that
involve the import quota limits. As a result, they are capable to plan several supply tiers
in relation to the list of materials, available resources, and transportation capability. And
just recently, the supply chain environmental management is the issue to concern where
the LP model is used to optimally allocate the resource (Sazvar et al., 2014, and
Ameknassi et al., 2016) ; such as the inventory replenishment policy that balance
between financial and various greenhouse gas emission (Sazvar et al., 2014). The
predictive MILP model was also developed by Perea-lopez et al., (2003) for the supply
chain dynamic characterization, and for the economic inventory control (Subramanian
et al., 2014). The MILP model is also capable to solved and illustrated by several
techniques apart from the optimization; such as Lagrangian and heuristic relaxation
techniques (Barbarosoglu and Ozgur, 1999) , genetic algorithms and fuzzy techniques
(Gen and Syarif, 2005), and Lagrangian decomposition (Eksioglu et al., 2006). The multi-
stage supply chain problems are also represented by the MILP model (Dhaenens-flipo
and Finke, 2001, Bredstrom and Ronnqvist, 2002, and Park, 2005. For example, Bilgen
and Ozkarahan (2007) considered a model that integrate the mixed load that transport
between different sea ports in a cereal industry, Meijboom and Obel (2007) studied the
coordination between different stages of a mid-range supply chain planning. Lastly,
Rizk et al., (2008) suggested the MILP model for a single product and several
distribution centers.
Ref. code: 25605422300342CMU
30
Based on the selected literature review, Table 2.3 summarizes the modeling
approach of the reviewed works, purpose of modeling, and shared process information
that contained in the objective of using the MILP model in the supply chain planning.
From tables 2.3 and 2.4, we can conclude that the majority of the supply chain planning
problem are using the MILP model to minimize cost, which most of the information
concern about the production cost, transport cost, inventory cost, production capacity,
and demand planning.
Table 2.3: The modeling approach and purpose of the model
Authors Modeling approach Purpose of the model
Linear
Programming
(LP)
Integer
Programming
Model (ILP)
Minimize
Cost
Maximize
Benefits
Chen and Wang (1997) x x
Timpe and Kallrath (2000) x x
Dhaenens-flipo and Finke
(2001) x x
Bredstrom and Ronnqvist
(2002) x x
Jayaraman and Pirkul
(2001) x x
Jang et al., (2002) x x
Ryu et al., (2004) x x
Perea-lopez et al., (2003) x x
Kanyalkar and Adil, (2005) x x
Gen and Syarif, (2005) x x
Park, (2005) x x
Oh and Karimi, (2006) x x
Eksioglu et al., (2006) x x
Bilgen and Ozkarahan (2007
x x
Meijboom and Obel (2007) x x
Jung et al., (2008) x x
Rizk et al., (2008) x x
Torabi and Hassini (2008) x x
Peidro et al., (2010) x x
Wu (2010) x x
Ref. code: 25605422300342CMU
31
Kabak and Ulengin, (2011) x x
Shaw et al., (2012) x x
Sazvar et al., (2014) x x
Subramanian et al., (2014) x x
Syam and Bhatnagar
(2015) x x
Diabat and Theodorou
(2015) x x
Ameknassi et al., (2016) x x
Shakourloo et al., (2016) x x x
Mogale et al., (2017) x x
Amirtaheri et al., (2017) x x
Ref. code: 25605422300342CMU
32
Table 2.4: The shared information process contained in the supply chain planning model
Authors Shared process information
Production
Cost
Transpor
t
Cost
Lead
time
Set Up
Cost
Replenishment
cost
Inventory
Level
Inventory
Cost
Back order
Cost
Production
Capacity
Demand
Planning
Chen and Wang (1997) x x x x x
Timpe and Kallrath (2000) x x x x x x x x
Dhaenens-flipo and Finke
(2001) x x x x x x
Bredstrom and Ronnqvist
(2002) x x x x x x
Jayaraman and Pirkul
(2001) x x x x x
Jang et al., (2002) x x x x x x
Ryu et al., (2004) x x x x x
Perea-lopez et al., (2003) x x x x x x x x
Kanyalkar and Adil, (2005) x x x x x
Gen and Syarif, (2005) x x x x x
Park, (2005) x x x x x x
Oh and Karimi, (2006) x x x x x
Eksioglu et al., (2006) x x x x x x
Bilgen and Ozkarahan (2007)
x x x x x
Meijboom and Obel (2007) x x x
Jung et al., (2008) x x x x x
Rizk et al., (2008) x x x x x x
Torabi and Hassini (2008) x x x x x x
Peidro et al., (2010) x x x x x
Wu (2010) x x x x
Kabak and Ulengin, (2011) x x x
Ref. code: 25605422300342CMU
33
Authors Shared process information
Production
Cost
Transpor
t
Cost
Lead
time
Set Up
Cost
Replenishment
cost
Inventory
Level
Inventory
Cost
Back order
Cost
Production
Capacity
Demand
Planning
Shaw et al., (2012) x x x x
Sazvar et al., (2014) x x x x x
Subramanian et al., (2014) x x x x x
Syam and Bhatnagar
(2015) x x x
Diabat and Theodorou
(2015) x x x x
Ameknassi et al., (2016) x x x
Shakourloo et al., (2016) x x x x
Mogale et al., (2017) x x x x
Amirtaheri et al., (2017) x x x x x x
Ref. code: 25605422300342CMU
34
Based on the implication of the MILP models that were appeared in several academic
papers, the conclusions drawn affirm that;
1. the MILP model is appropriate for the supply chain production, and transport
planning in the tactical decision level.
2. The optimization technique stand out as the suitable approach to solve the
MILP model.
3. The majority of the models were proposed to minimize the total supply chain
costs, following by to maximize the supply chain profit.
4. The demand, the production costs, the transportation, the inventory, and the
production capacities are considered in terms of the problem constraints or
limitation of resources, and lastly
5. Most of the problems were supported by numerical case studies.
With the characteristics of the problem that need to be modelled in this dissertation, it
is the motivation of this dissertation to use the MILP model to represent the production
system, and to use as a predictive model for the supply chain performance evaluation
and improvement purpose. In the next section, uncertainty in the production system and
the methodology to handle the uncertainty is discussed.
2.4 Uncertainty in the supply chain system and fuzzy set theory
In the traditional design of the supply chain planning problem, the issue of
uncertainty has not yet been embraced within the scientific community (Klir and Yuan,
1995). So, in the traditional view of science, uncertainty can be thought of the
information that is incomplete, imprecise, and unreliable that must be avoided. So, when
the MILP model is used, the traditional mathematical modelling usually assumes the
parameters to be deterministic. Until the 20th century, the statistical mechanics were
developed and the issue of uncertainty is reconsidered by using the probability theory.
This type of uncertainty is generally referred as random uncertainty. Particularly in the
manufacturing system, several authors have analyzed the sources of uncertainty
presented in the supply chain system (Davis, 1993, Lee and Billington, 1993,
Ref. code: 25605422300342CMU
35
Childerhouse and Towill, 2000, and Wang and Shu, 2005), and most of the researchers
classified them into three groups which are;
Uncertainty from the supply, which is caused by the faults or delays in the
supplier’s delivery.
Uncertainty from the process and manufacturing activity, which occurs as a
result of unreliable production process
Uncertainty from the demand, which arises from the inaccurate demand
forecasting.
As uncertainties in a supply chain system serve as one of the main factors that can
influence the effectiveness of operations, they lead to an increasing interest to model
the supply chain design by different modelling techniques that are closer to the real
situations. The probability theory dominated the mathematics of uncertainty for more
than five centuries (Lindley, 1987), and the leading publications in the 20th century
quantified the uncertainty using this technique. Recent publications of the supply chain
planning that include uncertainties, and model it according to the probabilistic
distribution and the stochastic modelling included Alonzo-Ayuzo et al., (2003), Gupta
and Maranas (2003), and Guillen et al., (2005). However, with the introduction of the
fuzzy sets by Zadeh in 1965, the expression of uncertainty by using the probability was
challenged when it is proved that the probability theory resulted from a special case of
fuzzy sets (Klir and Wierman, 1996). Also, once the probabilistic distribution is used, it
required a lot of statistical data from the past which is sometime not available (Wang
and Shu, 2005). The Fuzzy set theory (FST), which is introduced by Zadeh (1965), is an
alternative modelling technique which is simpler and less data demanding. The
following section gives the introduction to the fuzzy set theory (FST), and its application
that used in the supply chain research field to model the issue of uncertainty.
Ref. code: 25605422300342CMU
36
2.4.1 Fuzzy set theory (FST)
Fuzzy set theory (FST) was introduced by Zadeh (1965) as a technique to deal with
the imprecise data and uncertainty that cannot be avoided in a practical situation. Fuzzy
sets provide a mathematical way to represent vagueness in the humanistic systems. The
idea proposed by Lotfi Zadeh suggested that a set membership is the key to decision
making when face with uncertainty. The notion of set membership is the central
representation of objects within a universe by sets defined on the universe. While the
classical sets contain the objects that satisfy precise properties of membership, the fuzzy
sets contain objects that satisfy the imprecise properties of membership which can be
approximated. The following example illustrates the definition of crisp set and fuzzy
set. Assume that a set of heights from 5 to 7 feet is precise (crisp), the set of heights in
the region around 6 feet is imprecise or fuzzy, and suppose that a single collection of
individual element x, which make up a universe of information (discourse) X., and
various combinations of these individual elements make up a set A on the universe. A
crisp set is where the element x in the universe X is either a member of some crisp set
A or not. This binary issue of membership property is mathematically expressed as;
Ax
AxxAX
,0
,1)( (2)
The symbol )(xAX gives the indication of a definite membership of element x in set
A, and the symbol and indicate contains, or not contains in the set. In continuing
with the height example, suppose that set A is a crisp set of all people with 0.70.5 x
feet according to Fig. 4. If a particular x1 has a height of 6.0 feet, the membership in crisp
set A is equal to 1 and it is expressed as 11xAX . In contrast, if x2 has a height of
4.99 feet, the membership of this individual in set A is 0, expressed as 02xAX , as
shown in Fig. 2.3. In these cases, the membership is binary, which is either an element
is a member of a set or not.
Ref. code: 25605422300342CMU
37
Figure 2.3: Set A (top), and the crisp set A (bottom)
However, Zadeh extended the notion of binary membership to accommodate the
“various degrees of membership” to describe the uncertainty parameters which valued
in the real unit interval 1,0 , where the endpoints of 0 and 1 conform to no membership
and full membership. Similar to the crisp set, but the infinite number of values in
between the endpoints can represent various degrees of membership for an element x.
Such a membership function is displayed by Fig. 2.4.
Figure 2.4: A fuzzy set H
Consider a set H consisting of heights near 6 feet, the property of near 6 feet is fuzzy
and there is no unique membership function for H. The sets on the universe X that can
accommodate the “degrees of membership” were termed as the fuzzy sets, and with the
above example, it is denoted as H in the figure 2.4. The fuzzy set H is the function
H that carries X into 1,0 . The mathematical representation of a fuzzy set used in
this dissertation is denoted as A~
, where the functional mapping is given as the
following equation.
1,0)(~ xA
(3)
Ref. code: 25605422300342CMU
38
The symbol )(~ xA
is the degree of membership of the element x in a fuzzy set A~
, hence
)(~ xA
is a value on the unit interval that measures the degree to which element x belongs
to fuzzy set A~ . In this dissertation, the fuzzy parameters are used to represent the
sources of uncertainty in the production system, and they are described as triangular
fuzzy numbers (TFNs). The TFNs are denoted by fuzzy set A~
, and they are defined as
(a, b, c) as depicted in Eq. (4)
otherwise,0
,
,
)(~ cxbbc
xc
bxaab
ax
xA
(4)
2.4.2 Defuzzification to crisp sets.
In practice, there may be situations where the output of a fuzzy process needs to
transform to a scalar quantity. Let’s consider a fuzzy set A~
, the set A , is a crisp set
called the lampda (λ)-cut or (alpha-cut) set of the fuzzy set A~
, and it is used to represent
uncertainty. The A is derived from the parent fuzzy set A~
, where 10 and
)(~| xA
xA . The crisp set A is exhibited in Fig 2.5.
Figure 2.5: A fuzzy set with λ cut
There are actually a lot of defuzzification techniques being provided in the literature.
Common techniques include the center of gravity (Amnar and Wright, 2000, and Arikan
and Gungor ,2001), mean of maximum method (Bojadziev, 1995), and the weighted
average of maximum values of membership functions method (Siler, 1987). The first
method is computational intensive, while the other two are less complicated. However,
Ref. code: 25605422300342CMU
39
to avoid the complexity of computation, and as there are only three discrete fuzzy
numbers to represent uncertainty in the production system that obtained based on the
A . We select a centroid method (Chou and Chang, 2008) which is the simplest
technique to defuzzify the fuzzy numbers. To defuzzify the TFNs, the centroid of
cbaA ,,~ is determined by Eq. (5)
3
~cba
AC
(5)
In the next section, the application of the fuzzy set theory to model the supply chain
uncertainties problems, and the integration of the fuzzy set theory in the quality
deployment function to improve the product quality are introduced.
2.4.3 The fuzzy MILP model for supply chain planning under uncertainties.
FST has provided an efficient evaluation of a system, and was continuously used
until present, for example, in a control system (Iijima et al., 1995, Monfared and Steiner,
2000), resource allocation (John and Bennett, 1997), cellular manufacturing for small
batch production (Arikan and Gungor, 2001), performance evaluation (Ammar and
Wright, 2000), planning and scheduling (Majozi and Zhu, 2005), supply chain
production planning (Mula et al., 2010, Bilgen, 2010), supplier selection (Yucel and
Guneri, 2011, Lima-Junior, 2016), and system design (Ubando et al., 2016). The
popularity of the fuzzy MILP model to work with the supply chain uncertainty has
broadly appeared in various supply chain research fields as discussed in the following
literature review.
(1) In inventory management, Petrovic et al., (1999) used the fuzzy modeling
simulation of the supply chain to determine the stock levels and order quantities to
achieve the acceptable level of deliver performance that minimize the total cost for the
whole supply chain. Giannoccaro et al., (2003) developed a method to define inventory
management policy and use the FST to model uncertainty based on demand and
inventory cost. Carlsson and Fuller (2002) proposed the fuzzy logic approach to reduce
Ref. code: 25605422300342CMU
40
the bullwhip effect. Wang and Shu (2005) used the FST to represent uncertainty of the
customer demands, processing times, and reliable delivery, and then use the genetic
algorithms to propose a decision-making model.
(2) In supplier selection, Kumar et al., (2004) presented a fuzzy goal programming
model to select the supplier in the supply chain by using the triangular membership
functions for each fuzzy objective. The objective is to solve the three basic problems of
minimizing the total cost, rejects selection within the network, and delays in delivery.
Amid et al., (2006) addressed the same problem of adequately selecting suppliers within
the supply chain. They devised the fuzzy multi-objective goal programming model to
consider the cost cuts, and to increase the quality and service of the supplier selected.
Yucel and Guneri (2011) also proposed a method to select the supplier. The linguistic
values are expressed as trapezoidal fuzzy numbers, and it is used to assess weight of
the factors. The approach is followed by fuzzy MILP model to overcome the supplier’s
constraints and to assign the optimum order quantities to each supplier.
(3) In transportation planning, Chanas et al., (1993) considered assumptions related
to the supply and demand for a transportation problem. Three cases of the crisp values,
interval values, and fuzzy models for the transportation problem were proposed. Shih
(1999) studied the cement transportation problem in Taiwan using the fuzzy linear
programing based on the constraints of port capacities, demand fulfillment, loading-
unloading capacities, and constraints related to traffic control (Zimmermann, 1978,
Chanas, 1983, and Julien, 1994). Liu and Kao (2004) developed the method to obtain the
membership function of the total transport cost by considering the shipment cost,
supply, and demand as the fuzzy numbers, and Liang (2006) proposed an interactive
multi-objective LP model to solve the fuzzy transportation problems with a piecewise
linear membership function.
(4) In production-distribution planning, Sakawa et al., (2001) and Liang (2008)
proposed the production and transportation problems using the deterministic model and
fuzzy multi objective LP model to minimize costs in the integrated production-
Ref. code: 25605422300342CMU
41
transportation supply chains. Selim et al., (2008) applied the fuzzy goal-based
programming approach to a planning problems of a collaborative production-
distribution supply chain. The fuzzy elements considered in the objective functions are
to maximizing profits for manufacturers and distribution centers, retailer cost cuts, and
minimize the delay in retailers. Lastly, Aliev et al., (2007) developed an integrated multi-
period, multi-product fuzzy production and distribution aggregate planning model by
trading off between the fuzzy market demand and the profit.
(5) In procurement-production-distribution planning, Chen and Chang (2006)
developed an approach to derive the membership function of the fuzzy minimum total
cost of multi-products, multi-echelon, and multi-period supply chain model where the
cost of raw materials, unit transportation cost, and the demand quantity were fuzzy
numbers. And just recently, Torabi and Hassini (2008) proposed a new multi-objective
possibilistic MILP model for procurement, production, and distribution planning with
uncertainty in market demand, cost-time coefficients, and the capacity levels. The
proposed method was validated by the numerical tests.
From the introduction of uncertainty in the supply chain system to the applications
of FST in the supply chain planning problems, it can be seen that FST has provided as
the efficient way to manage the uncertainties in many supply chain environments. In
this dissertation, we aim to use the MILP model with the fuzzy parameters, as a
predictive model, to solve the production planning problem of a case study according
to the proposed methodology. The aim of the MILP model is to determine the optimal
plan for the limited production resources that satisfy the market demands at a minimum
cost. Fuzzy parameters are used to represent the sources of uncertainty in the production
system, and they are described as triangular fuzzy numbers (TFNs). The results from the
MILP model with fuzzy parameters are analyzed further for the evaluation of supply
chain performance by using the SCOR KPIs, and to improve the performance by using
the quality function deployment methodology.
Ref. code: 25605422300342CMU
42
2.5 Quality function deployment (QFD)
The quality function deployment (QFD) is considered as a useful tool that can help
a company to move forwards to a more proactive product development. Originated in
Japan in 1970s, the QFD has been applied successfully by many Japanese, American,
and European companies for their product development (Chan and Wu, 2005). QFD
integrates the customer requirements into every aspect of the product design by
outlining the needs of the customer, and translating them into technical requirements,
so that the end products meet customer expectations. (Liu and Wang, 2010). In the
following literature review, the introduction of QFD process is introduced, and the
application of QFD in fuzzy logic is presented.
2.5.1 Fundamentals of the Quality function deployment (QFD)
Quality function deployment (QFD) is a planning tool that is used to fulfill customer
expectations. It is initiated by Prof. Yoji Akao and Mr. Oshiumi of Bridgestone Tire
(Hauser and Clausing, 1988, and Akao, 1990). The original purpose is to show the
connections between the quality, quality characteristics, and process characteristics. In
1979, Mr. Sawada of Toyota Auto Body used the matrix in a reliability study which
addressed the technical trade-offs in the quality characteristics. It was done by adding
the roof to the QFD matrix, which is named later as a “House of Quality (HOQ)
The QFD process is capable of transforming customer requirements to
implementable actions. Thus, the approach is famous in the traditional product
development process (Shen et at., 2000, Kahraman et al., 2006, Botttani and Rizzi, 2006,
Amin and Razmi, 2009, Liu, 2011, and Zhang et al., 2014). Applications of QFD range
from product development, quality management, customer needs analysis, product
design, engineering decision making, supplier selection, budget allocation, and
strategic management in logistic services (Cristiano et al., 2001a, Tsai, 2003, Lager,
2005, Botttani and Rizzi ,2006, Bevilacqua et al., 2006, Amin and Razmi, 2009, Liu,
2009, Chen and Ko, 2009, Bhattacharya et al., 2010, and Mayyas et al., 2011). A central
Ref. code: 25605422300342CMU
43
element of the QFD planning process that enables the transformation of customer
requirement to design specification is the House of Quality (HOQ) matrix that contains
the information on “Whats”, “Hows”, and the interrelationship between them to
determine the output priority level to fulfil the needs of a customer (Chen and Ko, 2010).
The typical HOQ comprises of six parts with the explanation as follows, and it is also
shown in Fig. 2.6. (Chan and Wu, 2005)
Figure 2.6: A House of Quality (HOQ) (Left), and the HOQ with detailed description
(Right)
A. The customer needs (Whats)
The customer needs (also termed as the “voice of customer”) or customer
requirement is the first, and the most important part of the HOQ matrix. It structures
the list of product’s customer requirements in their own words. The list of
requirements is usually gathered using a tree diagram. Customer needs can be
collected by various method such as survey, focus groups, interviews, observing,
feedback, and sale records (Bicknell and Bicknell, 1995). The next QFD step is to
Ref. code: 25605422300342CMU
44
rate for the relative importance. Customers are asked to give the opinion for each
“WHAT” using a five, seven, or nine-point scales, and a sufficient number of
customers should be provided to give the statistical significance. The fuzzy method
can be applied to address the vagueness and subjectivity in the people’s assessment.
B. Planning Matrix
The planning matrix is responsible for quantifying the customer’s requirement from
part A. The most important measure is the “importance weighting”, in which
generally obtained from the average of the sample gathered. In this part, the
competitive evaluation is performed. The evaluation is assessed by asking the
customers to rate the relative importance of the company’s product and its
competitors on each WHAT, and then to aggregate the customer’s rating. From this
comparative evaluation, the strategic goals can be set. These goals are numerical
and should be consistence with the rating scale that has already established. The
next step is to determine the sales points based on the previous information. A sales
point contains the information that characterizes the company’s ability to sell the
product to meet customer satisfaction, and eventually the improvement factor can
be calculated. The overall rating for each requirement is placed into the planning
matrix as indicated in Fig. 2.7.
Ref. code: 25605422300342CMU
45
Figure 2.7: The HOQ planning martix (Bozdana, 2007)
C. Technical Measures (HOWs)
The technical measures (also named as “the engineering characteristics), structures the
“technical requirements” (HOWs) which is to identify the measurable characteristics of
the product that related to a specified customer requirement. The HOWs are usually
methods, company measures, design requirements, and some substitute quality
characteristics that the company can perform to achieve the customer needs. In practice,
the technical measures usually be generated from current product standards. For the
HOWs to be properly defined, the measurement should be associated with a unit and
direction. For example, voltage in volts, time in minutes, and capacity in gallons.
D. Relationship matrix between WHATs and HOWs
The relationship matrix of WHATs-HOWs, is a systematic means for identifying the
degree of relationship, or the linkage between each WHAT and each HOW. Completing
this matrix is the vital step in the QFD process as the final analysis rely heavily on this
part. Filling in the relationship matrix by looking at each HOW to each WHAT works
better since the HOW can be defined once, and then we can determine the impact level
of a particular HOW to WHAT. Usually, there are 4 relationship levels of no, weak,
Ref. code: 25605422300342CMU
46
medium, and strong relationship. The frequent relationship scale that is used to express
the relationship is (0, 1, 3, 9) (American Supplier Institute, 1994, Cohen, 1995, and
Vasilash, 1989). The FST is also used to express this relationship in a Likert scale.
E. Technical correlation matrix
The triangular roof matrix is used to identify where the technical requirements
characteristics impede each other. Simply, it is the matrix to indicate the
interrelationship between the HOWs themselves. After the HOWs have been identified,
the technical team will determine if one HOW is changed, how the others will be effect.
The degrees, and the direction of influences have serious impacts on the development
effort. For example, the negative impacts of one HOW to the others represent
bottlenecks in the design that call for special attention. A set of symbol is usually used
to represent the impact, where these impacts can be converted to the fuzzy numerical
scales for further analysis.
F. Technical matrix
The technical matrix contains much of the information that is linked to both customer
needs and the parts characteristics in the QFD’s second phase. It provides the initial
rank of the technical measures based on the previous information. Actually, the
technical matrix comprises of 3 parts, which are the technical priorities, competitive
benchmarking, and target settings. The technical priority is the relative importance of
each technical requirement of the product, obtained by calculating the weight in the
planning matrix, and the relationship matrix. Each interrelationship weighting is
multiplied by the overall weight from the planning matrix, and the values are summed
down to give a priority score for each technical requirement. The above statement is
illustrated by the following simple additive weighting formula.
WHAT
HOW and ATbetween WH valueiprelationsh WHATof rating importance Final HOW ofpriority Technical
Ref. code: 25605422300342CMU
47
After determining the HOWs’ relative importance, it is possible to conduct a
competitive technical assessment to compare our product’s technical performance to
the competitors to achieve a better position in a new product, and this is called the
competitive benchmarking. Even though this step can be done by marketing, but it is
still difficult to obtain all of the information from the competitors as some technical
parameters and the know-hows are not available. A careful technical assessment should
be conducted to give a reliable score that can represent the technical performance of
competitors. Lastly, the final output of the HOQ matrix is to set the engineering target
that can be met by the new product design. Nevertheless, not all parts of the technical
matrix assessment need to be conducted. Researchers can choose to work only the first
and second part of the technical matrix if the outputs from the QFD processes are
satisfied with the situations.
2.5.2 Further process after the QFD
The procedure until here is not the end of the QFD process. The output of the HOQ
matrix can be further utilized as the first stage of the four phases model (Hauser and
Clausing, 1988, and Chen and Ko, 2010) which consisting of the product planning, part
deployment, process planning, and the production planning phases. The four phases
model is exhibited in Fig. 2.8 and explaining as follows.
Figure 2.8: The typical 4 phases QFD model
Ref. code: 25605422300342CMU
48
Phase 1: The product planning phase translates the qualitative customer
requirements (CRs) into the measurable engineering characteristics (ECs), and
then to identify the important ECs.
Phase 2: The part deployment phase converts the output of the product
planning phase into the critical part characteristics (PCs), and then to explore
the relationship between ECs and PCs.
Phase 3: The process planning phase established the relationship between PCs
and the manufacturing operations that related to a part, and the critical process
parameters are identified in the operation instructions.
Phase 4: The production planning phase translates the manufacturing
operations into the production standards or the work instructions. For example,
number of parts to be checked, types of tools to be used, and the inspection
method to be performed.
The four phases of QFD share a similar structure and analysis processes. Each phase is
composed of its WHATs and HOWs, and each phase focuses on the priority analysis
of these items based on the information available. However, most of the existing works
that relate the QFD process to other issues focus on only the first phase of QFD, as it is
adequate to provide a systematic way to translate the voice of customer (VOC) into
engineering characteristics (ECs) (Chen and Weng, 2003, 2006, Kwong et al., 2007,
Chen and Ko, 2008). In the following section, the literature review regarding of the
application of the QFD, and QFD with the fuzzy logic are presented.
2.5.3 Fuzzy QFD
For traditional QFD process, the customer rating and relationship rating in the HOQ is
expressed by a point system such as 1-3-5 or 1-5-9. This indicates a linguistic judgment
such as “weak”, “moderate”, and “strong”. However, when a human decision is imprecise,
the fuzzy set theory is introduced as a suitable method to process these decisions
numerically (Liu, 2009). As discussed previously in the last literature review section,
Ref. code: 25605422300342CMU
49
fuzzy set theory (FST) involves a set with elements that have the degree of membership
valued in the real unit interval 1,0 , and the membership function is expressed as )(x
(Zadeh, 1965). A major contribution of FST is that it represents vague data and it
resembles human thought when generating decisions. However in the study of fuzzy
QFD, most of the fuzzy opinions are presented as triangular fuzzy numbers (TFNs),
denoted by A~
, and they are defined as (a,b,c). The membership function )(~ xA
is
presented in Eq. (6) and exhibited in Fig. 2.9.
otherwise,0
,
,
~ cxbbc
xc
bxaab
ax
xA
(6)
Figure 2.9: A triangular fuzzy number
From the literature review, the fuzzy QFD method has been successfully integrated with
other fields of study such as part deployment (Sohn and Choi, 2001, Chen and Weng,
2006, Liu, 2009), material selection (Mayyas et al., 2011), service assessment (Lapidus
and Schibrowsky, 1994, Stuart and Tax, 1996), supplier selection (Bevilacqua et al.,
2006, Amin and Razmi, 2009, Karsak and Dursun, 2015), service provider selection
(Amin and Razmi, 2009, Bhattacharya et al., 2010, Liao and Kao, 2014, Wang, 2015),
lean and agility (Bottani, 2009, Zarei et al., 2011), and a strategic planning and decision
making process (Yang et al., 2003, Partovi, 2006, Jia and Bai, 2011). The application of
QFD is also related to operation research tools to enhance many of the business
Ref. code: 25605422300342CMU
50
processes. The summarized literature reviews on types of QFD methodology and the
purpose of the study are depicted in Table 2.5.
Table 2.5: Types of QFD methodology and purpose of study
Types of QFD methodology Purpose Authors
Typical QFD model Customer Satisfaction
Service Quality
Supply chain leanness
Lapidus and Schibrowsky (1994)
Stuart and Tax (1996)
Zarei et al., (2011) Fuzzy QFD Part deployment
Supplier Selection
Logistic Service
Agility measurement
Manufacturing Strategy
New Product Design
Environment Consideration
Customer Requirement
Software Selection
Target Setting
Performance Measurement
Sohn and Choi (2001) Yang et al., (2003) Chen and Weng (2006) Liu (2009)
Bevilacqua et al., (2006) Amin and Razmi (2009) Karsak and Dursun (2015) Wang (2015)
Liao and Kao (2014)
Bottani (2009)
Jia and Bai (2011)
Chen and Ko (2009)
Kuo et al., (2009)
Ramasary and Selladurai (2004)
Sen and Baracli (2010) Senar and Karsak (2011) Kannan et al., (2013
Kano based QFD Product Life Cycle Management
Prioritize CR
Lee et al., (2008)
Nahm (2013) AHP integrated QFD models Material Selection
Supplier Selection
Strategic Sourcing
Mayyas et al., (2011)
Bhattacharya et al., (2010)
Ho et al., (2011) Ho et al., (2012)
Ref. code: 25605422300342CMU
51
ANP integrated QFD models Product Planning
Strategic Vision
Environment Consideration
Manufacturing Strategy
Product Development
Karsak et al., (2002) Büyüközkan et al., (2004), Kahraman et al., (2006)
Partovi (2006)
Lin et al., (2010)
Liu and Wang (2010)
Zaim et al., (2014)
From table 2.5, it can be seen that most of the QFD methodology is mostly
related with the fuzzy number, and becoming a fuzzy QFD to solve many problems in
the supply chain. For the extended literature review, it can be found as follows. Chen
and Ko (2009) developed two phases of QFD that involves the fuzzy nonlinear
programming model based on Kano’s analysis to determine the fulfillment level of part
characteristics that effectively meet the target design requirements in QFD phase 1. In
2010, the same authors have extended the study to consider all phases of the QFD by
using a means-end chain model to connect the relationship between attributes of the
requirements to achieve higher design requirements. (Chen and Ko, 2010). Jia and Bai
(2011) proposed an approach to derive a manufacturing strategy using QFD
methodology. For part deployment, Liu (2009) proposed a modified fuzzy k-means
clustering QFD method and FMEA analysis to classify the bottleneck groups of part
characteristics. An α-cut operation is used to manipulate the fuzzy sets instead of regular
algebraic operations of fuzzy numbers. The advanced QFD model with an integration
of ANP is also studied by Liu and Wang (2010). ANP is used in the stage of establishing
the interrelationship in the QFD components to provide the product developer with
more information and the bottleneck level of part characteristics. Other issues that
integrate the ANP with QFD can be found in Partovi (2001), Karsak et al., (2002),
Büyüközkan et al., (2004), Kahraman et al., (2006), and Lin et al., (2010) where ANP is
used to perform the pairwise comparison for the degree of interdependence among
Ref. code: 25605422300342CMU
52
criteria. The QFD framework is also combined with AHP modeling to rank candidate
suppliers under multi-criteria environment. Bhattacharya et al., (2010) and Mayyas et al.,
(2011) also integrated the same methodology for the selection of material for automotive
parts, and they found that the hierarchical QFD methodology allows the decision maker
to select and rank choices that meet the functional objective. The two-QFD matrices
with the fusion of fuzzy information and MCDM approach are also applied by Karsak
and Dursun (2015) to not only establish the supplier selection, but also consider the
impact of inner dependence among them. The extended study of QFD methodology
includes the work by Nahm (2013) which proposed a new approach to prioritize CRs
based on company competitiveness and Kano’s analysis. Moreover, just recently, the
relationship of QFD methodology in the performance measurement field begins to gain
more attention when Kannan and Jafarian (2013) applied the ANFIS and fuzzy QFD
method to define a relationship between strategic planning and operational budgeting
using the balanced scorecard as the performance framework.
From the literature review since the introduction of SCOR model, the use of
model in performance evaluation, and the successful application of fuzzy QFD by using
the House of Quality approach, our research foresees the potential benefits of extending
the utilization of QFD philosophy to the SCOR model that would profoundly support
professionals in the supply chain management field. We propose the Fuzzy QFD
methodology to manage the SCOR KPIs for performance improvement. The concluding
remark draws from the literature review is presented in the next section before the
methodology is proposed.
2.6 Concluding remarks
From the literature review, it is recognized that the APICS SCOR model is a globally
accepted model that has been used by most of the academicians and practitioners to
address many supply chain issues. However, the literature review of SCOR model
discloses that the method to estimate the SCOR KPIs is still limited in the literature.
Based on the literature review, even though the model has provided a definition that is
Ref. code: 25605422300342CMU
53
ready to use, and practitioners can assess the performance of the supply chain
straightaway by means of data collection to fit with the definition, but such approach
lack of a procedural methodology and the output KPIs cannot be further analyzed. In
this dissertation, we realize that without the decent method to estimate the performance
that can link from the manufacturing system parameters to the SCOR KPIs, the output
KPIs will imply only on a performance of a particular state. Consequently, the
identification of direction for performance improvement to match with the strategic
goals of organization is mostly unclear. Therefore, this dissertation aims to address this
research problem by proposing a method with a model that can relate between a
manufacturing system parameters to the SCOR KPIs, so that the systematic evaluation
of the SCOR KPIs is attainable. By reviewing the fundamental of MILP model and
together with the characteristic of the research problem that need to be modelled, it is
found that the MILP model is a suitable modelling technique to represent the
manufacturing system, and to use as a predictive model for SCOR KPIs evaluation. The
predictive model is useful since it helps the company to set up the relationship between
manufacturing system parameters and the supply chain performances. As a result, when
the company performs a what-if analysis by changing or improving the manufacturing
parameters in the model, the new SCOR KPIs will be predicted. With this mechanism,
it notifies changes to the management team before making decision and without
conducting a real experiment on the manufacturing system. However, as uncertainties
in a supply chain modelling is unavoidable, so we examined several modeling
techniques and the fuzzy set theory (FST) is chosen as the method to work with
uncertainty in our research problem. In this dissertation, we aim to use the MILP model
with the fuzzy parameters to solve the production planning problem of a case study
according to the proposed methodology. The aim of the MILP model is to determine
the optimal plan for the limited production resources that satisfy the market demands
at a minimum cost. Fuzzy parameters are used to represent the sources of uncertainty in
the production system, and they are described as triangular fuzzy numbers (TFNs). The
results from the MILP model with fuzzy parameters are analyzed further for the
Ref. code: 25605422300342CMU
54
evaluation of SCOR KPIs based on the proposed methodology. Finally, to enable the
SCOR KPIs to be improved successfully according to the organization requirement, the
fuzzy QFD approach is integrated to the SCOR model. As the QFD process is renowned
for its capability to transform the customer requirements to the implementable actions
that have been cited in many research works, so our dissertation anticipates the potential
advantages of extending the QFD philosophy to the performance management field,
especially to manage the SCOR KPIs for improvement. In the next chapter, the
methodology for evaluation of SCOR KPIs by using a MILP predictive model is
proposed, whereby the QFD approach to manage these SCOR KPIs for improvement
are presented accordingly in the consequence chapter.
Ref. code: 25605422300342CMU
55
Chapter 3
Evaluation of SCOR KPIs using a predictive MILP model with fuzzy
parameters
The objective of this chapter is to propose a methodology to assess the SCOR KPIs
under uncertainties based on level 2 of the SCOR-Make process metric, including nine
KPIs. The proposed methodology consists of predictive MILP models with fuzzy
parameters and some algorithms to assess the KPIs related to agility. The novelty of this
chapter in terms of the contribution of work is to relate the manufacturing parameters
to the SCOR KPIs, and use the MILP model with fuzzy parameters to enable the
performance prediction process in many what-if scenarios. The proposed method is new
in the performance evaluation framework by using a SCOR model. A case study of a
bottled-water factory is conducted to demonstrate the application of the proposed
methodology. The findings of this chapter indicate that the proposed methodology is
capable of developing the relationship between the manufacturing parameters and the
SCOR KPIs, which enable the effective prediction process, especially when the
manufacturing parameters are changed or improved.
3.1 The proposed methodology to evaluate the SCOR KPIs
The proposed methodology for SCOR KPIs evaluation consists of two parts. The
first part is to formulate the predictive MILP model with fuzzy parameters, and the
second part is to propose the method to evaluate the SCOR KPIs based on level 2 of
the SCOR-Make process metric, including nine KPIs. Before the methodology is
presented, this research present a block diagram to explain the overall research
procedure, and it is exhibited in Fig 3.1.
Ref. code: 25605422300342CMU
56
Figure 3.1: Block diagram of the SCOR KPIs evaluation procedure
3.2 The predictive model
In this dissertation, a predictive model is used because the relationships between the
values of SCOR KPIs and the manufacturing parameters are not known. The aim of the
predictive model is to represent the manufacturing system to be studied. This is used as
a foundation to assess the SCOR KPIs of the SCOR-make process. Also, there are agility
measures in the dissertation, and without the procedural methodology, the measurement
of agility is almost impossible. The structure of the manufacturing system, the MILP
model, and the fuzzy parameters are described as follows.
3.2.1 The MILP model
The MILP model is used to determine optimal plans that are most favorable to the stated
objective function. In this case, the optimal plans involve raw material ordering,
production, and inventory planning that meet the demand requirements in each period.
The structure of the manufacturing system is presented in the Fig 3.2. In this dissertation,
the manufacturing system is a make to stock flow shop. It produces i products to fulfill
the demand Dit over T planning periods. The manufacturing process consists of K
production stages. The raw material is planned and ordered using a material requirement
planning (MRP) system. The amount of plastic resin in grams to produce each size of
the plastic bottle is τi. The machine at each stage is specific to the operation and there
Develop the MILP model (predictive model) to represent the manufacturing
system under consideration.
Apply uncertainties in the manufacturing system to the MILP model, using
TFNs, and solve the model for the optimal outputs based on uncertainties.
Evaluate the level 2 SCOR KPIs, based on the outputs of the MILP with fuzzy
parameters, and the proposed methodology.
Ref. code: 25605422300342CMU
57
are nk identical machines at each production stage k. There is a work in -process (WIP)
between production stages, and Wt workers are available in period t. The manufacturing
system operates ht shifts in period t, and each shift has δ working hours. The parameters,
decision variables, objective function, and constraints of the model are defined as
follows.
Figure 3.2 : Structure of manufacturing system
Identification of the manufacturing parameters
i Product index i =1,2,..,I
t Period index t =1,2,..,T
k Production stage index k = 1,2,..K
nk Number of machines at production stage k.
cmi Material cost of product i (Baht/pack)
cui Utility and production overhead cost of product i (Baht/pack)
cr Labour cost per one shift (Baht/person)
cii Inventory carrying cost of product i (Baht/pack/period)
cji WIP Inventory carrying cost of product i (Baht/bag/period)
cl Raw material Inventory carrying cost (Baht/ton/period)
csi Subcontract cost of product i (Baht/pack)
cbi Backorder cost of product i (Baht/pack)
cki Standard cost of WIP inventory of product i (Baht/bottle)
cnt Standard cost of Raw Material inventory of product i (Baht/kg)
ei Hours of labour per unit of product i (man-hour/unit)
Ref. code: 25605422300342CMU
58
Wt Total workforce in period t (workers)
δ Working time per one worker per shift (hours/shift)
γ Machine operating hours per day (hours/day)
Ck Production capacity of each machine in stage k (units/hour)
ht Number of shifts per day in period t
dt 1 if period t is a working day, 0 otherwise
ρi Number of units per pack of product i
θ Number of units per bag of WIP of product i
Dit Demand of product i at period t (packs)
ФDi Total number of order of product i in all periods
(orders)
Ri Selling price of product i (Baht/pack)
iI Level of safety stock of product i, according to
company policy (packs)
Smit Maximum allowable subcontract amount of product i at period t (packs)
tM Maximum raw material inventory at the end of period t: beyond this level there
is a cost penalty (tonnes)
itJ Maximum WIP inventory of product i at the end of
period t in any stage: beyond this level there is a cost penalty (units)
itI Maximum finished product inventory of product i
at the end of period t: beyond this level there is a cost penalty (packs)
tM Safety stock of raw material at the end of period t (tonnes)
itJ Safety stock at of WIP of product i at the end of period t in any stage (units)
itI Safety stock of finished product i at the end of period t (packs)
iI Target ending inventory of product i according to company policy (packs)
Gt Amount of raw material based on MRP system to be received at period t
(tonnes)
i Amount of raw material used to produce product i (grams per unit)
Ref. code: 25605422300342CMU
59
TFi Fixed component cycle time i.e. schedule time, issue material time, and release
product time per lot of product i (min)
Lik
Tik
Lot size of product i at process k (packs)
Unit processing time of product i at process k
Identification of the decision variables
Pitk Amount of product i produced at period t in stage k (units)
Sit Subcontract amount of product i at period t (packs)
Iit Inventory of product i at the end of period t (packs)
Jitk WIP Inventory of product i at the end of period t in stage k (units)
Bit Backorder amount of product i at period t (packs)
Mt Raw material inventory left at the end of period t (tonnes)
ФBi Total number of orders, with backorder of product i in all period (orders)
Objective Function
T
ttclM
I
i
T
titBicbitSics
I
i
T
t
K
k i
kitJ
icjI
i
T
titIici
I
i
T
t Kk i
kitP
icuicmT
ttdtWcr
I
i
T
titDiRMAX
11 11 1
1
11 1
1 1)(
11 1
(7)
Constraints
1. Raw material balance
tMtGtMkitP
I
ii
1
1
610 , t , k=1 (8)
2. Inventory balance
1)1(
kitPk
itPkti
JkitJ , ti , , k= 1,..,K-1 (9)
itDitS
i
kitP
tiB
tiIitBitI
)1()1( , ti , , k=K (10)
Ref. code: 25605422300342CMU
60
3. Production capacity constraint
kntdkCI
i
kitP
1
, t , k (11)
4. Workforce- production constraint
tWthtdI
i
kitPie
1 , t , k=K (12)
5. Safety stock and maximum inventory policies
5.1 Raw material inventory
tMtMtM , t (13)
5.2 WIP inventory
itJk
itJitJ , i , t ,k= 1,..,K-1 (14)
5.3 Finished products inventory
itIitIitI , i , t (15)
6. Target ending inventory of finished products
iIitI ˆ , i , t=T (16)
7. Subcontracting limitation
itSmitS , ti , (17)
8. Backordering is not allowed at the end of planning horizon
0itB , i , t=T (18)
The objective function in Eq. (7) is to maximize profit, which consists of total sale
revenues minus total manufacturing costs, including the labor cost, direct material and
production overhead costs, inventory holding cost for all production stages,
subcontracting cost, and backordering cost. Constraints (8-10) explain the inventory
balance of raw materials, WIP, and finished products. Note that constraint (10) allows
backordering of finished products. Constraint (11) represents a machine capacity that
limits the production quantity of each stage based on the machine operating hours,
Ref. code: 25605422300342CMU
61
machine capacity, workday per period, and number of machines at each stage.
Constraint (12) limits production quantity of finished products, based on available
workforce level. Constraints (13-15) control raw material, WIP, and finished product
inventory levels based on the safety stock and maximum stock policies of the company.
Constraint (16) sets the target finished product inventory at the end of planning horizon,
based on the company policy. Constraint (17) restricts the subcontracting level in each
period. Constraint (18) states that backordering is allowed in all periods except at the
end of the planning horizon, to ensure that all demands must be satisfied, although it
may be satisfied late.
3.2.2 The MILP model with fuzzy parameters.
The output obtained from the MILP model is the optimal plans that the company should
follow to get the maximum profit, but in reality, there are uncertainties in the
manufacturing system that prevent the manufacturing process from reaching the
planned outputs. In this dissertation, we consider uncertainties from manufacturing
processes, demand, and supply. The crisp set A at =0.8, based on the fuzzy set A~
, is
used to represent uncertainty. Zadeh’s notation is used to present a crisp set 8.0A
according to Eq. (19).
cbaA ,,8.0 (19)
Equation (19) explains that each fuzzy parameter contains three finite numbers, which
represent uncertainties of three scenarios. The MILP model with a, b, and c values of
fuzzy parameters is solved separately to obtain the outputs under uncertainties. To be
specific, three MILP models with three sets of parameters are solved to determine the
company’s actual output in this case. The fuzzy parameters and decision variables are
defined below.
Fuzzy parameters for uncertainty
Uncertainties from the manufacturing process
Ref. code: 25605422300342CMU
62
kn~ Number of machines in working conditions at production stage k
tW~
Total workforce that is really available in period t (workers)
~
Working time that one worker really works per shift (hours/shift)
~ Number of hours that a machine really operates per day (hours/day)
0~
M Real initial raw material inventory (tonnes)
0~
ik
J Real initial WIP inventory of product i at stage k (bottles)
0~
ik
I Real initial finished product inventory of product i (packs)
Uncertainties from the supply side
tG~
Amount of raw material really received at period t (tonnes)
itmS ~ Real maximum allowable subcontract amount of product i at period t (packs)
Uncertainties from the demand side
itD~
Real demand of product i at period t (packs)
Fuzzy decision variables for uncertainty
kitP
~ Finished product i, which is really produced at period t in stage k (units)
kitJ
~
Real WIP Inventory of product i at the end of period t in stage k (units)
itI~
Real inventory of product i at the end of period t (packs)
itS~
Real subcontracting amount of product i at period t (packs)
itB~
Real backorder amount of product i at period t (packs)
tM~
Real raw material inventory left at the end of period t (tonnes)
The fuzzy set of parameters and the decision variables are replaced in the MILP
model to solve for the optimal outputs under uncertainties. However, we input the
additional constraints to the MILP model with fuzzy parameters to ensure that the
cumulative production quantities under uncertainties do not exceed the cumulative
planned production quantity in each period. The reason is that the company cannot
Ref. code: 25605422300342CMU
63
practically produce faster than the production plan to compensate for the delay that may
occur in the future, which is not known at the present time. This is explained by
constraint (20)
tt
t
kitP
t
kitP
11
~ , i , t , k (20)
The LINGO code and are presented in the Appendix A. The outputs from the
MILP model with fuzzy parameters are then defuzzified using a centroid method which
is presented by Chou and Chang (2008). For TFNs, the centroid of cbaA ,,~ is
determined by Eq.(21)
3
~cba
AC
(21)
3.3 The proposed methodology to evaluate the SCOR KPIs
This part consists of the proposed methodology to evaluate the SCOR KPIs based
on SCOR version 10.0 (APICS,2016), and a mechanism to assess the agility measures.
The scope of this research is the manufacturing process, therefore, the level 2 SCOR
KPIs of the make process are focused on. Table 3.1 illustrates the SCOR performance
attributes, level 1 strategic metrics, and the level 2 SCOR KPIs, used in this dissertation.
Ref. code: 25605422300342CMU
64
Table 3.1: SCOR performance attributes and level 2 KPIs used in this dissertation.
3.3.1 Percent of orders delivered in Full (RL2.1)
RL 2.1 measures the percentage of orders of each product that is delivered in full with
a committed quantity within the period. It is computed as:
%100i
Di
D
iB , i (22)
3.3.2 Make cycle time (RS2.2)
Make cycle time is the average cycle time associated with the make process. It consists
of the fixed component cycle time and the variable cycle time per lot. The calculation
is expressed in Eq. (23)
K
kik
TI
iiik
LI
iiTF
1)
1(
1 (23)
Ref. code: 25605422300342CMU
65
3.3.3 Upside Make Flexibility (AG2.2)
Upside make flexibility is the average number of days that a company requires to satisfy
a demand increase of 20% from the current level. The proposed procedural methodology
to evaluate AG2.2 is presented in Fig. 3.3
Figure 3.3 : The proposed procedure to evaluate Upside Make Flexibility
Table 3.2: Options to increase production capacity and the estimated lead time.
Resources Options Lead Time
Raw material Order additional raw material using MRP 10 days
Workforce Add four more skilled workers for production. 15 days
Subcontract Increase subcontracting by 10% from current
subcontract level
21 days
Safety Stock Increase safety stock of finished products by
25% from the current level. 21 days.
Machines Purchase more production machines. Up to 4 months.
No
In a company, list the options to increase the production capacity and their lead times
(LT)., Then rank these options in ascending order of the lead times. See Table 3.2.
Apply the first option to the MILP model. Increase the demand parameters by 20%, starting at the date of the lead time of the first option.
Is the solution from the MILP model
feasible?
Add the next option to the LP model. Shift the starting date for increasing the demand
to the date of LT of this option.
Include uncertainties into the MILP model using TFNs. Solve the MILP model with
fuzzy parameters for the outputs under uncertainties.
Upside Make Flexibility is the lead time of the last option applied to the MILP
model.
Yes
Ref. code: 25605422300342CMU
66
3.3.4 Upside Make Adaptability (AG2.7)
Upside make adaptability is the maximum sustainable increased percentage of the
demand that the company can satisfy given a preparation time of 30 days. The proposed
methodology to evaluate this agility measures is explained by Fig 3.4
Figure 3.4: The proposed procedure to evaluate Upside Make Adaptability
3.3.5 Downsize Make Adaptability (AG2.12)
Downsize make adaptability is the maximum reduction percentage of demand that the
company can achieve within a preparation time of 30 days, and the reduction must not
incur extra cost on inventory holding and other penalties. The procedural evaluation of
AG2.12 is presented in Fig. 3.5
List options to increase the production capacity where lead times are within 30 days. See Table 3.2, the first four options.
See table 1. Apply all options to the MILP model. Increase the demand level after 30 days by a
small percentage, and solve the model for the planned outputs.
Include uncertainties into the MILP model using TFNs, Solve the MILP model with
fuzzy parameters for the outputs under uncertainties.
Are the solutions feasible
under uncertainties?
Gradually increase demand and solve for
planned outputs.
Upside Make Adaptability is the maximum percentage of demand that can be
increased before an infeasible solution occurs
No
Yes
Ref. code: 25605422300342CMU
67
Table 3.3: Options to decrease production capacity and the estimated lead time.
Resources Options Lead
Time
Workforce Move three skilled workers to other activities in the
factory
15 days
Working time Reduce working time from 12 to 8 hours/shift 15 days
Subcontracting Reduce subcontract level up to 10% of the initial
demand
21 days
Figure 3.5: The proposed procedure to evaluate Downsize Make Adaptability
List options to reduce production capacity where lead times are within 30 days. See
Table 3.3
See table 1. Apply all options to the MILP model. Reduce the demand level after 30 days by a small
percentage, and solve the model for the planned outputs.
See table 1. Include uncertainties into the MILP model using TFNs, Solve the MILP model with
fuzzy parameters for the outputs under uncertainties.
See table 1.
Are the solutions feasible
under uncertainties?
Is there significant additional cost
due to worker idle time?
Gradually decrease demand and solve for planned outputs
Downsize Make Adaptability is the current percentage of demand reduction before an
infeasible solution or significant additional cost occurs.
See table 1.
No
Yes
Yes
No
Ref. code: 25605422300342CMU
68
3.3.6 Cost to make (CO2.3)
Cost to make or Cost of goods sold, measured in percentage of sales revenue, is the cost
associated with buying raw materials and producing the finished products. It includes
the direct cost of labor and materials, and the indirect cost of overhead. The evaluation
is explained by Eq.(24).
I
i
T
titDiR
T
ttclM
I
i
T
titBicbitSics
I
i
T
t
K
k i
kitJ
icjI
i
T
titIici
I
i
T
t Kk i
kitP
icuicmT
ttdtWcr
1 1/
11 1
1 1
1
11 1
1 1)(
1
(24)
3.3.7 Inventory days of supply (AM2.2)
The measure of cash-to-cash cycle time actually includes the inventory days of supply,
days sales outstanding (DSO), and days payable outstanding (DPO). However, this
dissertation aims to predict the SCOR KPIs from the MILP model, so we neglect the
effect of DSO and DPO. This is expressed by Eq. (25).
(Baht) periodper make Cost to
11 1)(
1 1
1
M
mtcntM
T
t
I
iickitJ
T
t
I
iicuicmitI
T (25)
3.3.8 Return on make fixed assets
The return on make fixed assets indicates the return on the capital invested to the make
fixed assets. It is calculated as the fraction of the net profit to the fixed assets in
manufacturing facilities. The formula is presented by Eq. (26).
assets fixed make Total
costsadmin and Sales (Baht) make Cost to1 1
I
i
T
titDiR
(26)
Ref. code: 25605422300342CMU
69
3.3.8 Return on make working capital
The return on make working capital compares the revenue generated from the
manufacturing facilities to the amount of working capital. The computation is expressed
by Eq. (27), while the AP and AR are assumed to be constant in this case.
APAR11 1
)(1 1
1
costsadmin and Sales (Baht) make Cost to1 1
M
mtcntM
T
t
I
iickitJ
T
t
I
iicuicmitI
T
I
i
T
titDiR (27)
In this work, since the SCOR KPIs are evaluated based on the outputs of the
MILP model with fuzzy parameters, the outputs are also fuzzy numbers. The SCOR
KPIs need to be defuzzifed using the centroid method in Eq. (21). Results from the
proposed methodology is presented in the next section.
3.4 Data collection and case study
In this dissertation, the proposed methodology suggests that the manufacturing
system must be firstly analyzed to generate the decision variables, parameters, and
system constraints. Therefore, to test the validity of the proposed method, this
dissertation selects a simple manufacturing system, which is a make-to-stock, flow shop
configuration. The case study of the bottle-water factory, located in Sri-Racha district,
Chonburi was considered.
The data collection process was carried out by two methods of (i) the factory site
visit, and (ii) personal interview to gather the related information. The aim of the factory
site visit is to gather the operational information, such as, the current capacity, machine
configuration, current work-flow, working time and workforce requirement at each
station, cycle time, idle time, and expected machine downtime. This dissertation has
also arranged the interview session with several parties. First, with the accounting
department for the past demand information, estimated value of current fixed assets,
and other financial information. Next, with the production manager for the production
Ref. code: 25605422300342CMU
70
and inventory management policies at each production stage from the raw material
acquisition of the PET plastic resin, the inventory buffer and bottle transfer, and the
packaging of bottled water as finished products before transfer to the warehouse.
Thirdly, this dissertation interview the randomly selected labors in the factory to
observe about the issue of work transfer if they are skilled workforces. Lastly, the
researcher discussed with the owner of the factory on the issue of current operational
satisfaction, the vision for improvement when the demand is increased or decreased,
and the possibility of investment to expand current capabilities.
The bottled drinking water factory under consideration has the manufacturing
process configured according to Fig 3.6. The company produces 2 sizes (i1 = 1500cc, i2
=600 cc) of drinking water in bottles. The amount of plastic resin in grams to produce
each size of the bottle is τ1 = 4.17 and τ2 = 1.58, respectively. The manufacturing facilities
are arranged as a flow shop that consists of 2 stages (K=2), which are a bottle blowing
process and a water filling process. The company orders raw material of plastic resin to
produce the bottles based on the material requirement planning (MRP) at an amount of
2 tonnes per lot. There are 4 blowing machines for producing bottles (n1 = 4). Each has a
capacity of 1,600 bottles per hour (C1= 1,600), and they are operated for 24 hours a day
(γ=24). Empty bottles, which are a work-in-process (WIP), are stored between two
production stages, and wait to be transferred to a fill line. The water filling line is
operated by a conveyor system. The empty bottles are conveyed to a wash, filled with
water, covered with a cap, seal, inspected, shrink-wrapped into bundles, and transferred
to stock in a warehouse area. There are two fill lines (n2 = 2). Each line has a capacity of
2,400 bottles per hour (C2= 2,400), and they are operated for 24 hours per day. Currently,
13 workers are involved in the production (Wt = 13). Each unit of bottles requires on
average 0.05 man-hours (ei=0.05) and the employees work two shifts per day (ht=2), at 12
hours/shift from Monday to Friday (δ=8). The labor cost (cr) is 300 Baht/day. The
company is now subcontracting for extra capacity on average at 30% of the current
Ref. code: 25605422300342CMU
71
demand. The cost structure, inventory holding policy, options to increase and decrease
capacity, and total asset values are discussed next.
3.4.1 Cost structure and inventory holding policy.
The finished products are sold in packs, which are 6 bottles per pack for 1,500 cc (ρ1=6),
and 12 bottles per pack for 600 cc (ρ2=12). Estimated demand per day is 805 packs per
day for 1,500 cc (D1t = 805), and 3,198 packs per day for 600 cc (D2t = 3,198). The selling
price (Ri) is 40 Baht/pack for both products. Table 3.4 shows the related operating costs.
The unit for all costs is Baht/pack except the finished product and WIP inventory
holding cost, which are Baht/pack/period, and Baht/bag/period, respectively. The
standard cost for WIP inventory is in Baht/bottle, and standard cost for raw material
inventory (cnt) is 60 Baht/kg. The raw material inventory holding cost (cl) is 20
Baht/tonne/period.
Figure 3.6: The manufacturing process of a case study
Table 3.4: Operating cost information
The company’s inventory holding policy is shown in Table 3.5.
Bottle
(CC) Material
Cost
cmi
Overhead
Cost
cui
Subcontracting
Cost
csi
Backorder
Cost
cbi
FP
Inventory
holding
Cost
cii
WIP
Inventory
holding
Cost
cji
Standard
Cost of
WIP
Inventory
cki
1,500 21.68 3 35.96 7 0.72 0.8 1.35
600 23.2 4 37.13 9 0.8 0.833 2.95
Ref. code: 25605422300342CMU
72
Table 3.5: Inventory holding policy of the company.
The WIP in between the process is stored and transferred in bags, which are 380 bottles
per bag for 1,500 cc (ϴ1=380), and 720 bottles per bag for 600 cc (ϴ2=720). The options
to increase and decrease production capacity to analyse the agility measures are
presented in Tables 3.2 and 3.3.
3.4.2 Current fixed assets, estimated accounts receivable, and accounts payable.
From the collected data, the company can estimate total fixed assets as shown in Table
3.6. In this case study, the estimated accounts receivable and accounts payable are
5,286,107 Baht and 2,509,905 Baht, respectively.
Table 3.6: Estimated company's total fixed assets. Make Fixed Assets Value
(THB) 1. Land 15,000,000
2. Building, factory, office, and warehouse 5,000,000
3. Four blowing machines at current book value 2,200,000
4. Two fill lines at current book value 2,800,000
Estimated Total Make Fixed Assets 25,000,000
The sources of uncertainty are presented by TFNs, using a crisp set A at =0.8. The
fuzzy parameters used in the MILP model are presented in Table 3.7
Inventory Maximum inventory limit Minimum inventory limit
1,500 cc 600 cc 1,500 cc 600 cc
WIP (bottles) 19,000 ( tJ1 ) 50,000 ( tJ 2 ) 0 ( tJ1 ) 15,000 ( tJ 2 )
Finish
products
(packs)
2,500 ( tI1 ) 5,000 ( tI 2 ) 375 ( tI1 ) 600 ( tI 2 )
Raw materials 0.82 tonnes ( tM ) 0.13 tonnes ( tM )
Ref. code: 25605422300342CMU
73
Table 3.7: The fuzzy parameters used in the MILP model
3.5 Results and discussion
The proposed methodology is applied to the case study to demonstrate the practicality
of the method. Results are presented in two parts; first is the outputs from the predictive
model and second is the outcomes of the SCOR KPIs based on the proposed method.
3.5.1 Outputs from the predictive model
The optimal outputs based on the provided data and MILP model are presented in Table
3.8 in terms of the total cost structure, according to the stated objective functions and
model constraints.
Fuzzy Parameters 8.0A
Number of blowing machines in working condition (n1) 4,4,3
Number of fills line in working condition (n2) 2,2,1
Total workforce that is available (W’t) 13,13,12
Working time for one worker work per shift (δ') 12,12,11
Number of hours that a machine operates per day (γ') 24,24,22
Real initial raw material inventory (M’0) 53.0,5.0,47.0
Real initial WIP inventory of product 1 at stage 1 (J’110) 400,9,000,9,600,8
Real initial WIP inventory of product 2 at stage 1 (J’120) 600,37,000,36,400,34
Real initial finished product inventory of product 1 (I’10) 020,1,000,1,980
Real initial finished product inventory of product 2 (I’20) 140,6,000,6,860,5
*Amount of raw material really received at period t (G’t) 632.1,451.1,179.1
Real maximum allowable subcontract amount of product 1 at
period t (Sm’1t) 824,792,760
Real maximum allowable subcontract amount of product 2 at
period t (Sm’2t) 058,3,940,2,822,2
Real demand of product 1 at period t (D’1t) 875,805,734
Real demand of product 2 at period t (D’2t) 390,3,200,3,009,3
Ref. code: 25605422300342CMU
74
Table 3.8: Outputs from the MILP model, and MILP model with fuzzy parameters.
MILP model
Defuzzified outputs of
MILP model with fuzzy
parameters. Total Revenue
7,689,600 (Baht) 7,688,883
(Baht) Total Cost
1) Production 5,131,522 5,022,173
2) FP Inventory cost
3,447
3,548 3) Backorder cost - 2,234 4) Subcontract cost 45 149,742 5) Labor cost 187,200 184,320
6) WIP Inventory 1,138 1,156 7) Raw material inventory 841 855
Total (COGS) 5,324,193 5,364,028 Gross Profit 2,365,407 2,324,856 Operating expenses (10% of
revenue) 768,960 768,888
Net Profit 1,596,447 1,555,967
The revenue and cost structure of the planned outputs from the MILP model is
compared to the outputs under uncertainties. The results indicate that the average net
profit is decreased when uncertainties exist. This is because there is a variation in the
production resources, which is sometimes up or down, and the company cannot manage
to produce according to the plan. Therefore, to meet the required demand in each period,
subcontracting is needed and backordering is unavoidable, which result in higher
subcontracting and backorder costs.
3.5.2 The SCOR KPIs
From the outputs of the predictive model and the proposed methodology to evaluate the
SCOR KPIs, the performance of the company is presented in Table 3.9, and illustrated
graphically in Fig 3.7.
Ref. code: 25605422300342CMU
75
Table 3.9: SCOR KPIs of the company
Figure 3.7: Graphical representation of the SCOR KPIs
Since the MILP model with fuzzy parameters is used to determine the output under
uncertainty, the SCOR KPIs derived from the proposed methodology are TFNs, as
depicted in column 2 of Table 3.9. The advantage of the TFNs is that they allow a
management team to understand the ranges of SCOR KPIs under uncertainties that
occur in the manufacturing system. The fuzzy solution is defuzzified as shown in
Level 2 SCOR KPIs (make process) Outputs of the SCOR
KPIs at 8.0A Defuzzified
SCOR KPIs
Value of each scale in
the spider diagram
Percent of Orders Delivered in Full (i=1) %100,%100,%33.98 99.44 % 0, 80, 85, 90, 95, 100
Percent of Orders Delivered in Full (i=2) %100,%100,%67.96 98.89% 0, 80, 85, 90, 95, 100
Make Cycle Time 9.7,7.88.8 8.46 hours
10,8,6,4,2,0
Upside Make Flexibility 1,5,5 3.67 days
200,160,120,80,40,0
Upside make adaptability %129%,132%,60 107%
0,40,80,120,160,200
Downsize make adaptability %20,%37,%43 33%
0 ,10, 20, 30, 40, 50
Cost to serve %38.79%,24.79%,59.80 79.74%
90, 85, 80, 75, 70, 0
Inventory Days of Supply 0.19,8.17,4.17 18 days 25,20,15,10,5,0
Return on make fixed assets %56,35%,31.38%,15.38
37.3%
0,20,25,30,35,40
Return on make working capital %4.27,%0.27%,2.26 27% 0,20,25,30,35,40
Ref. code: 25605422300342CMU
76
column 3. The SCOR KPIs based on the proposed method and above case study
indicates that the company can now fulfil 99.44%, and 98.89% of orders for 1,500 cc and
600 cc bottles, respectively. The actual cycle time to produce bottles of water is
approximately 8.46 hours per cycle. When the demand is suddenly increased by 20%,
the company takes around 4 days to response to this change. Due to a sufficient capacity
and given a preparation time of 30 days, the upside make adaptability or the ability that
the company can cope with the increase in demand is 107%. In contrast, the company
can reduce the production capacity by 33% without an additional cost or inventory
penalty. The cost to make, calculated as a percentage of total revenue, is 79.74%. The
inventory day of supply is only 18 days. The return on make fixed assets, and return on
make working capital are estimated at 37.3% and 27%, respectively. From the numerical
results, a spider diagram is presented to display the value of SCOR KPIs based on the
9 metrics. The scale in column 4 of Table 3.9 is obtained from the opinion of the
management team, based on a satisfaction level for each KPI. The diagram is also used
for comparison when there is an improvement of KPIs in the future. For example, the
scale of the percent of orders delivered in full starts from 80% because the management
team feels that 80% is the minimum acceptable level for their company. The scale of
some KPIs starts from the maximum to the minimum, such as the total cost to serve,
because lower is the better. According to the spider diagram, it is seen that most of the
KPIs are located quite far from the center. This indicates that the operating
performance, based on the SCOR KPIs of this company, is satisfactory. Based on the
results obtained from the predictive model and the achievement of SCOR KPIs from
the proposed method, the findings indicate that the proposed method is effective to
predict the SCOR performances in a real situation. Moreover, since the MILP model is
a predictive model, it can be used to perform a what-if analysis to estimate the KPIs
under different situations. For example, when the management team needs to invest in
more assets and needs to know the consequences of future performances, or when the
management team want to improve the SCOR KPIs which will be discussed in the next
Ref. code: 25605422300342CMU
77
chapter. For a measurement of agility, as flexibility analysis is a key strategic role to
improve responsiveness, the proposed method here can be applied to answer other
agility questions that may be different from the definitions of the SCOR model.
However, the MILP model presented in this dissertation is only applied to the current
situation. It is suggested that the model should be further applied to various situations
to establish a stronger relationship between the predictive model and the SCOR KPIs,
to make the evaluation of SCOR KPIs more accurate. Lastly, the model and proposed
methodology can be a good foundation to evaluate performance in a supply chain
system that is not limited to the make process.
3.6 Concluding remarks
The SCOR model is a process reference model that is widely recognized in the supply
chain research field, and the framework has been successfully used to improve
businesses in various industries. However, among the current research works, the
method for evaluation of the SCOR KPIs is still limited. The SCOR model has provided
a definition to assess these KPIs directly, but without a procedural methodology, the
resulting KPIs cannot be further analyzed. This chapter proposes a method to evaluate
the SCOR KPIs based on the predictive model. It consists of the MILP model that is
used to represent the operations of the company, the MILP model with fuzzy parameters
to address the uncertainties from the operations, and a methodology to evaluate the
SCOR KPIs based on the level 2 of the SCOR-Make process with some algorithms to
assess the KPIs related to agility. TFNs with a specific crisp set are used to represent
uncertainties. A case study of a make-to-stock, bottled water manufacturer is used to
demonstrate an application of the method. The proposed methodology provides
theoretical and practical contributions to the field of supply chain management and
performance measurements as follows:
1. The proposed methodology to evaluate the SCOR KPIs based on the predictive
model is new and original.
Ref. code: 25605422300342CMU
78
2. The proposed approach is capable of establishing the relationship between the
SCOR KPIs and manufacturing parameters. Thus, it enables the prediction of the
performance when the manufacturing parameters are changed.
3. The proposed methodology consists of a procedural method and a model to evaluate
the agility in the SCOR metrics.
4. A real industrial case study is used to demonstrate that the SCOR KPIs of the
company can be evaluated based on the proposed approach.
This theoretical contribution of the dissertation still has some limitations that can be
improved further. First, when the characteristics of the manufacturing system are
changed, the parameters and constraints of the MILP models need to be adjusted to the
particular case. A further research to construct a software to automatically generate the
MILP model based on manufacturing system structure and parameters is recommended.
Second, the value of each scale of the spider diagram is obtained based on an opinion
of the management team of the company. Thus, it should be revised when applied to
other companies. In this case, it is suggested that some visualization technique such as
R-statistical modelling can be applied to the spider diagram to demonstrate a real-time
performance comparison when the manufacturing parameters are changes. And lastly,
the current scope of this dissertation considers only the manufacturing aspect of the
SCOR-Make process, therefore further research can be extended to cover the evaluation
of other processes, namely, plan source deliver, and return, in a supply chain system.
Ref. code: 25605422300342CMU
79
Chapter 4
Fuzzy QFD approach for managing SCOR performance indicators
The Supply Chain Operations Reference (SCOR) model and their KPIs are well-
recognized model and widely used to assess the supply chain performances by many
industrial practitioners. In chapter 3, this dissertation has successfully applied the
SCOR model and proposes the predictive model to evaluate the SCOR KPIs. However,
based on these KPIs, it is still difficult to determine what actions should be carried out
in order to improve them. This chapter proposes a new fuzzy QFD approach to manage
the SCOR KPIs. The SCOR KPIs are specified as “Whats” and the technical
improvement actions (TIs) are specified as “Hows”. The proposed fuzzy QFD approach
will prioritize the TIs to be implemented to achieve the target SCOR KPIs. The same
case study of bottled water manufacturing is used to demonstrate the application of the
proposed approach. The dissertation is the first attempt to develop the fuzzy QFD
approach to manage SCOR KPIs using real industrial case study.
4.1 The proposed methodology to manage SCOR KPIs using fuzzy QFD
Before proposing the fuzzy QFD approach to manage the SCOR KPIs, a research
flowchart to establish the sequence of the methodology is presented. Fig. 4.1 shows the
block diagram of the overall research methodology.
Ref. code: 25605422300342CMU
80
Figure 4.1: Block diagram of the research methodology
From the block diagram it is explained that, in any manufacturing system to be studied,
the current SCOR KPIs of the system should be measured first. When the current SCOR
KPIs are disclosed, the management team is now able to apply the fuzzy QFD approach
to manage these KPIs for improvement. Generally, the technical improvement actions
(TIs or Hows) for each manufacturing system are identified in this step. The company
can practically implement the suggested TIs, and then evaluate the new SCOR KPIs.
However, it can take a long time to accomplish this. Alternatively, it is possible to
predict the new SCOR KPIs by using the predictive model which takes a shorter time.
This dissertation applies the latter case. The mixed integer linear programming (MILP)
model which is proposed earlier in chapter 3 is used to determine the operations of the
manufacturing system, to validate the current SCOR KPIs and to evaluate the new
SCOR KPIs after the TIs are applied. Once the new SCOR KPIs are determined, the
management team can consider whether they are satisfied with the results. If not, the
procedure is repeated by setting the new target KPIs, and the fuzzy QFD approach is
re-applied to improve the performance continuously until the company obtains
Ref. code: 25605422300342CMU
81
satisfactory KPIs. The proposed method of using the Fuzzy QFD to manage the SCOR
KPIs is exhibited in detail in Section 4.1.1
4.1.1 Fuzzy QFD approach for managing SCOR KPIs
In this section, the eight-step QFD approach for managing the SCOR KPIs for
performance improvement based on the HOQ planning matrix is established. Fig. 4.2
presents the overall structure of the method.
Figure 4.2: Fuzzy QFD approach for managing SCOR KPIs
Step1: Identification of SCOR KPIs as “Whats” and evaluate the existing SCOR KPIs
This dissertation uses 9 SCOR KPIs from SCOR 10.0, level 2, which allows separate
study on each process, and focuses on only Make processes. The dissertation exclude
the value-at-risk indicator as the study of risk analysis is beyond the scope. The SCOR
KPIs are identified as “Whats”. The list of SCOR KPIs and the definitions used in this
dissertation are presented in Table 4.1. The company collects the required information
and data from the existing system to evaluate the values of SCOR KPIs (Wm), according
to the definitions in Table 4.1.
Ref. code: 25605422300342CMU
82
Table 4.1: SCOR KPIs focusing on Make process and definitions used in this
dissertation
Attributes SCOR KPIs Definition
Reliability (W1) Make order fulfillment (%) Percentage of demand that is satisfied on time (not
back ordered), Responsiveness (W2) Make cycle time (hours per
cycle)
A period of time required to setup and produce one
production batch of all products.
Agility (W3) Upside make flexibility
(days)
Number of days required to achieve an unplanned
sustainable 20% increase in production quantity
(W4) Upside make adaptability (%) Maximum sustainable percentage increase of
production quantity to be achieved in 30 days. (W5) Downsize make adaptability
(%)
Maximum sustainable percentage decrease of
production quantity to be achieved in 30 days,
without excess inventory and/or cost penalty.
Costs (W6) Cost to make (%) Total production cost as a percentage of sale
revenue. It includes material, labor, inventory
holding, utility, subcontracting, and backordering
costs. Assets (W7) Inventory days of supply
(days)
Total inventory values divided by average cost of
goods sold per day.
(W8) Return on make fixed assets
(%)
Percentage of net profit per fixed assets used in
production.
(W9) Return on make working
capital (%)
Percentage of net profit per total inventory value
Step 2: Calculate the relative importance of SCOR KPIs.
Management consensus is the most important input factor to manage the performance
indicators in the QFD process, and it usually comes from the discussion and decision
making that is agreed among the management team (Liu, 2009). Since the relative
importance of each SCOR KPI is not equally important, and each decision maker may
impose different opinions on the relative importance, a company should form a group
to manage performances that include experts from various departments. Each expert
expresses an opinion by rating the relative importance of SCOR KPIs using linguistic
judgment. The fuzzy set theory is used to capture the uncertainty of human thought. Let
Ref. code: 25605422300342CMU
83
U = {VL, L, M, H, VH} be the linguistic sets to express opinions on a 5-point scale. U
is quantified using triangular fuzzy numbers (TFNs) as shown in Fig. 4.3
Figure 4.3: Linguistic representation of U
Wm denotes the mth SCOR KPI. The expert DMk rates the relative importance as mkg~ ,
then the average relative importance rating mg~ is calculated by Eq. (28) (Bevilacqua et
al., 2006, Amin and Razmi, 2009)
mkmmm gggK
g ~~~1~21 , m (28)
MKMMM
K
K
K
gggW
gggW
gggW
DMDMDM
~.~~.....
~.~~
~.~~.
21
222212
112111
21
Step3: Conduct best practice analysis
The APICS Supply Chain Council (SCC) has provided guidelines for industry best
practices on performance indicators for a specific industry at a particular size. The
dissertation obtains the average SCOR performance outputs from the SCC organization
database in Thailand is that registered with APICS for the small-medium sized beverage
industry. The best practices of SCOR KPIs are summarized and denoted by zm, and
presented in the case study section.
Ref. code: 25605422300342CMU
84
Step4: Perform competitive analysis to obtain improvement rating
When the management team gets the values of SCOR KPIs of the current system and
the best practices, a competitive analysis is performed by comparing them. Then, the
management team sets the target of SCOR KPIs for improvement. In the traditional
QFD method, the improvement rating um is computed by determining the ratio between
the target and the current performance (Chan and Wu, 2005, Liu and Wang, 2010,
Nahm, 2013). The higher the ratio, the wider the gap between the target and current
performance, and that KPI should get a higher priority for improvement. However, this
method is not suitable due to the following reasons.
1. The SCOR KPIs are measured using various scales. For some indicators,
higher values mean better performance, and vice versa. Hence, it is not practical
to compare using the ratio method.
2. For some KPIs, after comparison with the best practices, the improvement
ratio may not be so high, but that KPI might contribute to significant
improvement as it reflects the core competency of the company. These KPIs
should have a high improvement rating.
With these reasons, this dissertation proposes a new approach for the competitive
analysis to obtain the improvement rating. Instead of using the ratio method, the
management team will firstly analyze the KPIs between the current one and the best
practices to identify the target SCOR KPIs which are denoted as am and then evaluate
the degree of significance for improvement ( mu~ ) for each KPI. In this case, mu~ indicates
the improvement rating which represents the opinion of the management team using
the same linguistic scale in Fig. 20. The details and demonstration of the method are
shown in the case study section.
Ref. code: 25605422300342CMU
85
Step5: Determine final importance rating of SCOR KPIs
In general, the SCOR KPI that receives relatively high degree of relative importance (
mg~ ) and improvement rating ( mu~ ) should gain a high priority for improvement.
According to the multiplication laws of two TFNs in Eq.(29), the final importance rating
mW *~ is calculated by the relative importance mg~ , and the improvement rating mu~ , as
shown in Eq.(30). mW *~ is defuzzified as DF
mW * using Eq. (31). Then,
DF
mW *is normalized
as NORM
mW * using Eq. (32)
),,(),,(),,(~~
332211321321 babababbbaaaBA (29)
mmm ugW ~~~ * , m (30)
*
3
*
2
*
1
*
3
1mmm
DF
m WWWW , m (31)
DF
mm
DF
mm
DF
mm
DF
mNORM
mWW
WWW
**
***
minmax
min
, m (32)
Step6: Identification of technical improvement actions (TIs)
The technical improvement actions (TIs) are the list of actions that can be implemented
to improve the performance of the enterprise. It is defined as a set of “HOWs”, and it
should be controllable. For the production system, Table 4.2 shows examples of possible
actions to be performed. The technical improvement list is dependent on opinions of
experts and the existing facilities of the company.
Ref. code: 25605422300342CMU
86
Table 4.2: List of possible Technical Improvement actions (TIs) in a manufacturing
system
Technical Improvements (TIs) Description
Increase production capacity Invest in more production machines to increase overall
capacity
Increase production workforce Hire more workers to increase capacity
Increase working days per week For example, employ every alternate Saturday as a
working day.
Increase working hours per day Increase working hour per shift, or utilize overtime.
Increase subcontract level Increase the use of subcontractors to increase capacity
Increase the safety stock level Stock more finished products to avoid a backorder.
Step7: Determine the “SCOR KPIs-TIs” relationship scores.
The relationship between each pair of “SCOR KPIs-TIs” explains the degree of influence
that TIs can technically influence SCOR KPIs, and it could be both positive and
negative. In this step, the degree of influence between each pair is evaluated by the
decision maker in the related working area. Let the relationship value between the
performance indicator Wm, and the technical improvement Hn be mnr~ , and assume that
it follows the triangular fuzzy linguistic scale. Fig. 4.4 presents the general form of the
relationship matrix.
Figure 4.4: Relationship matrix between “Whats” and “Hows”
Ref. code: 25605422300342CMU
87
In the standard HOQ process for new product development, the determination of
“HOW-HOW” relationship is established since one TI may affect other TIs (Bevilacqua
et al., 2006, Mayyas et al., 2011) However, this dissertation excludes this step since it is
not necessary.
Step8: Obtain final ratings of the TIs.
We complete the HOQ process by calculating the final rating of the TIs ( nf~
), using the
final importance rating of SCOR KPIs ( *~mW ), and the relationship score ( mnr~ ) according
to Eq. (33). These variables are declared as the TFNs, and the final ratings of the TIs are
shown at the base of the matrix where it is the main outputs of the proposed
methodology.
mnmnnn rWrWrWM
f ~~~~~~1~ *
2
*
21
*
1 , n (33)
There are a lot of techniques to defuzzify the fuzzy numbers, such as α-cut, centroid
method, and hamming distance. However the most popular and simplest method is the
centroid method (Chou and Chang, 2008). The defuzzified value of nf~
, denoted by DF
nf
, is calculated by Eq. (34). Then, DF
nf is normalized as NORM
nf on a 0-1 scale by Eq. (35).
The resulting NORM
nf is ranked in descending order to determine the priority of the TIs.
3213
1nnn
DF
n ffff , n (34)
DF
nn
DF
nn
DF
nn
DF
nNORM
nff
fff
minmax
min
, n (35)
Finally, the management team selects only a set of high priority TIs for implementation.
After they are implemented, the new SCOR KPIs are re-evaluated to compare with the
target SCOR KPIs. If they are significantly different, a new cycle of the proposed
Ref. code: 25605422300342CMU
88
method is repeated. A case study will be used to demonstrate how the proposed
methodology is applied in practice.
4.2 Data collection and case study
For the continuity of the study, this dissertation uses the same case study of the
bottle water factory to demonstrate the effectiveness of the method. However, during
the development of the fuzzy QFD process, this dissertation applied the current
situation with the data collected from the beginning to formulate the MILP model as a
predictive model, and then identify the possibility for technical improvements in the
proposed methodology to test what actions that the proposed methodology has
recommended before actually apply to improve the operation in the factory. The data
for constructing the MILP model is presented accordingly. Firstly, The company
produces 2 sizes (i1 = 1500cc, i2 =600 cc) of drinking water in bottles. The structure of the
manufacturing process is presented in Fig. 4.5.
Figure 4.5: Current production process of case study
In the previous situation, the manufacturing process is a flow shop with 2 stages (K=2),
which are a bottle blowing process and a water filling process. The company orders raw
material as a PET plastic resin according to the material requirement planning (MRP) at
an amount of 2 tonnes per lot to produce bottles. The amount of PET in grams to produce
each size of the bottle is τ1 = 4.17 and τ2 = 1.58, respectively. Currently, there are two
blowing machines for producing the bottles (n1 = 2), where each machine has a capacity
of 1,600 bottles per hour (C1 = 1,600). Empty bottles are temporarily stored in a work in
process area, waiting to transfer to the fill line. The water filling line is operated by a
conveyor system where the empty bottles are conveyed to wash, filled with water,
covered with a cap, sealed, inspected, and wrapped into bundles, and finally transferred
Ref. code: 25605422300342CMU
89
to a warehouse area. There is only one fill line (n2=1) with the capacity of 2,400 bottles
per hour (C2=2,400). The company works 16 hours per day (γ=16). Currently, 13 workers
are involved in the production (Wt = 13). Each unit of bottles requires on average 0.05
man-hour (ei=0.05) and the employees work two shifts per day (ht=2), at 8 hours/shift from
Monday to Friday (δ=8). The labor cost (cr) is 300 Baht/day. The company is now
subcontracting for extra capacity on average at 30% of the current demand. The cost
structure, inventory holding policy, options to increase and decrease capacity, and total
asset values will be discussed next.
4.2.1 Cost structure and inventory holding policy.
The finished products are sold in packs, which are 6 bottles per pack for 1,500 cc (ρ1=6),
and 12 bottles per pack for 600 cc (ρ2=12). Estimated demand per day is 805 bottles per
day for 1,500 cc (D1t = 805), and 3,198 bottles per day for 600 cc (D2t = 3,198). The selling
price (Ri) is 40 Baht/pack for both products. Table 4.3 shows the related operating costs.
The unit for all cost is Baht/pack except the finished product and WIP inventory holding
cost that is Baht/pack/period, and Baht/bag/period, respectively. The raw material
inventory holding cost (cl) is 20 Baht/tonne/period. The company’s inventory holding
policy is shown in Table 4.4. The WIP in between the process is stored and transferred
in bags, which are 380 bottles per bag for 1,500 cc (ϴ1=380), and 720 bottles per bag for
600 cc (ϴ2=720),
Table 4.3: Operating cost information
Bottle
(CC) Material
Cost
cmi
Overhead
Cost
cui
Subcontracting
Cost
csi
Backorder
Cost
cbi
FP
Inventory
holding
Cost
cii
WIP
Inventory
holding
Cost
cji
1,500 21.68 3 35.96 7 0.72 0.8
600 23.2 4 37.13 9 0.8 0.833
Ref. code: 25605422300342CMU
90
Table 4.4: Inventory holding policy
Inventory Maximum inventory limit Minimum inventory limit
1,500 cc 600 cc 1,500 cc 600 cc
WIP (bottles) 19,000 ( tJ1 ) 50,000 ( tJ 2 ) 0 ( tJ1 ) 15,000 (
tJ 2 )
Finish goods
(packs) 2,500 ( tI1 ) 5,000 ( tI2 ) 375( tI1 ) 600 ( tI2 )
Raw material 0.82 tonnes ( tM ) 0.13 tonnes ( tM )
4.2.2 Options to increase and decrease the production capacity
Based on the data collected and the available production resources, the management
team has agreed that the factory can perform the tasks listed in Tables 4.5 and 4.6 to
increase or decrease the production capacity for determining the SCOR agility
measures.
Table 4.5: Options to increase production capacity
Resources Actions Lead Time Investment (Baht) Machine Invest in two more blowing
machines
Up to 4
months 1,000,000/machine
Invest in one more fill line Up to 4
months 2,000,000/fill line
Workforce Hire four more skilled workers
for production. 15 days 300
Baht/worker/day
Work day Use every other Saturday as a
working day
10 days to
inform -
Working
hours
Increase working hours per shift
from 8 to 12
10 days to
inform -
Subcontract Increase subcontracting level by
10% from the current one
21 days -
Safety
Stock
Keep stock of finished products
25 % more than the current level
21 days. -
Ref. code: 25605422300342CMU
91
Table 4.6 : Options to decrease production capacity
Resources Actions Lead Time Investment
(Baht) Workforce The company can move three skilled
workers to other activities in the factory
15 days -
No. of
Shifts
Reduce number of shift to only 1 shift
per day. 15 days -
Subcontract Reduce subcontract level up to 10% of the
initial demand
21 days -
4.2.3 Current fixed assets.
From the collected data, the company can estimate total fixed assets as shown in Table
4.7.
Table 4.7: Estimated company total fixed assets
Fixed Assets Value (Baht) 1. Land 15,000,000
2. Factory, office, and warehouse 5,000,000
3. Two blowing machines at depreciation cost 1,200,000
4. One fill line at depreciation cost 800,000
5. Miscellaneous cost 600,000
The production system has to be modelled realistically to assess the existing SCOR
KPIs accurately. We employ the MILP model to represent the process and simulate the
operation of the system. The definitions of parameters and variables, objective
functions, and constraints, and method to assess the SCOR KPIs are discussed in
Chapter 3.
4.2.4 Opinions of Decision Makers (DMs)
In this dissertation, the company forms a team of managers or decision makers from
three departments including production (DM1), sales and marketing (DM2), and
accounting (DM3). These 3 DMs represent most of important activities in the company.
Ref. code: 25605422300342CMU
92
Table 4.8 shows the level of importance or weight of “WHATs” or SCOR KPIs,
expressed using the fuzzy linguistic scale.
Table 4.8 : DMs’ relative importance on SCOR KPIs
SCOR KPIs DM1 DM2 DM3
Increase make order fulfillment (W1) M H VH
Decrease make cycle time (W2) H M M
Decrease upside make flexibility (W3) H L H
Increase upside make adaptability (W4) H M H
Increase downsize make adaptability (W5) VH L M
Decrease cost to make (W6) H H VH
Decrease inventory days of supply (W7) H VH VH
Increase return on make fixed assets (W8) M L M
Increase return on make working capital (W9) H M H
The technical improvement actions (TIs) for the production system as well as the
relationship between SCOR KPIs and TIs are determined by the production manager.
Table 4.9 shows the list of TIs that can be implemented in the factory, and Table 4.10
exhibits the relationship matrix between SCOR KPIs and TIs. By following the
proposed methodology, the results of the case study are discussed in the next section.
Ref. code: 25605422300342CMU
93
Table 4.9: :List of corresponding TIs and their implementation lead time
Technical Improvements actions
(TIs) Description Lead Time
Increase production capacity (H1) Increase more production capacity by
investing in: -two more blowing machines
-one more fill line
Up to 4
months
Increase production workforce (H2)
Increase by four more skilled workers 15 days
Increase working days per week (H3)
Use every other Saturday as a working day 10 days
Increase working hours per day (H4)
Increase the shift hours from 8 hours/day to
12 hours/day
10 days
Increase subcontract level (H5) Increase 10% additional from current
subcontract level
21 days
Increase the safety stock level (H6)
Stock 25% finished products more from
current level
21 days
Table4.10 : Degree of influence of TIs on SCOR KPIs
H1 H2 H3 H4 H5 H6
W1 + VH + H + H + H + M + VH
W2 + VH + H + M + M + L + M
W3 + L + M + M + H + M + H
W4 + H + H + H + M + M + M
W5 + VL - H + VH + H + VH - H
W6 + L - M - M - VH - H + M
W7 + L + L + L + L + VL - H
W8 - H + M + L + L + H + VL
W9 + H + M + L + L + H - L
4.3 Results and discussions.
The proposed methodology is applied to the case study and the results are presented in
three parts according to the study scope. This consists of the assessment of current
SCOR KPIs, the suggested TIs as a result of the fuzzy QFD approach, and the
evaluation of the new SCOR KPIs after the TIs are implemented.
Ref. code: 25605422300342CMU
94
4.3.1 Current SCOR KPIs of the company.
Based on the provided data, the MILP model, and the procedure presented in
Appendices B and C, the current SCOR KPIs of the company are shown in Table 4.11,
and presented graphically as a spider diagram in Fig 4.6.
Table 4.11: Current SCOR KPIs of the company
SCOR-Make level 2
performance indicators.
Unit per scale in the spider
diagram
Make order fulfillment (W1) 85.0 % 0, 80, 85, 90, 95, 100
Make cycle time (W2) 10.4 hours 10, 8, 6, 4, 2, 0
Upside make flexibility (W3) 180 days 200, 160, 120, 80, 40, 0
Upside make adaptability (W4) 0 % 0, 40, 80, 120, 160, 200
Downsize make adaptability
(W5)
33.0 % 0 ,10, 20, 30, 40, 50
Cost to make (W6) 81.2 % 90, 85, 80, 85, 70, 0
Inventory days of supply (W7) 16.0 days 25, 20, 15, 10, 5, 0
Return on make fixed assets
(W8)
34.8 % 0, 20, 25, 30, 35, 40
Return on make working capital
(W9)
25.0 % 0, 20, 25, 30, 35, 40
Ref. code: 25605422300342CMU
95
Figure 4.6 : Graphical representation of current SCOR KPIs
Currently, the company can produce the products on time only 85%. The make cycle
time to setup and produce all products in one cycle (a day) is 10.41 hours. If the demand
is suddenly rising, the company cannot response in a short time due to inadequate
capacity. The company has to invest in more machines, which take up to 6 months and
this makes their upside make adaptability equal to 0%. However, it is possible to reduce
a production capacity by 33% without additional cost or inventory penalty. Total cost to
make, calculated as a percentage of total revenue, is 81%. This dissertation considers
only make processes. Thus, the cash-to-cash cycle time is equivalent to the inventory
days of supply, which is 16 days on average. Return on fixed assets and return on
working capital are estimated at 34.83% and 25%, respectively.
Table 4.12 shows the revenue-cost structure, obtained from the outputs of the MILP
model. A high proportion of the cost contributes to the subcontract, other operating
expenses, and the back-ordering cost. This figure also supports the statement that the
company is now encountering capacity problems.
Ref. code: 25605422300342CMU
96
Table 4.12 : Current revenue-cost structure of the company.
Total Revenue (Baht) 7,048,142.93
Cost to make (Baht) 5,019,810.14
1) Production 3,647,299.23
2) FP Inventory cost 2,997.54
3) Backorder cost 419,047.63
4) Subcontract cost 780,124.09
5) Labor cost 167,200.00
6) WIP Inventory 2,153.58
7) Raw material
inventory
988.07
Gross Profit 2,028,332.80
Sales and admin cost (10% of rev)
704,814.29
Net Profit 1,323,518.50
Cost to make
(% of total revenue) 71.20%
4.3.2 Selection of high priority TIs to improve SCOR KPIs
The relative importance of SCOR KPIs from three DMs, presented in Table 4.8, is
converted to TFNs. Next, Equation 29 is used to calculate the average relative
importance score. A competitive analysis is then conducted to compare the current
performance to the best practices, and identify the priority for improvement using the
fuzzy linguistic scale. The final importance rating is computed by Eq. (31), and the TFNs
are de-fuzzified and normalized. The SCOR KPIs are ranked as shown in Table 4.13.
Ref. code: 25605422300342CMU
97
Table 4.13: Derivation of the average importance rating, competitive analysis, and
final importance rating.
WHATs
SCOR KPIs
Average
Rating
Competitive Analysis Final Importance rating
mg~ Wm zm am
mu~ mu~
(TFNs)
mmm ugW ~~~ * DF
mW *
NORM
mW *
Rank
Increase make order fulfillment (W1)
(3.6, 4.2, 4.8) 85% 95% 90% H (6, 7, 8) (21.6, 29.4, 38.4) 29.8 0.800 4
Decrease make cycle time (W2)
(2.8, 3.4, 4.0) 25 16 20 H (6, 7, 8) (16.8, 23.8, 32.0) 24.2 0.567 6
Increase upside make flexibility (W3)
(2.8, 3.4, 4.0) 180 10.00 10.00 VH (8, 9, 10) (22.4, 30.6, 40.0) 31 0.850 3
Increase upside make adaptability (W4)
(3.2, 3.8, 4.4) 0 50% 30% VH (8, 9, 10) (25.6,34.2,44.0) 34.6 1.000 1
Increase downsize make adaptability (W5)
(2.8, 3.4, 4.0) 33% 50% 33% L (0, 1, 2) (5.6, 10.2, 16.0) 10.6 0.000 9
Decrease cost to make (W6)
(4.0 ,4.6, 5.2) 81% 70% 75% H (6, 7, 8) (24.0 ,32.2, 41.6) 32.6 0.917 2
Decrease inventory days of supply (W7)
(4.4 ,5.0 ,5.6) 16.00 15.00 15.00 M (4, 5, 6) (17.6, 25.0 ,33.6) 25.4 0.617 5
Increase return on make fixed assets (W8)
(2.0 ,2.6, 3.2) 0.10 0.10 0.10 M (4, 5, 6) (8.0, 13.0, 19.2) 13.4 0.117 8
Increase return on make working capital (W9)
(3.2 ,3.8, 4.4) 0.25 0.20 0.25 M (4, 5, 6) (12.8, 19.0, 26.4) 19.4 0.367 7
The top three SCOR KPIs that should be improved with priority suggested by the QFD
methodology are W3, to lower the number of the day for upside make flexibility, W4, to
increase the capability for upside make adaptability, and W6, to reduce the total cost to
make. By following the proposed approach after the identification of the “SCOR KPIs-
TIs” relationship, the dissertation complete the HOQ process by computing the final
rating of TIs ( nf~
) by the final importance of SCOR KPIs ( *~mW ) and the relationship score
( mnr~ ), according to Eq. (35). Table 4.14 depicts the relationship matrix as the TFNs based
on the opinions of the expert team of the company as shown by the linguistic scale in
Table 4.10. The final rating nf~
is defuzzified in crisp values, normalized, and presented
in Table 4.15.
Ref. code: 25605422300342CMU
98
Table 4.14 : Relative importance (weight) of WHATS ( *~mW ) and the relationship score (
mnr~ )
SCOR
KPIs
*~mW 1
~mr 2
~mr 3
~mr 4
~mr 5
~mr 6
~mr
W1 (21.6, 29.4,
38.4) (8,9,10) (6,7,8) (6,7,8) (6,7,8) (4,5,6) (8,9,10) W2 (16.8, 23.8,
32.0) (8,9,10) (6,7,8) (6,7,8) (4,5,6) (2,3,4) (4,5,6) W3 (22.4, 30.6,
40.0) (2,3,4) (4,5,6) (4,5,6) (6,7,8) (4,5,6) (6,7,8) W4 (25.6,34.2,44.0) (6,7,8) (6,7,8) (6,7,8) (4,5,6) (4,5,6) (4,5,6) W5
(5.6, 10.2, 16.0) (0,1,2) (-8,-7,-6) (8,9,10) (6,7,8) (8,9,10)
(-8,-7,-6)
W6 (24.0 ,32.2,
41.6) (2,3,4) (-6,-5,-4)
(-6,-5,-4)
(-10,-9,-8)
(-8,-7,-6) (4,5,6)
W7 (17.6, 25.0
,33.6) (2,3,4) (2,3,4) (2,3,4) (2,3,4) (0,1,2) (-8,-7,-6)
W8
(8.0, 13.0, 19.2) (-8,-7,-
6) (4,5,6) (2,3,4) (2,3,4) (6,7,8) (0,1,2) W9 (12.8, 19.0,
26.4) (6,7,8) (4,5,6) (2,3,4) (2,3,4) (6,7,8) (-4,-3,-2)
Table 4.15 : Final rating and ranking of TIs
The results from the proposed methodology suggest a priority list of TIs to achieve the
desired performance level. In this case, the top three TIs, which have significantly
higher values of NORM
nf than other TIs, include H1, to increase more production
capacity, H2, to increase the working days per week, and H3, to increase the skilled
H1 H 2 H 3 H 4 H 5 H 6
nf~
(66.85,
114.87,
182.76)
(44.8,
85.27,
144.54)
(50.14,
96.29,
162.85)
(33.78,
73.63,
132.8)
(32.18,
73.09,
133.87)
(37.34,
71.05,
121.96) DF
nf 121.49 91.54 103.09 80.07 79.71 76.78
NORM
nf 1.0000 0.3300 0.5885 0.0735 0.0655 0.0000
Ranking 1 3 2 4 5 6
Ref. code: 25605422300342CMU
99
workforce. Therefore, these three TIs should be selected for implementation. The
dissertation determines the new SCOR KPIs after implementation of the selected TIs
using the predictive MILP model presented in section 3.2 of Chapter 3. The related
parameters of the MILP model are modified to represent the selected TIs (H1, H2, and
H3). The model is then solved for the optimal solution and the new SCOR KPIs are
determined using formulas and steps presented in section 3.3 of Chapter 3. The new
SCOR KPIs are shown in Table 4.16 and illustrated as a spider diagram in Fig. 4.7.
Table 4.16 : The new SCOR KPIs after improvement
SCOR-Make level 2 performance
indicators.
Unit per scale in the spider
diagram
Make order fulfillment (W1) 100.0 % 0, 80, 85, 90, 95, 100
Make cycle time (W2) 7.3 hours 10, 8, 6, 4, 2, 0
Upside make flexibility (W3) 7 days 200, 160, 120, 80, 40, 0
Upside make adaptability (W4) 131.0 % 0, 40, 80, 120, 160, 200
Downsize make adaptability (W5) 32.0 % 0 ,10, 20, 30, 40, 50
Cost to make (W6) 79.0 % 90, 85, 80, 85, 70, 0
Inventory days of supply (W7) 17.0 days 25, 20, 15, 10, 5, 0
Return on make fixed assets (W8) 31.2 % 0, 20, 25, 30, 35, 40
Return on make working capital (W9) 26.0 % 0, 20, 25, 30, 35, 40
Ref. code: 25605422300342CMU
100
Figure 4.7 : Graphical representation of new SCOR KPIs
When the factory increases production resources, including machines, labor, and
working time, apparently most of the capacity issues are solved. Therefore, the MILP
model which does not consider uncertainties results in 100% make order fulfillment.
However, in reality where uncertainties exist, it will prohibit the company from
achieving 100% fulfillment. The results from the model also illustrate that the company
can improve their upside make flexibility to 7 days, and the upside make adaptability
to 131%. Regarding the total cost to make, if the demand is stable, the investment in
machines and other production assets lead to a cost reduction from 81% to 79%.
However, more demand may leverage the asset utilization and increase profit. As a
result, the cost to make may be further reduced if the demand is increased due to sales
growth in the future. Table 4.17 exhibits the new cost structure after implementation of
the selected TIs. More proportion of the cost is allocated to the production while the
subcontracting and the backordering cost are mostly eliminated.
Ref. code: 25605422300342CMU
101
Table 4.17 : The total Revenue-Cost structure obtained from LP model after
performance improvement.
Based on the results from the MILP model, it can be seen that many KPIs after
improvement are quite close to the target KPIs, and some KPIs are better than the
targets. This finding encourages the management team to really implement the
suggested TIs and measure real KPIs after the implementation. However, there are two
main reasons that the real KPIs after implementing the TIs may be significantly
different from the target KPIs. First, the MILP (predictive) model is validated (tuned)
only under the current situation. Thus, it may not be totally accurate to predict the KPIs
under different situations when the TIs are implemented. It is suggested that the MILP
model should be validated under various situations dependent on the TIs. Second, the
relationship between SCOR KPIs and TIs, which is obtained from opinions of the
management team (in step 7 of the fuzzy QFD), may be inaccurate. This data (or
knowledge) should be updated if the management team learns that a more accurate one
is available based on real experiences after implementation of TIs for some periods.
Total Revenue (Baht) 7,688,883.20
Cost to make (Baht) 5,321,124.22
1) Production 5,131,069.66
2) FP Inventory cost 3,513.64
3) Backorder cost
4) Subcontract cost 14.98
5) Labor cost 184,320.00
6) WIP Inventory 1,239.69
7) Raw material inventory 966.24
Gross Profit 2,367,758.98
Sales and admin cost (10% of rev) 768,888.32
Net Profit 1,598,870.66
Cost to make
(% of total revenue) 69.2%
Ref. code: 25605422300342CMU
102
4.3.3 Relationships between results in chapters 3 and 4
In this chapter, there are two situations of the SCOR KPIs, namely, the current SCOR
KPIs, and the new SCOR KPIs after the implementation of TIs. As this research has
been conducted for about 6 years, and the dissertation wants the reader to perceive the
current KPIs of the company which is not good at the beginning, so the researcher uses
the data collected in 2012 and applies the predictive model to determine the current
SCOR KPIs as shown in figure 4.6. The aim of this chapter is to propose the fuzzy QFD
methodology to suggest the TIs for performance improvement. Hence, the dissertation
apply the selected lists of TIs to the predictive model to predict the new SCOR KPIs,
as shown in figure 4.7. Nevertheless, there are relationships between the results in
chapters 3 and 4. In 2014, there was a real decision in the factory to invest in more
machines and hire more workforce to increase the production capacity. Therefore, the
data was recollected in 2014, and the predictive model was used to determine the SCOR
KPIs in chapter 3, as shown in figure 3.7. To compare the SCOR KPIs in figure 3.7,
which involve the real investment, and those in figure 4.7, which involve the suggested
TIs, it can be seen that the SCOR KPIs are not the same. This is because the decisions
are not the same. While the real action in the factory in chapter 3 was to install more
machines and hire more workforces, the proposed methodology in chapter 4 suggests
to install more machines, increase the workdays per week, and to hire more skilled
workers.
4.4 Concluding remarks
The SCOR model is a process reference model that is widely recognized in the field of
supply chain management to improve the overall business process and performance of
a company. However, when the SCOR KPIs are used to measure the performance, only
the current operational results are disclosed. Managing the SCOR KPIs is not an easy
task because they involve many metrics and are interrelated. Therefore, it requires a
suitable methodology to lead the organization to the proper direction for performance
improvement. This chapter presents a new approach to manage the SCOR KPIs for
Ref. code: 25605422300342CMU
103
performance improvement by employing the fuzzy QFD methodology. Quality function
deployment (QFD) is generally an efficient tool for a product design process that can
efficiently translate the customer needs into design requirements. In this chapter, the
QFD process with the use of House of Quality (HOQ) planning matrix is combined with
the SCOR KPIs, which are defined as “WHATs” while the technical improvement
actions (TIs) are defined as “HOWs”. The eight-step fuzzy QFD approach to manage the
SCOR KPIs is proposed. The TFNs are used to handle the uncertainty of the human
opinions. The application of the method is demonstrated by a case study of a make-to-
stock, bottled water manufacturer. The predictive method (including the MILP model
and steps for determining SCOR agility measures) is used to represent the existing
manufacturing system, to evaluate current SCOR KPIs, and to predict the new SCOR
KPIs, after the TIs are implemented.
The proposed methodology provides theoretical and practical contributions to
the field of supply chain management and the QFD application as follows:
1. The proposed QFD approach to manage SCOR KPIs for performance
improvement in this paper is new and original.
2. The proposed approach is capable of prioritizing the TIs. When the
management team selects some high priority TIs to be implemented, the new SCOR
KPIs will be changed toward the target SCOR KPIs.
3. Although the typical QFD framework is adopted, this dissertation develops a
new method for determining the improvement rating of the target SCOR KPIs using
opinions of the management team following the linguistic scale.
4. This dissertation uses a real industrial case study to demonstrate that when the
proposed approach is applied, the SCOR KPIs are managed in the desired direction.
The proposed methodology has some limitations as follows. First, the
management team should be able to identify the list of TIs (Hows) that can be really
Ref. code: 25605422300342CMU
104
implemented in the factory, and they should significantly affect the SCOR KPIs
(Whats). Second, if the management team expresses the relationship between “Whats”
and “Hows” incorrectly or without enough knowledge to make a judgment, it will
consequently affect the priority list of TIs to be implemented, and the company may
not get the KPIs after improvement according to the target. Nevertheless, this situation
can be regarded as a learning process where the management team can revise the
relationship between “Whats” and “Hows”, based on experiences of implementing the
TIs. This cycle can be repeated as a continuous improvement process following the
block diagram in Fig. 4.1
The recommendations for further studies are as follows. First, a Kano analysis,
which is a tool to prioritize the customer requirements for product design purpose, may
be further adapted and integrated with the proposed fuzzy QFD approach to determine
the relative importance of the SCOR KPIs, and choose the desired TIs that are suitable
for the situation in the factory. Secondly, the dissertation only applies the proposed
methodology to a single case study, and found that the methodology can be effectively
applied to get a satisfactory result using the predictive (MILP) model. Therefore, it
should be practically applied in various situations of many industries. Lastly, our
dissertation only focuses on the SCOR-make process, so the SCOR KPIs and the list of
TIs are only derived based on the manufacturing situation. Further studies can be
extended to cover SCOR-plan, source, deliver, and return processes in supply chain
management.
Ref. code: 25605422300342CMU
105
Chapter 5
Conclusions
The objective of this chapter is to recap what have been completed in the PhD
dissertation. First, the summary of the research from the beginning are discussed,
following by the findings of each research chapter. Next, the theoretical and practical
knowledge contributions which is an original contribution of this PhD dissertation are
presented. Lastly, the limitation and recommendation for further study is addressed for
the benefits of fellow researchers who is interesting to continue the research.
5.1 Summary of the research
This dissertation aims to focus on the topic of performance evaluation and improvement
in the supply chain system. Based on the literature study, there were a lot of articles that
have proposed the supply chain performance as a metric design or a requirement to
compose a good performance measurement system, however the consideration of the
evaluation method such as the mechanism for assessment is still limited. The
dissertation realizes that a good performance measurement system should be able to tell
the company on the improvement areas and the decision to move on rather than just
monitoring. This becomes the research interest of this dissertation.
From the first chapter, we started from the identification of the performance
measurement system (PMS) and their role in the supply chain management, and then the
subject of various PMSs that are available in the context of supply chain management
are introduced. The pros and cons of each PMS technique is discussed, and this
dissertation selects The Supply Chain Operations Reference (SCOR) Model as a
framework to use throughout the dissertation. The SCOR model consists of an explicit
supply chain processes, with the standard definitions of each process, the KPIs are
classified based on the attributes and metrics, and there are the best practices where the
company can follow to achieve the good performances. The advantage of using the
SCOR model in comparison to others performance measurement system model is, the
model creates the same language that can be used thoroughly in any supply chain
Ref. code: 25605422300342CMU
106
members, and the model makes it feasible for organizations to determine and compare
a performance of a supply chain process within their organizations and against others.
Next, to focus on the issue of method of evaluation and improvement technique, the
research problems are epitomized. With the use of SCOR model, this dissertation
summarizes the research problems as followed;
1. The SCOR model is just a reference model. It lacks of the interrelationship
between the model and the system under studied, as well as the procedural
methodology for the performance evaluation.
2. There are the agility measures in the model. These KPIs are difficult to evaluate
without a mechanism and some models.
3. As the company cannot be the best in all SCOR KPIs, there should be a
methodology for the company to trade off among the improved KPIs that can
satisfy the need of the organization.
4. There is a complex interrelationship between the variable in the SCOR KPIs
and the system under studied, so the management of KPIs for improvement
needs a detailed methodology to determine the direction of improvement.
From the above research problems, the literature reviews of the related subjects are
conducted in Chapter2, where the research problems (1) and (2) are addressed by the new
methodology contribution to evaluate the SCOR KPIs by using a predictive MILP
model with fuzzy parameters in Chapter 3. Then, problems (3) to (4) are resolved by the
new approach to manage the SCOR KPIs by fuzzy QFD approach in Chapter 4. Lastly,
the summary of this dissertation and the overall contributions are stated in this Chapter.
In Chapter 2, this dissertation firstly provides the comprehensive review of the
SCOR model and their application in performance measurement to support the
dissertation of why the SCOR model is appropriated to use as a supply chain
performance evaluation framework in our dissertation. Next, the reviews of the MILP
model in the supply chain planning problem is discussed. Based on the previous
implication of the MILP model, it is found that the model is suitable for the supply
Ref. code: 25605422300342CMU
107
chain planning in the tactical decision level whereby the optimization technique stand
out as the suitable approach to solve the MILP model. With the characteristic of the
problem in this dissertation, so the dissertation applies the MILP model to use as the
predictive model in this research. To manage the issue of supply chain uncertainty, the
types of uncertainty are firstly introduced, then we present the pros and cons of each
technique to manage uncertainties before the Fuzzy Set Theory (FST) are selected as the
method to combine with the MILP model and fuzzy parameters that is used as a
predictive model to evaluate the SCOR KPIs, and to predict the future performance in
many what-if scenarios for performance improvement. The last section of the literature
reviews described about the Quality Function Deployment (QFD), and the fuzzy QFD
approach which is a tool in the quality management. The QFD approach is popular in
transforming the need of the customer, and translate it into the technical requirement,
so that the end products meet the customer expectation, so our research aims to employ
the fuzzy QFD approach, to combine with the SCOR model to manage the SCOR KPIs
for performance improvement that are satisfying with the need of the organization.
Nevertheless, as the SCOR model consists of many processes, so to be able to
demonstrate the practicality of the proposed method efficiently, the research is scoped
down to focus on the Make process of the SCOR model, while the other processes are
exempted. In this PhD dissertation, the Make process is selected as the preliminary
study, specify the manufacturing system to be studied, develop the proposed
methodology for performance evaluation and improvement, and demonstrate the
practicality of the method based on the selected Make-to-stock process. In Chapter 3,
the dissertation has reviewed from the previous academic papers that the SCOR model
is a widely-recognized model that has been successfully implemented to improve the
business in many industries. So, the objective of this chapter is to resolve the research
problems (1) and (2) by proposing a method to evaluate the SCOR KPIs based on the
predictive model. It consists of the MILP model that is used to represent the system
under studied, and a methodology to evaluate the SCOR KPIs based on the level 2 of
the SCOR-Make process, and some algorithms to assess the agility-related measures.
Ref. code: 25605422300342CMU
108
The Triangular Fuzzy Numbers (TFNs) with a specific crisp set are used to represent
uncertainties in the supply chain system. A case study of a bottled-water factory is
conducted to demonstrate the application of the method, and the SCOR KPIs outputs
are illustrated by the spider diagram. The findings of this chapter indicate that the
proposed methodology is capable of developing the relationship between the
manufacturing parameters and the SCOR KPIs, which enable the effective prediction
process especially when the manufacturing parameters are changed or improved.
From chapter 3, even now it is possible to evaluate the SCOR KPIs outputs
based on the proposed methodology, but only the current performance outputs are
disclosed. Managing the SCOR KPIs for performance improvement to meet with the
requirement of organization is still a challenging task as the SCOR model composes of
many metrics, and it has a wide range to improve without a structural methodology. The
aim of this chapter is to propose a new approach to manage the SCOR KPIs for
performance improvement by utilizing the fuzzy QFD methodology that would be able
to guide the organization to work on the preferred direction of the performance
improvement. In this chapter, the eight-step fuzzy QFD approach to manage the SCOR
KPIs is proposed. The SCOR KPIs are defined as “WHATs”, while the technical
improvement actions (TIs), which is the list of actions that can be implemented in the
factory to improve the performance of the organization, are defined as “HOWs”. The
TFNs are used to handle the linguistic judgement of the human opinions, and transform
into the mathematical values that can be managed. The application of the method is
well demonstrated by a similar case study of a make-to-stock, bottled water
manufacturer. For a comparison purpose between the current and the improved
performance, the predictive model and the proposed methodology in Chapter 3 that is
used to evaluate the SCOR KPIs are applied to assess current performance and to
predict the new SCOR KPIs, after the TIs are implemented. The spider diagram with a
scalar value derived from the acceptance level of the organization are used to portrait
the SCOR KPIs of before and after performance assessment. The findings of the chapter
reveal that the proposed methodology is capable to assist the company when they have
Ref. code: 25605422300342CMU
109
a lot of performance criteria to be improved, and the SCOR model can be well-
integrated with the QFD method to manage the performance as demonstrated by the
case study.
Overall, the dissertation resolved all of the research problems and
accomplished all of the objectives as stated in this PhD research. The dissertation
provides a knowledge contribution which is new and original to the field of supply
chain management as summarized in the next section.
5.2 Key Contributions of the dissertation.
The contributions are summarized as follows.
1. The proposed methodology to evaluate the SCOR KPIs based on the predictive
MILP model with fuzzy parameters is new and original in the field of supply chain
management.
2. The proposed approach is capable for determining the relationship between the
SCOR KPIs, and the system under studied which is the manufacturing parameters.
So, when the SCOR KPIs and the production system are related, it enables the
performance prediction process when the manufacturing parameters are changed or
improved.
3. The proposed methodology consists of a procedural method and a supportive
model, which is the predictive model, to evaluate the measurement of supply chain
agility of upside make flexibility, upside make adaptability, and downsize make
adaptability in the SCOR metrics. Without the model and some methods, the
measurement of agility KPIs is almost impossible.
4. The spider diagram, which is developed in this dissertation with a scale that
represents satisfactions of DMs is useful since it is not only used to portrait the
Ref. code: 25605422300342CMU
110
KPIs, but it can be used for a comparison purpose of the performance before and
after improvement.
5. The proposed methodology of using the fuzzy QFD approach to manage the SCOR
KPIs for performance improvement in this dissertation is also new and original.
6. The proposed approach of using the fuzzy QFD approach is capable of prioritizing
the technical improvement actions (TIs). So, when the management team selects
some high priority TIs to be implemented, the new SCOR KPIs will be changed
toward the target SCOR KPIs.
7. In the proposed fuzzy QFD methodology, this dissertation develops a new method
for determining the improvement rating of the target SCOR KPIs by using opinions
of the management team which reflect directly to the need of organization for
performance improvement.
8. A real industrial case study is used to demonstrate that the SCOR KPIs of the
company can be efficiently evaluated based on the proposed approach in Chapter 3,
and when the similar case study is applied in Chapter 4, the SCOR KPIs can be
managed to obtain satisfactory KPIs.
Publications based on the results of this research are presented as follows.
Akkawuttiwanich, P. and Yenradee, P., 2017. Evaluation of SCOR KPIs using a
predictive MILP model under fuzzy parameters. International Journal of Supply Chain
Management, 6(1), 172-185.
Akkawuttiwanich, P. and Yenradee, P., (Under review). Fuzzy QFD Approach for
Managing SCOR Performance Indicators. Computers and Industrial Engineering.
Ref. code: 25605422300342CMU
111
5.3 Limitations, and recommendation for further study.
This dissertation still has some limitations that can be improved further. They are
summarized as follows.
1. This dissertation uses the MILP model to establish the predictive model, when the
characteristics of the manufacturing system are changed, the parameters and
constraints of the MILP models need to be adjusted to the particular case. This
method lacks of flexibility. For further research, we recommend that the researchers
should create a generalized model first by using more flexible modelling method
such as system simulation, and then customize the system to match with a case study
later.
2. The scaling of the spider diagram that is used to illustrate the SCOR KPIs is
obtained based on an opinion of the management team of the company. Thus, it
should be revised when applied to other companies. In this case, it is suggested that
some visualization technique such as R-statistical modelling can be applied to the
spider diagram to demonstrate a real-time performance comparison when the
manufacturing parameters are changed.
3. The current scope of this dissertation considers only the manufacturing aspect of
the SCOR-Make process as the exploratory study of using the proposed method to
evaluate and improve the performance, as well as to derive the list of TIs based on
the specific manufacturing situation. Therefore, further research can be extended to
cover the evaluation of other processes, namely, plan, source, deliver, return, and
enabler in a supply chain system.
4. Based on the proposed method of fuzzy QFD approach to manage the SCOR KPIs,
there are some limitations that the management team should have enough
knowledge that can be able to identify the list of TIs (Hows) that can be really
implemented in the factory, and they should significantly affect the SCOR KPIs
(Whats).
Ref. code: 25605422300342CMU
112
5. From the proposed method of using the fuzzy QFD approach, if the management
team expresses the relationship between “Whats” and “Hows” incorrectly or without
enough knowledge to make a judgment, it will consequently affect the priority list
of TIs to be implemented, and the company may not get the KPIs after improvement
according to the target. Nevertheless, this situation can be regarded as a learning
process where the management team can revise the relationship between “Whats”
and “Hows”, based on experiences of implementing the TIs. This cycle can be
repeated as a continuous improvement process following the block diagram in Fig.
4.1
6. A Kano analysis, which is a tool to prioritize the customer requirements for product
design purpose, may be further adapted and integrated with the proposed fuzzy QFD
approach to determine the relative importance of the SCOR KPIs, and choose the
desired TIs that are suitable for the situation in the factory.
7. This dissertation only applies the proposed methodology to a single case study, and
found that the methodology can be effectively applied to get a satisfactory result
using the predictive (MILP) model. Therefore, it should be practically applied in
various situations of many industries.
Ref. code: 25605422300342CMU
113
References
Agami, N., Saleh, M., Rasmy, M., 2014. An innovative fuzzy logic based approach for
supply chain performance management. IEEE System Journal, 8 (2), 336–342.
Akao, Y., 1990. Quality function deployment: Integrating customer requirements into
product design. Cambridge, MA: Springer
Akyuz, GA., and Erkan, TE., 2010. Supply chain performance measurement: a literature
review. International Journal of Production Research, 48, 5137–5155.
Aliev, R.A., Fazlollahi, B., Guirimov, B.G., and Aliev, R.R., 2007. Fuzzy-genetic
approach to aggregate production–distribution planning in supply chain
management. Information Sciences, 177, 4241–4255.
Aloma,M. and Pasek, Z.J., 2014. Linking Supply Chain Strategy and Processes to
Performance Improvement. Variety Management in Manufacturing. Proceedings of the 47th CIRP Conference on Manufacturing Systems, 17, 628
– 634.
Alonso-Ayuso, A., Escudero, L., Garín, A., Ortuño, M.T., and Pérez, G., 2003. An
Approach for Strategic Supply Chain Planning under Uncertainty based on
Stochastic 0-1 Programming. Journal of Global Optimization, 26(1), 97-124.
Ameknassi, L., Aït-kadi, D., & Rezg, N., 2016. Integration of logistics outsourcing
decisions in a green supply chain design: A stochastic multi-objective multi-
period multi-product programming model. International Journal of Production
Economics, 182, 165–184.
American Supplier Institute, 1994. Quality Function Deployment (Service QFD): 3-Day
Workshop. ASI Press, Dear- born, MI.
Amid, A., Ghodsypour, S. H., and O’Brien, C., 2006. Fuzzy multi-objective linear model
for supplier selection in a supply chain. International Journal of Production
Economics, 104(2), 394–407.
Amin,S.H., and Razmi,J., 2009. An integrated fuzzy model for supplier management: A
case studyof ISP selection and evaluation. Expert Systems with Applications,
36, 8639–8648.
Ref. code: 25605422300342CMU
114
Amirtaheri, O., Zandieh, M., Dorri, B., & Motameni, A. R., 2017. A bi-level
programming approach for production-distribution supply chain problem.
Computers & Industrial Engineering, 110, 527–537.
Ammar, S., and Wright, R., 2000. Applying fuzzy-set theory to performance evaluation.
Socio-Economic Planning Sciences, 34, 285–302.
APICS Supply Chain Council, 2016. The Supply Chain Operations Reference model
(SCOR) framework. http://www.apics.org/sites/apics-supply-chain-
council/frameworks/scor. Access online 23-11-2016
Arikan, F., and Gungor, Z., 2001. An application of fuzzy goal programming to a
multiobjective project network problem. Fuzzy Sets and Systems, 119, 49–58.
Ashayeri, J., Tuzkaya, G., and Tuzkaya, U.R., 2012. Supply chain partners and
configuration election: an intuitionistic fuzzy Choquet integral operator based
approach. Expert Systems with Applications, 39, 3642–3649.
Bai, C., and Sarkis, J., 2010. Integrating sustainability into supplier selection with grey
system and rough set methodologies. International Journal of Production
Economics, 124, 252–264.
Banomyong, R., and Supatn, N., 2011. Developing a supply chain performance tool for
SMEs in Thailand. Supply Chain Management: An International Journal, 16 (1),
20–31.
Barbarosoglu, G., and Ozgur, D., 1999. Hierarchical design of an integrated production
and 2-echelon distribution system. European Journal of Operational Research,
118, 464–484.
Beale, E.M.L. and Forrest, J.J.H., 1976. Global optimization using special ordered sets.
Mathematical Programming, 10(1), 52-69.
Beamon, B.M., 1999. Measuring supply chain performance. International Journal of
Operations and Production Management, 19 (3), 275–292.
Bevilacqua, M., Ciarapica, F. E., and Giacchetta, G., 2006. A fuzzy-QFD approach to
supplier selection. Journal of Purchasing and Supply Management, 12(1), 14–27.
Bhagwat, R., and Sharma, M.K., 2009. An application of the integrated AHP-PGP model
for performance measurement of supply chain management. Production
Planning & Control, 20(8), 678–690.
Ref. code: 25605422300342CMU
115
Bhattacharya, A., Geraghty, J., and Young, P., 2010. Supplier selection paradigm: An
integrated hierarchical QFD methodology under multiple-criteria environment. Applied Soft Computing, 10, 1013–1027.
Bicknell, B.A. and Bicknell, K.D., 1995. The Roadmap to Repeatable Success, CRC
Press, Boca Raton, FL.
Bilgen,B, 2010. Application of fuzzy mathematical programming approach to the
production allocation and distribution supply chain network problem. Expert
Systems with Applications, 37(6), 4488-4495.
Bilgen, B., and Ozkarahan, I., 2007. A mixed-integer linear programming model for bulk
grain blending and shipping. International Journal of Production Economics,
107, 555–571.
Bojadziev, G., and Bojadziev, M., 1995. Fuzzy sets, fuzzy logic, applications. World
Scientific Publishing Co. Pte Ltd.
Bolstorff, P. and Rosenbaum, R.G., 2003. Supply chain excellence–A handbook for
dramatic improvement using the SCOR model. New York: American
Management Association.
Bottani, E., 2009. A fuzzy QFD approach to achieve agility. International Journal of
Production Economics, 119(2), 380–391.
Botttani,E. and Rizzi,A., 2006. Strategic management of logistics service: A fuzzy QFD
approach. International Journal of Production Economics, 103, 585–599.
Bozdana, A., 2007. Quality Function Deployment.
http://sixsigma123.blogspot.com/2007/04/quality-function-deployment-
qfd.html. Access online 23-11-2016
Bredstrom, D., and Ronnqvist, M., 2002. Integrated production planning and route
scheduling in pulp mill industry. In: Proceedings of the 35th Annual Hawaii
International Conference on System Sciences, 2002, HICSS.
Burgess, K., and Singh, P. J., 2006. A proposed integrated framework for analyzing
supply chains. Supply Chain Management: An International Journal, 11(4), 337–
344.
Büyüközkan, G., Feyziohlu, O., and Ruan, D. 2004. Fuzzy group decision-making to
multiple preference formats in quality function deployment. Computer in
Industry, 58, 392–402.
Ref. code: 25605422300342CMU
116
Carlsson, C., and Fuller, R.,2002. A fuzzy approach to taming the bullwhip effect.
Advances In Computational Intelligence and Learning: Methods and
Applications International series in intelligent technologies, 18, 247–262.
Cavalieri, S., Gaiardelli, P., and Ierace, S., 2007. Aligning strategic profiles with
operational metrics in after-sales service. International Journal of Productivity
and Performance Management, 56 (5–6), 436–455.
Chan, F.T.S., 2003. Performance measurement in a supply chain. International Journal
of Advanced Manufacturing Technology, 21 (7), 534–548.
Chan, F. and Qi, H., 2003. An innovative performance measurement method for supply
chain management. Supply Chain Management: An International Journal, 8 (3),
209-23.
Chan,L.K., and Wu M.,L., 2005. A systematic approach to quality function deployment
with a full illustrative example. The international journal of management
science, Omega 33, 119-139.
Chanas, S., 1983. The use of parametric programming in fuzzy linear-programming,
Fuzzy Sets and Systems,11, 243–251.
Chanas, S., Delgado, M., Verdegay,J.L., and Vila, M.A.,1993. Interval and fuzzy
extensions of classical transportation problems, Transportation Planning and
Technology, 17, 203–218.
Charkha,P.G., and Jaju.,S.B., 2014. Supply chain performance measurement system: an
overview. International Journal of Business Performance and Supply Chain
Modelling. 6(1).
Chen, H., Amodeo, L., Chu, F., and Labadi, K. (2005). Modeling and Performance
Evaluation of Supply Chains Using Batch Deterministic and Stochastic Petri
Nets, IEEE Transactions on Automation Science and Engineering, 2(2), 132–
144.
Chen, S.P., and Chang, P.C.,2006. A mathematical programming approach to supply
chain models with fuzzy parameters, Engineering Optimization 38, 647–669.
Chen, L.H., and Ko, W.C., 2008. A fuzzy nonlinear model for quality function
deployment considering Kano’s concept. Mathematical and Computer
Modeling, 48, 581–593.
Chen, L.H., and Ko, W.C, 2009. Fuzzy approaches to quality function deployment for
new product design. Fuzzy Sets and Systems, 160(18), 2620–2639.
Ref. code: 25605422300342CMU
117
Chen, L.H., and Ko, W.C., 2011. Fuzzy linear programming models for NPD using a
four-phase QFD activity process based on the means-end chain concept.
European Journal of Operational Research, 201(2), 619–632.
Chen, M., and Wang, W., 1997. A linear programming model for integrated steel
production and distribution planning. International Journal of Operations and
Production Management, 17, 592–610.
Chen, L.H., Weng, M.C., 2003. A fuzzy model for exploiting quality function
deployment. Mathematical and Computer Modeling, 38, 559–570.
Chen, L. S., and Weng, M. C., 2006. An evaluation approach to engineering design in
QFD processes using fuzzy goal programming models. European Journal of
Operational Research, 172(1), 230–248.
Cheng, J.C.P., Law, K.H., Bjornsson, H., Jones, A., and Sriram, D., 2010. Modelling and
monitoring of construction supply chains. Advanced Engineering Informatics,
24, 435–455.
Chiang, W.Y.K., and Monahan, G.E., 2005. Managing inventories in a two-echelon dual-
channel supply chain. European Journal of Operational Research, 162(2), 325-
41.
Childerhouse, P. and Towill, D., 2000. Engineering supply chains to match customer
requirements. Logistics Information Management, 13 (6), 337–345.
Chou, S. Y., and Chang, Y. H., 2008. A decision support system for supplier selection
based on a strategy-aligned fuzzy SMART approach. Expert Systems with
Applications, 34(4), 2241–2253.
Clivillé, V., and Berrah, L., 2012. Overall performance measurement in a supply chain:
towards a supplier-prime manufacturer based model. Journal of Intelligent
Manufacturing, 23, 2459–2469.
Cohen, L., 1995. Quality function deployment: How to make QFD work for you. Reading, MA: Addison-Wesley.
Cristiano J, J.J., Liker J, J.K., White III, C.C., 2001a .Key factors in the successful
application of quality function deployment (QFD). IEEE Transactions on
Engineering Management, 48 (1), 81–95.
Crowder, H., Johnson, E.L., and Padberg, M., 1983. Solving large-scale zero-one linear
programming problems. Operations Research, 31, 803–834.
Ref. code: 25605422300342CMU
118
Davis, T., 1993. Effective supply chain management. Sloan Management Review, 34,
35-46.
De Toni, A., and Tonchia, S., 2001. Performance measurement systems: models,
characteristics and measures. International Journal of Operations and
Production Management, 21 (1–2), 46–70.
Diabat, A., & Theodorou, E., 2015. A location – inventory supply chain problem:
Reformulation and piecewise linearization. Computers and Industrial
Engineering, 90, 381–389.
Dhaenens-flipo, C., and Finke, G., 2001. An integrated model for an industrial
production–distribution problem. IIE Transactions, 33, 705–715.
Dogan, K., and Goetschalckx, M., 1999. A primal decomposition method for the
integrated design of multi-period production–distribution systems. IIE
Transactions ,31, 1027–1036.
Dong, J., Ding, H., Ren, C., and Wang, W., 2006. IBM-mart SCOR—a SCOR based
supply chain transformation platform through simulation and optimisation
techniques. In: Proceedings of the 2006 Winter Simulation Conference, 650–659.
Eksioglu, S.D., Romeijn, H.E., and Pardalos, P.M., 2006. Cross-facility management of
production and transportation planning problem. Computers and Operations
Research, 33, 3231–3251.
Elgazzar, S.H., Tipi,N.S., Hubbard,N.J., and Leach,D.Z., 2012. Linking supply chain
processes’ performance to a company’s financial strategic objectives. European
Journal of Operational Research, 223(1), 276–289.
Ellram, L.M., Tate, W.L.,and Billington., 2004. Understanding and managing the
services supply chain. Journal of Supply Chain Management, 40, 17-32.
Floudas, C.A. and Anastasiadis, S.H., 1988. Synthesis of General Distillation Sequences
with Several Multicomponent Feeds and Products. Chemical Engineering
Science, 43,2407.
Ganga, G. M. D., and Carpinetti, L. C. R., 2011. A fuzzy logic approach to supply chain
performance management. International Journal of Production Economics,
134(1), 177–187
Gen, M.S., and Syarif, A., 2005. Hybrid genetic algorithm for multi-time period
production/distribution planning. Computers and Industrial Engineering, 48,
799–809.
Ref. code: 25605422300342CMU
119
Giannakis, M., 2011. Management of service supply chains with a service-oriented
reference model: the case of management consulting. Supply Chain
Management: An International Journal, 16(5), 346–361.
Giannoccaro, I., Pontrandolfo, P., and Scozzi, B., 2003. A fuzzy echelon approach for
inventory management in supply chains. European Journal of Operational
Research, 149, 185–196.
Goetschalckx, M., Vidal, C.J., and Dogan, K., 2002. Modeling and design of global
logistics systems: a review of integrated strategic and tactical models and design
algorithms. European Journal of Operational Research,143, 1–18.
Guillen, G., Mele, F.D., Bagajewicz, M.J., Espuna, A., and Puigjaner, L., 2005.
Multiobjective supply chain design under uncertainty. Chemical Engineering
Science, 60, 1535–1553.
Gulledge, T., and Chavusholu, T., 2008. Automating the construction of supply chain
key performance indicators. Industrial Management & Data Systems, 108 (6),
750–774.
Gumus, A. T., Guneri, A. F., and Ulengin, F., 2010. A new methodology for multi-
echelon inventory management in stochastic and neuro-fuzzy environments.
International Journal of Production Economics, 128(1), 248–260
Gunasekaran, A., Patel, C. and Tirtiroglu, E., 2001. Performance measures and metrics
in a supply chain environment. International Journal of Operations &
Production Management, 21 (1/2), 71-87.
Guruprasad, P., and Herrmann, J.W., 2006. A hierarchical approach to supply chain
simulation modelling using the supply chain operations reference model.
International Journal of Simulation and Process Modelling, 2 (3/4), 124–132.
.Gupta, A., and Maranas, C.D., 2003. Managing demand uncertainty in supply chain
planning. Computers and Chemical Engineering, 27, 1219–1227.
Han, S.H., and Chu, C.H., 2009. Developing a collaborative supply chain reference
model for a regional manufacturing industry in China. International Journal of
Electronic Customer Relationship Management, 3 (1), 52–70.
Harelstad, C., Swartwood, D., and Malin, J., 2004. The value of combining best
practices. ASQ Six Sigma Forum Magazine August, 19–24.
Hauser, J. R., and Clausing, D., 1988. The House of Quality. Harvard Business Review.
Ref. code: 25605422300342CMU
120
Ho, W., He, T., Lee, C.K.M., and Emrouznejad, A., 2012. Strategic logistics outsourcing: An integrated QFD and fuzzy AHP approach. Expert Systems with Applications,
39(12), 10841–10850.
Ho, W., Dey, P.K., and Lockstrom, M. Strategic sourcing: a combined QFD and AHP
approach in manufacturing. Supply Chain Management, 16(6), 446–461.
Huang, S.H., Sheoran, S.K., and Keskar, H., 2005. Computer-assisted supply chain
configuration based on Supply Chain Operations Reference (SCOR) model.
Computers & Industrial Engineering, 48 (2), 377–394.
Hwang, Y., Lin, Y., and Lyujr, J., 2008. The performance evaluation of SCOR sourcing
process—The case study of Taiwan’s TFT-LCD industry. International Journal
of Production Economics, 115(2), 411–423.
Hwang, Y.D., Wenb, Y.F., and Chen, M.C., 2010. A study on the relationship between
the PDSA cycle of green purchasing and the performance of the SCOR model.
Total Quality Management (TQM), 21 (12), 1261–1278.
Iijima, T., Nakajima, Y., and Nishiwaki,Y., 1995. Application of fuzzy logic control
system for reactor feed-water control. Fuzzy Sets and Systems, 74(1), 61-72.
Jalalvand, F., Teimoury, E., Makui, a., Aryanezhad, M. B., and Jolai, F., 2011. A method
to compare supply chains of an industry. Supply Chain Management: An
International Journal, 16(2), 82–97.
Jang, Y.J., Jang, S.Y., Chang, B.M., Park, J., 2002. A combined model of network design
and production/distribution planning for a supply network. Computers and
Industrial Engineering, 43, 263–281.
Jayaraman, V., and Pirkul, H., 2001. Planning and coordination of production and
distribution facilities for multiple commodities. European Journal of
Operational Research, 133, 394–408.
Jia, G.Z., and Bai, M., 2011. An approach for manufacturing strategy development based
on fuzzy-QFD. Computers & Industrial Engineering, 60(3), 445–454.
John, R., and Bennett, S., 1997. The use of fuzzy sets for resource allocation in an
advance request vehicle brokerage system—a case study. Journal of the
Operational Research Society, 48(2), 117-123.
Julien, B., 1994. An extension to possibilistic linear-programming, Fuzzy Sets and
Systems, 64, 195–206.
Ref. code: 25605422300342CMU
121
Jung, H., Jeong, B., Lee, C.G., 2008. An order quantity negotiation model for distributor-
driven supply chains. International Journal of Production Economics, 111, 147–
158.
Kabak, Ö., & Ülengin, F., 2011. Possibilistic linear-programming approach for supply
chain networking decisions. European Journal of Operational Research, 209,
253–264.
Kahraman,C.,Ertay,T., and Buyukozkan, G., 2006. A fuzzy optimization model for QFD
planning process using analytic network approach. European Journal of
Operational Research, 171 (2), 390–411.
Kannan, D., Jafarian, A., Khamene, H.A., and Olfat, L., 2013. Competitive performance
improvement by operational budget allocation using ANFIS and fuzzy quality
function deployment: a case study. International Journal of Advanced
Manufacturing Technology, 68 (1), 849-862.
Kanyalkar, A.P., and Adil, G.K., 2005. An integrated aggregate and detailed planning in
a multi-site production environment using linear programming. International
Journal of Production Research, 43, 4431–4454.
Kaplan, RS., and Norton, DP.,1997. Using the balanced scorecard as a strategic
management system. Harvard Business Review,74 (1), 75-85.
Karsak, E., and Dursun, M., 2015. An integrated fuzzy MCDM approach for supplier
evaluation and selection. Computers and Industrial Engineering, 82, 82–93.
Karsak, E. E., Sozer, S., and Alptekin, S. E., 2002. Product planning in quality function
deployment using a combined analytic network process and goal programming
approach. Computers and Industrial Engineering, 44(1), 171–190.
Kevan, T., 2005. Modeling the future. Frontline Solutions, 6 (1), 22–24.
Klir, G., and Wierman, M., 1996. Uncertainty-based Information, Physica-Verlag,
Heidelberg.
Klir, GJ., and Yuan, B. 1995. Fuzzy Sets and Fuzzy Logic: Theory and Applications.
Prentice-Hall PTR, Upper Saddle River, New Jersey
Kocaoğlu, B., Gülsün, B., and Tanyaş, M., 2011. A SCOR based approach for measuring
a benchmarkable supply chain performance. Journal of Intelligent
Manufacturing, 24 (2013), 113–132.
Ref. code: 25605422300342CMU
122
Kumar, M., Vrat, P., and Shankar, R., 2004. A fuzzy goal programming approach for
vendor selection problem in a supply chain. Computers and Industrial
Engineering, 46(1), 69–85.
Kuo, T.C., Wu, H.H., and Shieh, J.I., 2009. Integration of environmental considerations
in quality function deployment by using fuzzy logic. Expert Systems with
Applications, 36(3), 7148–7156.
Kwong, C.K., Chen, Y., Bai, H., Chan, D.S.K., 2007. A methodology of determining
aggregated importance of engineering characteristics in QFD. Computers and
Industrial Engineering, 53, 667–679.
Lager, T., 2005. The industrial usability of quality function deployment: A literature
review and synthesis on a meta-level. R&D Management, 35 (4), 409–426.
Land, A. A. H., and Doig, A. G., 1960. An Automatic Method of Solving Discrete
Programming Problems. Econometrica, 28 (3), 497-520.
Lapidus, R.S., and Schibrowsky, J.A., 1994. Aggregate complaint analysis: A procedure
for developing customer service satisfaction. Journal of Services Marketing, 8
(4), 50–60.
Lee, H.L., and Billington, C., 1993. Material management in decentralized supply chain.
Operations Research, 41 (5), 835–847.
Lee, Y.C., Sheu, L.C., and Tsou, Y.G. Quality function deployment implementation
based on Fuzzy Kano model: an application in PLM system. Computers &
Industrial Engineering, 55, 48–63.
Li, S., Rao, S. S., Ragu-nathan, T. S., and Ragu-nathan, B., 2005. Development and
validation of a measurement instrument for studying supply chain management
practices. Journal of Operations Management, 23 (6), 618–641.
Li, L., Su, Q., and Chen, X., 2011. Ensuring supply chain quality performance through
applying the SCOR model. International Journal of Production Research, 49
(1), 33–57.
Liang, T.F., 2006. Project management decisions using fuzzy linear programming.
International Journal of Systems Science, 37, 1141-1152.
Liang, T.F., 2008. Fuzzy multi-objective production-distribution planning decisions with
multi-product and multi-time period in a supply chain. Computers & Industrial
Engineering, 55 (3), 676-694.
Ref. code: 25605422300342CMU
123
Liao, C.N., and Kao, H.P., 2014. An evaluation approach to logistics service using fuzzy
theory, quality function development and goal programming. Computers &
Industrial Engineering, 68, 54–64.
Lin, Y., Cheng, H.P, Tseng, M.L, and Tsai, J.C., 2010. Using QFD and ANP to analyze
the environmental production requirements in linguistic preferences. Expert
Systems With Applications, 37(3), 2186–2196.
Lindley, D., 198 Comment: a tale of two wells. Stat. Sci., 2, 38–40. Quine
Lima-Junior, F. R., and Carpinetti, L. C. R., 2016. Combining SCOR model and fuzzy
TOPSIS for supplier evaluation and management. International Journal of
Production Economics, 174, 128–141.
Liu, H.T., 2009. Expert Systems with Applications the extension of fuzzy QFD: From
product planning to part deployment. Expert Systems with Applications, 36(8), 11131–11144.
Liu, H.T., 2011. Product design and selection using fuzzy QFD and fuzzy MCDM
approaches. Applied Mathematical Modelling, 35(1), 482–496.
Liu, S.T., and Kao, C., 2004. Solving fuzzy transportation problems based on extension
principle. European Journal of Operational Research, 153, 661–674.
Liu, H.T., and Wang, C.H., 2010. An advanced quality function deployment model using
fuzzy analytic network process. Applied Mathematical Modelling, 34(11), 3333–
3351.
Lockamy III, A. and McCormack, K., 2004. Linking SCOR planning practices to supply
chain performance, an explorative study. International Journal of Operations &
Production Management, 24 (12), 1192–1218.
Majozi, T., and Zhu, X. X., 2005. A combined fuzzy set theory and MILP approach in
integration of planning and scheduling of batch plants—Personnel evaluation
and allocation. Computers & Chemical Engineering, 29(9), 2029-2047.
Malin, J.H., and Reichardt, E., 2005. Strengthen the six sigma portfolio. Quality, 44 (6),
40–43.
Martin, C.H., Dent, D.C., and Eckhart, J.C., 1993. Integrated production, distribution,
and inventory planning at Libbey–Owens–Ford. Interfaces,23, 78–86.
Maskell, B. H., 1991. Performance measurement for world class manufacturing. USA:
Productivity Press.
Ref. code: 25605422300342CMU
124
Mayyas, A.,Shen,Q.,Abdelhamid,M., Shan,D., Qattawi,A., and Omar,M., 2011. Using
Quality Function Deployment and Analytical Hierarchy Process for material
selection of Body-In-White. Materials & Design, 32(5), 2771–2782.
McCormack, K., Ladeira, M.B., and Valadares de Oliveira, M.P., 2008. Supply chain
maturity and performance in Brazil. Supply Chain Management: An
International Journal, 13 (4), 272–282.
Mcdonald, C.M., and Karimi, I.A., 1997. Planning and scheduling of parallel
semicontinuous processes. 1. Production planning. Industrial and Engineering
Chemistry Research, 36, 2691–2700.
Meijboom, B., and Obel, B., 2007. Tactical coordination in a multi-location and multi-
stage operations structure: a model and a pharmaceutical company case. Omega-
International Journal of Management Science, 35, 258–273.
Mogale, D. G., Dolgui, A., Kandhway, R., Krishna, S., & Kumar, M., 2017. A multi-period inventory transportation model for tactical planning of food grain supply
chain. Computers & Industrial Engineering, 110, 379–394.
Monfared., M. A. S. and Steiner, S.J., 2000. Fuzzy adaptive scheduling and control
systems. Fuzzy Sets and Systems,115(2), 231–246.
Mula, J., Peidro, D., Madroñero, M.D., and Vicens, E., 2010. Mathematical
programming models for supply chain production and transport planning.
European Journal of Operational Research, 204(3), 377-390.
Nahm, Y.E., 2013. A novel approach to prioritize customer requirements in QFD based
on customer satisfaction function for customer-oriented product design. Journal
of Mechanical Science and Technology, 27, 3765-3777.
Neely, A., Gregory, M., and Platts, K., 1995. Performance measurement system design:
A literature review and research agenda. International Journal of Operations
and Production Management, 15 (4), 80-116.
Nemhauser, G. and Wolsey, L.A. 1988. Integer and Combinatorial Optimization. John
Wiley and Sons, New York.
Oh, H.C., and Karimi, I.A., 2006. Global multiproduct production–distribution planning
with duty drawbacks. AICHE Journal, 52, 595–610.
Palma-Mendoza, J. A., 2014. Analytical hierarchy process and SCOR model to support
supply chain re-design. International Journal of Information Management, 34
(5), 634–638.
Ref. code: 25605422300342CMU
125
Pan, N.H., Lin,Y.Y., and Pan, N.F. (2010). Enhancing construction project supply chains
and performance evaluation methods: a case study of a bridge construction
project. Canadian Journal of Civil Engineering. 37,1094–1116.
Papoulias, S.A. and I.E. Grossmann. (1983). A Structural Optimization Approach in
Process Synthesis. Part II: Heat Recovery Networks. Computers and Chemical
Engineering, 7,707.
Park, Y.B., 2005. An integrated approach for production and distribution planning in
supply chain management. International Journal of Production Research, 43,
1205–1224.
Parker, C., 2000. Performance measurement. Work Study, 49 (2), 63-66.
Parker, R. and Rardin, R., 1988. Discrete Optimization. Academic Press, San Diego.
Partovi, F.Y., 2001. An analytic model to quantify strategic service vision. International
Journal of Service Industry Management, 12 (5), 476–499.
Partovi,F.Y., 2006. An analytic model for locating facilities strategically. OMEGA - The
International Journal of Management Science.34, 41–55.
Peidro, D., Mula, J., Jiménez, M., & Botella, M., 2010. A fuzzy linear programming
based approach for tactical supply chain planning in an uncertainty
environment. European Journal of Operational Research, 205(1), 65–80.
Perea-lopez, E., Ydstie, B.E., Grossmann, I.E., 2003. A model predictive control strategy
for supply chain optimization. Computers and Chemical Engineering, 27, 1201–
1218.
Person, F., 2003. Supply chain simulation: experiences from two case studies. In:
Verbraeck, A., Hlupic, V. (Eds.). Proceedings from the 15th European Simulation
Symposium, Delft, The Netherlands, October 26–29,399–404.
Persson, F., and Araldi, M., 2009. The development of a dynamic supply chain analysis
tool—Integration of SCOR and discrete event simulation. International Journal
of Production Economics, 121(2), 574–583.
Petrovic, D., Roy, R., and Petrovic, R., 1999. Supply chain modeling using fuzzy sets.
International Journal of Production Economics, 59, 443–453.
Potthast, J.M., Gärtner, H., and Hertrampf, F., 2010. Allocation for manufacturing
companies. Electronic Scientific. Journal of Logistics, 6(2), 19–24.
Ref. code: 25605422300342CMU
126
Rabelo, L., Eskandari, H., Shaalan, T., and Helal, M., 2007. Value chain analysis using
hybrid simulation and AHP. International Journal of Production Economics,
105(2), 536–547.
Ramasamy, N.R., and Selladurai, V., 2004. Fuzzy logic approach to prioritize
engineering characteristics in quality function deployment (FL-QFD). International Journal of Quality and Reliability Management, 21(9), 1012–1023.
Reyes, H.G., and Giachetti, R., 2010. Using experts to develop a supply chain maturity
model in Mexico. Supply Chain Management: An International Journal,15 (6),
415–424.
Rizk, N., Martel, A., and D’amours, S., 2008. Synchronized production-distribution
planning in a single-plant multi-destination network. Journal of the Operational
Research Society, 59, 90–104.
Röder, A., and Tibken, B., 2006. A methodology for modelling inter-company supply
chains and for evaluating a method of integrated product and process
documentation. European Journal of Operational Research, 169 (3), 1010–1029.
Romo, F., Tomasgard, A., Hellemo, L., Fodstad, M., Eidesen, B.H., and Pedersen, B.,
2009. Optimizing the Norwegian Natural gas production and transport.
Interfaces, 39 (1), 46–56.
Russel, D.M., Ruamsook, K., and Thomchick, E.A., 2009. Ethanol and the petroleum
supply chain of the future: five strategic priorities of integration. Transportation
Journal, 48 (1), 5–22.
Ryu, J.H., Dua, V., and Pistikopoulos, E.N., 2004. A bilevel programming framework
for enterprise-wide process networks under uncertainty. Computers and
Chemical Engineering, 28, 1121–1129.
Sakawa, M., Nishizaki, I., and Uemura, Y., 2001. Fuzzy programming and profit and
cost allocation for a production and transportation problem. European Journal
of Operational Research, 131, 1–15.
Sazvar, Z., Al-e-hashem, S. M. J. M., Baboli, A., & Jokar, M. R. A., 2014. A bi-objective
stochastic programming model for a centralized green supply chain with
deteriorating products. International Journal of Production Economics, 150,
140–154.
Ref. code: 25605422300342CMU
127
Schmitz, P.M.U., 2008. The Use of Supply Chains and Supply Chain Management to
Improve The Efficiency and Effectiveness of GIS unit PhD thesis (unpublished). University of Johannesburg, South Africa, 523
Schnetzler, M.J., Lemm, R., Bonfils, P., and Thees, O., 2009. The supply chain
operations reference (SCOR) model to describe the value-added chain in forestry.
German Journal of Forest Research. 180 (1/2), 1–14.
Schrijver, A., 1986. Theory of Linear and Integer Programming. Wiley, Chichester.
Selim, H., Am, C., and Ozkarahan, I., 2008. Collaborative production–distribution
planning in supply chain: a fuzzy goal programming approach. Transportation
Research Part E-Logistics and Transportation Review, 44, 396–419.
Sellitto, M. A., Pereira, G. M., Borchardt, M., da Silva, R. I., and Viegas, C. V., 2015. A
SCOR-based model for supply chain performance measurement: application in
the footwear industry. International Journal of Production Research, 53(16),
4917–4926.
Sen, C.G., and Baracli, H., 2010. Fuzzy quality function deployment based methodology
for acquiring enterprise software selection requirements. Expert Systems with
Applications, 37, 3415–3426.
Sener, Z., and Karsak, E.E., 2011. A combined fuzzy linear regression and fuzzy
multiple objective programming approach for setting target levels in quality
function deployment. Expert Systems with Applications, 38(4), 3015–3022.
Shakourloo, A., Kazemi, A., Oroojeni, M., & Javad, M., 2016. A new model for more
effective supplier selection and remanufacturing process in a closed-loop supply
chain. Applied Mathematical Modelling, 40, 9914–9931.
Shan, N.and Pantelides, CC., 1991. Optimal long-term campaign planning and design of
batch-operations. Industrial and Engineering Chemistry Research, 30, 2308-2321.
Shaw, K., Shankar, R., Yadav, S. S., & Thakur, L. S., 2012. Supplier selection using fuzzy
AHP and fuzzy multi-objective linear programming for developing low carbon
supply chain. Expert Systems With Applications, 39(9), 8182–8192.
Shen,X.X., Tan, K.C., and Xie,M., 2000. Benchmarking in QFD for quality
improvement, Benchmarking: An International Journal, 7(4), 282-291.
Ref. code: 25605422300342CMU
128
Shepherd, C., and Gunter, H., 2006. Measuring supply chain performance: current
research and future directions. International Journal of Productivity and
Performance Management, 55 (3/4), 242-258.
Shih, L.H.,1999. Cement transportation planning via fuzzy linear programming.
International Journal of Production Economics, 58, 277–287.
Siler,W., 1987. Building fuzzy expert systems. http://users.aol.com/wsiler/
Soffer, P. and Wand, Y.,2007. Goal-driven multi-process analysis. Journal of the
Association for Information Systems, 8 (3),175–204.
Sohn, Y. S., and Choi, I. S., 2001. Fuzzy QFD for supply chain management with
reliability. Reliability Engineering and System Safety, 72, 327–334.
Stephens, S., 2001. Supply Chain Council and Supply Chain Operations Reference
Model Overview. Supply Chain Council, Inc.
Stuart, F.I., and Tax, S.S., 1996. Planning for service quality: An integrative approach. International Journal of Service Industry Management, 7 (4), 58–77.
Subramanian, K., Rawlings, J. B., & Maravelias, C. T., 2014. Economic model predictive
control for inventory management in supply chains. Computers and Chemical
Engineering, 64, 71–80.
Syam, S. S., & Bhatnagar, A., 2015. A decision support model for determining the level
of product variety with marketing and supply chain considerations. Journal of
Retailing and Consumer Services, 25, 12–21.
Thakkar, J., Patel, A.D., Kanda, A., and Deshmukh, S.G., 2009. Supply chain
performance measurement framework for small and medium scale enterprises.
Benchmarking: An International Journal, 16 (5), 702–723.
Theeranuphattana, A., and Tang, J.C.S., 2008. A conceptual model of performance
measurement for supply chains; alternative considerations. Journal of
Manufacturing Technology Management, 19 (1), 25–48.
Timpe, C.H., and Kallrath, J., 2000. Optimal planning in large multi-site production
networks. European Journal of Operational Research, 126, 422–435.
Torabi, S.A., and Hassini, E., 2008. An interactive possibilistic programming approach
for multiple objective supply chain master planning. Fuzzy Sets and
Systems,159, 193–214.
Ref. code: 25605422300342CMU
129
Tsai, C. Y., 2003. Using fuzzy QFD to enhance manufacturing strategic planning. Journal of the Chinese Institute of Industrial Engineers, 18(3), 33–41.
Ubando, A.T., Culaba, A.B., Aviso, K.B., Tan, R.R., Cuello, J.L., Ng, D.K.S., and
Halwagi, M.M.E., 2016. Fuzzy mixed integer non-linear programming model for
the design of an algae-based eco-industrial park with prospective selection of
support tenants under product price variability. Journal of Cleaner Production,
136, 183-196.
Van Landeghem, R., and Persoons, K., 2001. Benchmarking of logistical operations
based on a causal model. International Journal of Operations & Production
Management, 21(1/2), 254-267.
Van Roy, T.J, and Wolsey, L.A., 1987. Solving mixed 0-1 programs by automatic
reformulation. Operations Research, 35,45–57.
Vanany, I., Suwignjo, P., and Yulianto, D., 2005. Design of supply chain performance
measurement system for lamp industry. In: Proceedings of the 1st International
Conference on operations and supply chain management, Bali, H-78.
Vasilash, G.S., 1989. Hearing the voice of the customer. Production, 34(2), 66-8.
Wang, C.H, 2015. Using quality function deployment to conduct vendor assessment and
supplier recommendation for business-intelligence systems. Computers &
Industrial Engineering, 84, 24–31.
Wang, W. Y. C., Chan, H. K., and Pauleen, D. J., 2010. Aligning business process
reengineering in implementing global supply chain systems by the SCOR
model. International Journal of Production Research, 48(19), 5647–5669.
Wang, G., Huang, S. H., and Dismukes, J. P., 2004. Product-driven supply chain selection
using integrated multi-criteria decision-making methodology. International
Journal of Production Economics, 91(1), 1-15.
Wang, J., and Shu, Y.-F., 2005. Fuzzy decision modeling for supply chain management.
Fuzzy Sets and Systems, 150, 107–127.
Wu, Y., 2010. Computers & Industrial Engineering A time staged linear programming
model for production loading problems with import quota limit in a global
supply chain. Computers & Industrial Engineering, 59(4), 520–529.
Xiao, R., Cai, Z., and Zhang, X., 2012. An optimisation approach to risk decision-making
of closed-loop logistics based on SCOR model. Optimisation, 61(10),1221–1251.
Ref. code: 25605422300342CMU
130
Yang, Y. Q., Wang, S. Q., Dulaimi, M., and Low, S. P., 2003. A fuzzy quality function
deployment system for buildable design decision-makings. Automation in
Construction, 12, 381–393.
Yilmaz, Y., and Bititci, U., 2006. Performance measurement in the value chain:
manufacturing v.tourism. International Journal of Productivity and
Performance Management, 55 (5), 371–389.
Yücel, A., and Güneri, A. F., 2010. A weighted additive fuzzy programming approach
for multi-criteria supplier selection. Expert Systems with Applications, 38(5),
6281–6286.
Zadeh, L. A., 1965. Fuzzy sets. Information and Control, 8(3), 338–353.
Zaim, S., Sevkli, M., Camgo ¨z-Akdag, H., Demirel, O.F., Yayla, A.Y., and Delen, D., 2014. Use of ANP weighted crisp and fuzzy QFD for product development. Expert Systems With Applications. 41, 4464–4474.
Zangoueinezhad, A., Azary, Y., and Kazaziz, A., 2011. Using SCOR model with fuzzy
MCDM approach to assess competitiveness positioning of supply chains: focus
on shipbuilding supply chains. Maritime Policy & Management, 38 (1), 93–109.
Zarei, M., Fakhrzad, M.B., and Paghaleh, M.J., 2011. Food supply chain leanness using
a developed QFD model. Journal of Food Engineering, 102(1), 25–33.
Zhang, F., Yang, M., and Liu, W., (2014). Using integrated quality function deployment
and theory of innovation problem solving approach for ergonomic product
design. Computers & Industrial Engineering, 76, 60–74.
Zimmermann, H. J., 1978. Fuzzy programming and linear programming with several
objective functions. Fuzzy Sets and System, 1, 44–55.
Ref. code: 25605422300342CMU
131
Appendix
Ref. code: 25605422300342CMU
132
Appendix A
Lingo Code
A1: Predictive MILP model for SCOR performance evaluation : most likely.
MODEL: SETS: PERIOD1/1..60/:CR,day,Sbin,W,Craw; PERIOD2/1..59/; PRODUCT/1..2/:CM,CU,CS,CI,CB,e,PRICE,WIP,InvZero,BackZero,pack,rr,CJ,bag;
STAGE/1..3/; PROD_PERIOD1(PRODUCT,PERIOD1):S,Smax,B,D; PROD_PERIOD2(PRODUCT,PERIOD2); PROD_PERIOD1_STAGE(PRODUCT,PERIOD1,STAGE):J,Q,P,In; PROD_PERIOD2_STAGE(PRODUCT,PERIOD2,STAGE); PERIOD1_STAGE(PERIOD1,STAGE):G,M; PERIOD2_STAGE(PERIOD2,STAGE); !here are 2 machines to produce the products; MACHINE/1..2/; MACHINE_STAGE(MACHINE,STAGE):Capa; MACHINE_PERIOD1(MACHINE,PERIOD1):n; ENDSETS
!Objective Function is to maximize PROFIT; MAX=PROFIT;
PROFIT=REVENUE-COST;
REVENUE=@SUM(PROD_PERIOD1(i,t):PRICE(i)*D(i,t));
COST=@SUM(PERIOD1(t):CR(t)*W(t)*day(t))+@SUM(PROD_PERIOD1_STAGE(i,t,k):((CM(i)+CU(i))*(P(i,t,k)/pack(i)))+(CI(i)*In(i,t,k))+(CJ(i)*(J(i,t,k)/bag(i))))+@SUM(PROD_PERIOD1(i,t):(CS(i)*S(i,t))+(CB(i)*B(i,t)))+@SUM(PERIOD1_STAGE(t,k):Craw(t)*(M(t,k)/1000));
COST1=@SUM(PROD_PERIOD1_STAGE(i,t,k):((CM(i)+CU(i))*(P(i,t,k)/pack(i)))); COST2=@SUM(PROD_PERIOD1_STAGE(i,t,k):CI(i)*In(i,t,k)); COST3=@SUM(PROD_PERIOD1(i,t):CB(i)*B(i,t)); COST4=@SUM(PROD_PERIOD1(i,t):CS(i)*S(i,t)); COST5=@SUM(PERIOD1(t):CR(t)*W(t)*day(t)); COST6=@SUM(PROD_PERIOD1_STAGE(i,t,k):CJ(i)*(J(i,t,k)/bag(i))); COST7=@SUM(PERIOD1_STAGE(t,k):Craw(t)*(M(t,k)/1000));
!Profit can be negative; @FREE(PROFIT); !_________________________; !Material balance eq; !For stage Beginning, period1; @FOR(PERIOD1_STAGE(t,k)|k#EQ#1:
Ref. code: 25605422300342CMU
133
@FOR(PERIOD1(t):@SUM(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#1:Q(i,1,k+1)*rr(i))=INM+G(1,k)-M(1,k)));
!For stage beginning, period2-12; @FOR(PERIOD2_STAGE(t,k)|k#EQ#1: @FOR(PERIOD2(t):@SUM(PROD_PERIOD2_STAGE(i,t,k)|k#EQ#1:Q(i,t+1,k+1)*rr(i))=M(t,k)+G(t+1,k)-M(t+1,k)));
@FOR(PERIOD1_STAGE(t,k)|k#EQ#1:M(t,k)<=1500000); @FOR(PERIOD1_STAGE(t,k)|k#EQ#1:M(t,k)>=113400);
!____________________________________________________________;
!Inventory Balance Constraints for WIP; !For stage WIP, period1; @FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#2:J(i,1,k)=WIP(i)+Q(i,1,k)-P(i,1,k+1));
!For stage WIP, period2-12; @FOR(PROD_PERIOD2_STAGE(i,t,k)|k#EQ#2:J(i,t+1,k)=J(i,t,k)+Q(i,t+1,k)-P(i,t+1,k+1));
!For stage FIN,product1,period1; @FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#3:In(i,1,k)-B(i,1)=InvZero(i)-BackZero(i)+(P(i,1,k)/pack(i))+S(i,1)-D(i,1));
!For stage FIN,product1,period2_12; @FOR(PROD_PERIOD2_STAGE(i,t,k)|k#EQ#3:In(i,t+1,k)-B(i,t+1)=In(i,t,k)-B(i,t)+(P(i,t+1,k)/pack(i))+S(i,t+1)-D(i,t+1));
!There is a safety stock per day, which is derived from theoretical value;
@FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#3:In(1,t,k)>=750); @FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#3:In(2,t,k)>=1200);
!Upper bound for keeping FIN inv; @FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#3:In(1,t,k)<=5000); @FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#3:In(2,t,k)<=10000);
!Ending inv of Fin goods; @FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#3:In(1,60,k)=1000); @FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#3:In(2,60,k)=6000); !Inventory of WIP; !WIP is allow to store only 90 bag for 1500cc and 120 bag for 600cc; @FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#2:J(1,t,k)<=38000); @FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#2:J(2,t,k)<=100000);
!Safety stock of WIP derived from theory; @FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#2:J(1,t,k)>=0); @FOR(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#2:J(2,t,k)>=15000);
Ref. code: 25605422300342CMU
134
!Subcontract plan; @FOR(PROD_PERIOD1(i,t):S(i,t)<=(Sbin(t)*Smax(i,t)));
!Backorder is allowed but must be 0 at end month; @FOR(PROD_PERIOD1(i,t):B(i,60)=0); !Production Constraint; !For production at WIP stage; @FOR(MACHINE_STAGE(j,k)|j#EQ#1: @FOR(PERIOD1(t):@SUM(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#2:Q(i,t,k))<=gamma*Capa(j,k)*day(t)*n(j,t)));
!For production at FIN stage; @FOR(MACHINE_STAGE(j,k)|j#EQ#2: @FOR(PERIOD1(t):@SUM(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#3:P(i,t,k))<=gamma*Capa(j,k)*day(t)*n(j,t)));
!Workforce constraint; @FOR(PERIOD1(t):@SUM(PROD_PERIOD1_STAGE(i,t,k)|k#EQ#3:e(i)*P(i,t,k))<=roll*day(t)*shift*W(t));
!RM is to order at minimum 2 Ton each;
!@FOR(PERIOD1_STAGE(t,k)|k#EQ#1: @FOR(PERIOD1(t):G(t,k)=(1814370*Ia(t)))); !@FOR(PERIOD1(t):@BIN(Ia(t))); !Workdays are given;
!Demand and Number of working days per month data;
DATA: CM,CU,CR,CS,CI,CB,e,D,gamma,roll,shift,day,Smax,Sbin,price,capa,n,W,G
,InvZero,pack,BackZero,WIP,rr,INM,CJ,bag,Craw=@OLE('C:\Users\TOSHIBA\Google Drive\Paper2015\normal\mrp_most_likely_RM.xlsm');
@OLE('C:\Users\TOSHIBA\Google Drive\Paper2015\normal\mrp_most_likely_RM.xlsm')=P,S,B,In,Q,J,M;
ENDDATA
END