A modied failure mode and effects analysis method for supplier

selection problems in the supply chain risk environment: A case study

q
Ping-Shun Chen

, Ming-Tsung Wu
Department of Industrial and Systems Engineering, Chung Yuan Christian University, Chung Li 320, Taiwan, ROC
a r t i c l e i n f o
Article history:
Received 13 October 2012
Received in revised form 15 August 2013
Accepted 22 September 2013
Available online 30 September 2013
Keywords:
Analytic hierarchy process (AHP)
Failure mode and effects analysis (FMEA)
Supply chain risk management
Supplier selection
a b s t r a c t
In the emerging supply chain environment, supply chain risk management plays a more important role
than ever. Companies must focus not only on the efciency of supply chain, but also on its risks. If an
unanticipated event occurs, all of the supply chain members will be impacted, and the result will cause
signicant loss. Therefore, this research proposes a modied failure mode and effects analysis (MFMEA)
method to select new suppliers from the supply chain risks perspective and applies the analytic
hierarchy process (AHP) method to determine the weight of each criterion and sub-criterion for supplier
selection. An IC assembly company is then studied to validate this model. The result shows that the case
company can categorize its suppliers more effectively and at the same time select a low-risk supply chain
partner. Moreover, the case company can provide unsatisfactory suppliers with valuable feedback that
will help them improve and become its partners in the future.
2013 Elsevier Ltd. All rights reserved.
1. Introduction
Supply chain risk management (SCRM) has become an essential
issue for supply chain management (Neiger, Rotaru, & Churilov,
2009; Schoenherr, Rao Tummala, & Harrison, 2008; Tang, 2006;
Thun & Hoenig, 2011; Wu & Olson, 2008). Hallikas, Karvonen,
Pulkkinen, Virolainen, and Tuominen (2004) have described supply
chain not as a simple vertical chain, but as a multi-layer supply
network. The supply chain network has to confront four types of
risks: demand, due date, cost management, and risk associated
with production capability and operation exibility. Tang (2006)
described that the supply chain risk as comprised of operational
and disruption risks. Operational risk was associated with the
uncertainty of a process such as customer demand, the amount
of supply, and cost uctuations. Disruption risk encompassed
natural and human disasters, such as earthquakes, oods,
hurricanes, terrorist attacks, nancial crises, or labor strikes. For
example, the ood of Thailand in 2011 caused serious damage to
warehouses of hard drives suppliers. These suppliers were unable
to fulll PC customers orders on time during the ood. Since these
suppliers provided large quantities of hard drives for PC manufac-
turers, this created a shortage of hard drives throughout the entire
PC supply chain. This was a typical supply chain risk caused by a
natural disaster.
However, since disruption risk is difcult to predict and pre-
vent, this research focuses on decreasing the operational risk. In or-
der to reduce the uncertainty of the supply, the SCRM should
construct a good supplier selection and assessment system
(Srinivasan, Mukherjee, & Gaur, 2011; Tang, 2006). Collaborating
with suppliers at a low operational risk could reduce the chances
of losses for all members of the supply chain.
Based on the literature review, previous researchers have
viewed supplier selections as a multi-criteria decision problem
(Che & Wang, 2008; Ho, Xu, & Dey, 2010). The multi-attribute deci-
sion making technique is often used to solve this problem (Ho
et al., 2010; Ng, 2008). However, this research considers this type
(operational risk) of problem as a SCRM problem; few researchers
have focused on this eld (Sawik, 2011; Wu, Blackhurst, &
Chidambaram, 2006). Pillay and Wang (2003) found that the result
of the FMEA could assist managers in making the right decisions in
the face of supply chain risk. In practice, the FMEA has been used in
product design and manufacturing improvement. Therefore, intro-
ducing the FMEA into the supplier evaluation and selection is fea-
sible. Further, this study considers the SCRM in suppliers
evaluation and selection. In order to develop a supplier selection
procedure, this study proposes a modied FMEA (MFMEA) method,
which integrated the FMEA and AHP methods, to construct a
supplier evaluation system and to discuss potential failure factors
and their effects on the system in a risky supply chain environ-
ment. Moreover, this research will assist companies in improving
their ways of selecting and evaluating suppliers. Finally, preventive
strategies to the potential failure factors are identied, and the
results are discussed and summarized.
0360-8352/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.cie.2013.09.018
q
The manuscript was handled by the area editor Qiuhong Zhao.

Corresponding author. Tel.: +886 32654410; fax: +886 32654499.

E-mail address: pingshun@cycu.edu.tw (P.-S. Chen).
Computers & Industrial Engineering 66 (2013) 634642
Contents lists available at ScienceDirect
Computers & Industrial Engineering
j our nal homepage: www. el sevi er. com/ l ocat e/ cai e
The rest of this article is organized as follows. Section 2 summa-
rizes the literature review of supplier selection. Section 3 describes
the case company. Section 4 proposes a MFMEA method. Section 5
discusses the results of this study and summarizes the managerial
implications based on this case study. The conclusions and future
research are presented in Section 6.
2. Literature review
Selecting the right suppliers is an important step in supplier
management (Shu & Wu, 2009; Tseng, Chiang, & Lan, 2009; Wu
et al., 2006). The selection of the right suppliers can reduce opera-
tional costs and delivery time (Che & Wang, 2008). Similarly,
choosing the wrong suppliers may increase the number of defec-
tive products, unstable deliveries, or refabricated costs, all of which
can increase companys total cost and tarnish its reputation. There-
fore, how to select the right suppliers has become an essential to-
pic for companies wishing to minimize their supply chain risks.
Previous studies have identied some criteria for evaluating
suppliers. Based on the relationships between suppliers and man-
ufacturers, Dickson (1966) summarized 23 criteria, which fell into
four categories: quality, deliverability, performance, and warranty
policy. Tracey and Tan (2001) chose quality, delivery reliability,
product characteristics, and unit price as criteria when manufac-
turers assessed their suppliers ability to increase customers satis-
faction and improve companies performance. Similarly, Katsikeas,
Paparoidamis, and Katsikea (2004) considered suppliers reliability,
competitive price, service, and technical skills. In their review of
research studies from 2000 to 2008, Ho et al. (2010) cited quality
as the critical criterion, followed by deliverability, price, manufac-
turing capability, service, management, and technology.
In order to increase a companys competitive advantage in sup-
ply chain management, enterprises have to maintain long-term
relationships with their most reliable suppliers. When companies
select the right suppliers, cost is not the only criterion to be consid-
ered; companies also need to consider quality, deliverability, and
service (Ho et al., 2010). Moreover, as mentioned in the previous
literature reviews, researchers have focused on suppliers quality,
cost, deliverability, and service. Therefore, the criteria adopted by
this research to select preferred suppliers include these four
factors.
According to supplier selection criteria, the supplier risk de-
pends on the type and degree of risks. A failure caused by the sup-
plier is viewed as a risk for the manufacturer. The manufacturer
should then evaluate and score the impact of each failure; the
sum of the scores is the supplier risk. Therefore, the preferred sup-
plier selection procedure is equal to the supplier lowest risk assess-
ment procedure. Although many researchers have studied the
supplier evaluation and selection problems (Ho et al., 2010), few
have explored supplier selection problems based on supplier risk
or supply chain risk. For example, Wu et al. (2006) used the ana-
lytic hierarchy process (AHP) to assess supply risk. They considered
six factors of risks: internal controllable, internal partial controlla-
ble, internal uncontrollable, external controllable, external partial
controllable, and external uncontrollable. Through the supply eval-
uation, a company could understand its supply risks based on each
factor and decide which supplier was the most preferred. Schoenh-
err et al. (2008) studied how a US manufacturing company as-
sessed its supply chain risk and made its offshore sourcing
decisions. The case company adopted the AHP method to evaluate
the weights of its main objectives (such as product, partner, and
environment) and sub-objectives (such as quality, cost, service,
and management capabilities). Based on the weights of the 16 fac-
tors, the case company could evaluate several offshore alterna-
tives: nished goods from China; nished goods from Mexico;
parts from China, Maquiladora, no investment; parts from China,
Maquiladora, with investment; and parts from China, with assem-
bly in the US. The results showed that sourcing nished goods from
China would be the best offshore strategy for the case company.
Further, Thun and Hoenig (2011) surveyed 67 German automotive
manufacturers to investigate the supply chain vulnerability and
the key drivers of supply chain risks. They applied the probabil-
ity-impact-matrix to analyze the internal and external supply
chain risks. They then offered suggestions for mitigating these sup-
ply chain risks.
Beside these supplier risk assessments, the failure mode and ef-
fects analysis (FMEA) is a popular method of measuring preventive
risks (Ko, 2013; Liu, Liu, & Liu, 2013). The FMEA has been exten-
sively applied in product design and manufacturing process plan-
ning (Almannai, Greenough, & Kay, 2008; Chen & Ko, 2009;
Ekmekcioglu & Kutlu, 2012). The traditional FMEA evaluated risks
by calculating the risk priority number (RPN). The RPN was com-
puted by multiplying three factors (O, S, and D), where O and S rep-
resented the occurrence and severity of a failure, and D was
dened as the detection that meant the ability to detect the failure
before it reached the customer (Chin, Wang, Poon, & Yang, 2009).
Each factor was evaluated on a 10-point scale. After calculating
the RPNs of each failure, managers could sort the RPNs from largest
to smallest. Failures with higher RPNs could be viewed as more
important and as meriting greater attention. Therefore, the FMEA
could help managers assess the risks of failures and provide the
managers with guidelines for improvement. After the system was
improved, a reevaluated version of the FMEA could be imple-
mented. New RPNs of failures would be generated. The cycle would
continue until the system reached a level of low or acceptable risk
ranges. Except for the FMEA applications in the airplane industry,
the use of the FMEA has been introduced to other many industries,
with the notable exception of airlines (Almannai et al., 2008; Pillay
& Wang, 2003).
3. The case company
The case company is a well-known comprehensive semiconduc-
tor manufacturing service provider which offers engineering tests,
package design, integrated circuit (IC) assembly (or packaging),
wafer probing, nal test, and design manufacturing services. With
these manufacturing capabilities, the case company can provide its
customers with complete semiconductor turnkey solutions. The
main business services of the case company are IC and system ser-
vices. For the IC services, the case company offers substrate design
and manufacturing, engineering test, wafer probing, nal test,
package and module design, wafer bumping, and chip packaging
(or assembly). For the system services, the case company offers
turnkey solutions of module-to-systems products.
In order to provide fast services for global customers, the case
companys subsidiaries are based near their overseas customers
in South Korea, Japan, Singapore, Malaysia, China, the Americas,
and Europe. Since the semiconductor industry is a high-tech indus-
try, the case company invests heavily resources in research and
development (R&D). The case company also invests in equipment
and state-of-art facilities in order to satisfy its customers
demands.
The case companys main raw materials are lead frames, IC sub-
strates, epoxy, molding compounds, gold wires, and solder balls.
Raw material costs account for approximately 50% of the total
manufacturing cost of the case company. Except for the molding
compounds, the price of the other raw materials is closely related
to the price of the industrial and precious metals. However, in
2009, prices of the industrial and precious metals increased by
99.43% and 59.5%, respectively. This indicated that manufacturers
P.-S. Chen, M.-T. Wu/ Computers & Industrial Engineering 66 (2013) 634642 635
needed to pay almost twice as much for these raw materials. The
impact was huge for the case companies. Due to much higher pro-
duction cost, the case company incurred a signicant loss of prot.
In addition, most of the case companys suppliers are in foreign
countries. Thus, the case company faces some supply chain risks in
securing its raw materials. For example, the case company has
trouble lling its urgent orders by air and with its products cost
and quality. This research therefore proposes the MFMEA method
as a solution.
4. A decision-making framework for supplier selections
In the process of supplier selections, suppliers needs and risks
need to be considered. This research proposes a ve-step supplier
selection framework (Fig. 1).
Step 1: Identifying the needs of potential suppliers.
This project begins when the manager of the procurement
department considers selecting a new lead-frame supplier in order
to reduce the supply chain risk of the case company. The manager
and his or her subordinates should discuss key or requisite charac-
teristics for suppliers and survey available candidates A, B, and C
in the market.
Supplier A was established in the 1950s to repair and overhaul
compressors. At rst, Supplier A had a compressor factory for its
machinery division. It established its sporting goods division in
the 1960s and its precision products division in the 1970s. In the
2000s, Supplier A established its electronics division. Supplier A
has subsidiaries in China (Zhongshan and Beijing), Vietnam, and
the US (Los Angeles and St. Louis). The main products of Supplier
A are light-emitting diode (LED) lead frames, IC lead frames, and
precision molds.
Supplier B is a subsidiary of a Japanese company which is a
world-class supplier for IC lead frames. In the 1990s, Supplier B
established an Asia Pacic regional operations center in Singapore.
Supplier B has subsidiaries in Asian countries. Supplier B has plants
in Malaysia, Taiwan, Thailand, Indonesia, and China (Chengdu and
Soochow). Supplier B has advanced manufacturing technology. The
main products of Supplier B are lead frames and exible print cir-
cuit (FPC).
Supplier C is a Germany company which was funded in the
1950s. Supplier C providers high-quality, high-precision intercon-
nect materials to customers in the semiconductor or electronics
industry. Product categories of Supplier C are semiconductor lead
frame services, such as offering ICs lead frames, discrete and power
lead frames; connector services, such as offering customization
services of plating and stamping technologies to electronics com-
panies; and smartcard services, such as offering interconnects to
the nancial/banking security, and communication industries.
Supplier C has factories in Singapore, Malaysia, Hong Kong, China
(Shenzhen), France, the Netherlands, and the US (South Carolina
and Illinois).
Step 2: Establishing a team for supplier selections.
Once the needs of the new suppliers have been identied, the
case company creates a cross-functional team to assist the pro-
curement department. This team is comprised of experts from
the R&D, procurement, quality assurance, and engineering
departments.
Step 3: Determining criteria for supplier selections.
This case company considers not only the traditional criteria,
such as quality, cost, deliverability, and service, but also the prac-
tical criteria, such as technology and productivity. The reasons for
adding these two practical criteria are as follows.
According to the literature, technology and productivity are
essential for supplier selection. For example, Weber, Current, and
2. Establishing a team for supplier
selections
1. Identifying the needs of potential
suppliers
3. Determining criteria for supplier
selections
6 criteria and 17 sub-criteria
4. Conducting the MFMEA method
Qualified
suppliers?
Recommendations for needs-
improvement and unqualified suppliers
5. Categorizing qualified suppliers
Selecting a new lead-frame supplier
Staff of R&D, procurement, quality assurance,
and engineering departments
Modified FMEA and AHP
No
Yes
Fig. 1. A decision-making framework of supplier selections.
636 P.-S. Chen, M.-T. Wu/ Computers & Industrial Engineering 66 (2013) 634642
Benton (1991) work provided an overview of issues of supplier
selection up to 1991. The basic attributes were price, delivery,
quality, production facilities and capacity, geographical location,
and technical capability. Production facilities and capacity refers
to productivity; technical capability refers to technology. Ho
et al. (2010) reviewed on articles about supplier selection from
2000 until 2008 and concluded that the most popular criteria con-
sidered by the decision makers were quality (68 papers), price/cost
(63 papers), manufacturing capability (39 papers), service (35 pa-
pers), management (25 papers), and technology (25 papers). Here,
manufacturing capability means productivity. As a result, produc-
tivity and technology are important criteria in the literature on
supplier selection.
Further, the case company belongs to a high-tech industry. For
the technology criterion, the IC assembly industry is very compet-
itive and requires a long-term investment in R&D for the whole
supply chain members. Otherwise, lagging behind other competi-
tors may diminish the competitive advantage of the case company.
For the productivity criterion, the domestic semi-conductor indus-
try emerges due to the need of a exible (or agile) productivity. In
the past, the integrated device manufacturers (IDM) viewed the IC
design companies as competitors rather than as customers. Unless
the IC design companies have an extra demand, the IDMs will not
take productivity expansion into consideration. To solve this prob-
lem, the domestic semi-conductor industry allows the IC design
companies to provide manufacturing services without competition
in the industry. This differential strategy has contributed to the
success of the semi-conductor industry.
Consequently, technology and productivity are essential in the
semi-conductor supply chain. This case company considers these
practical criteria into its supplier selection.
Step 4: Conducting the MFMEA method.
Before conducting the MFMEA method, the team used surveys,
approval of written documents, and eld study to collect informa-
tion on the Suppliers A, B, and C. Then, the MFMEA method ana-
lyzes the information. The steps of the MFMEA method are as
follows:
1. Dening a scoring system for the supplier selection: The pro-
curement department is in charge of evaluating cost, deliver-
ability, and productivity of the supplier selection; the quality
assurance department is responsible for assess the quality of
new suppliers; the R&D department rates the technology score;
and the service score is rated by all departments.
2. Identifying potential failure modes: The project team deter-
mines potential failure modes based on each score. For the tech-
nology criterion, the case company nds that some suppliers
have less design capability than their competitors do. This will
be viewed as a failure mode for the candidate supplier.
3. Evaluating the impact of a failure mode and justifying its sever-
ity: If a candidate supplier has a weak design capability, it will
be not able to produce some components of products that are
requested by the case company. Since the market of the IC
industry changes when new products or technologies emerge,
all members of the supply chain should have a strong technol-
ogy capability. Therefore, the project teamneeds to evaluate the
effect (or severity) of the failure caused by the supplier weak
design capability and give it a score from 1 (lowest) to 4 (high-
est) points, as shown in Table 1.
4. Determining factors that cause to failures and classifying the
probability of occurrence of the failure: For example, high raw
material price offered by a supplier may be attributed to its high
direct labor costs. Then, the project team needs to evaluate how
frequently the raw material price offered by the supplier is
higher than that offered by other competitors. The score can
be classied into four levels from 1 (rarely) to 4 points (most
often), as shown in Table 2.
5. Delineating the detection of each failure and scoring each sub-
criteria: After the four steps have been carried out, the project
team evaluates whether or not the case company has the ability
to detect failures of candidate suppliers, as shown in Table 3.
The results show that the project team can detect all failures
of the candidate suppliers. Therefore, the detection of all sub-
criteria is set to 1.
6. Weighing each criterion based on its importance: This research
adopts the AHP method (Saaty, 1980) to determine the weight
of each criterion. The AHP method is developed by Thomas L.
Saaty in the 1970s and can be applied to solve complex deci-
sion-making problems. The AHP method is based on pairwise
comparisons (a
ij
) of experts opinions to evaluate the
Table 1
Scales for severity.
Severity Rating Descriptions
Remote 1 Cost: A supplier offers cheap raw materials
Quality: A supplier offers few defects among raw
materials
Deliverability: A supplier delivers raw materials with a
slight delay
Technology: A supplier has a strong capability to solve
technical problems caused by manufacturing the raw
materials
Productivity: A supplier has a strong ability to
manufacture customized raw materials for the case
company
Service: A supplier can handle the case companys
complaint within short time
Low 2 Cost: A supplier offers middle-priced raw materials
Quality: A supplier offers some defects among raw
materials
Deliverability: A supplier delivers raw materials with a
normal delay
Technology: A supplier has a normal capability to solve
technical problems caused by manufacturing the raw
materials
Productivity: A supplier has a normal ability to
manufacture customized raw materials for the case
company
Service: A supplier can handle the case companys
complaint within moderate time
Moderate 3 Cost: A supplier offers high-priced raw materials
Quality: A supplier offers many defects among raw
materials
Deliverability: A supplier delivers raw materials with a
long delay
Technology: A supplier has a low capability to solve
technical problems caused by manufacturing the raw
materials
Productivity: A supplier has a low ability to
manufacture customized raw materials for the case
company
Service: A supplier can handle the case companys
complaint within long time
High 4 Cost: A supplier offers very expensive raw materials
Quality: A supplier offers abundant defects among raw
materials
Deliverability: A supplier delivers raw materials with a
very long delay
Technology: A supplier has a very little capability to
solve technical problems caused by manufacturing the
raw materials
Productivity: A supplier has a very little ability to
manufacture customized raw materials for the case
company
Service: A supplier can handle the case companys
complaint within very long time
P.-S. Chen, M.-T. Wu/ Computers & Industrial Engineering 66 (2013) 634642 637
importance between two criteria. The meaning of a
ij
refers to
the compared importance of criterion i to criterion j, the ranking
from 9 (highest), 7, 5, 3, to 1 (lowest). After pairwise compari-
sons, the weights of criteria and sub-criteria are calculated by
Eq. (1). However, since there are only two sub-criteria of the
cost and productivity, the weight of each sub-criterion can be
decided by one comparison whether the AHP method is applied
or not.
For the other four criteria, k
max
is calculated by Eqs. (2) and (3).
Consistency index (C.I.) is calculated by Eq. (4), where n is the num-
ber of criteria. The value of random index (R.I.) is based on the
number of criteria. Here, the values 3, 4, and 5 of R.I. are 0.58,
0.90, and 1.12, respectively. Consistency ratio (C.R.) is calculated
by C.I. divided by R.I. The value of C.R. must be less than 0.1, which
means the consistency of comparisons among criteria. According to
the outcomes of Tables 48, all C.R. values are less than 0.1. Fur-
ther, the results of the AHP method are shown in Tables 49.
W
i

Q
n
j1
a
ij

1=n
P
n
i1
Q
n
j1
a
ij

1=n
8i 1; 2; . . . ; n 1
m
i

P
n
j1
w
j
a
ij
w
i
8i 1; 2; . . . ; n 2
k
max

P
n
i1
m
i
n
3
Table 2
Scales for occurrence.
Severity Rating Descriptions
Remote 1 Cost: A supplier rarely offers expensive raw materials
Quality: A supplier rarely offers defects of raw materials
Deliverability: A supplier rarely fails to deliver raw
materials on time
Technology: A supplier rarely cannot solve technical
problems caused by manufacturing the raw materials
Productivity: A supplier rarely cannot manufacture
customized raw materials for the case company
Service: A supplier rarely fails to handle the case
companys complaint within a reasonable time
Low 2 Cost: A supplier sometimes offers expensive raw
materials
Quality: A supplier sometimes offers defects of raw
materials
Deliverability: A supplier sometimes fails to deliver raw
materials on time
Technology: A supplier sometimes cannot solve
technical problems caused by manufacturing the raw
materials
Productivity: A supplier sometimes cannot manufacture
customized raw materials for the case company
Service: A supplier sometimes fails to handle the case
companys complaint within a reasonable time
Moderate 3 Cost: A supplier often offers expensive raw materials
Quality: A supplier often offers defects of raw materials
Deliverability: A supplier often fails to deliver raw
materials on time
Technology: A supplier often cannot solve technical
problems caused by manufacturing the raw materials
Productivity: A supplier often cannot manufacture
customized raw materials for the case company
Service: A supplier often fails to handle the case
companys complaint within a reasonable time
High 4 Cost: A supplier usually offers expensive raw materials
Quality: A supplier usually offers defects of raw
materials
Deliverability: A supplier usually fails to deliver raw
materials on time
Technology: A supplier usually cannot solve technical
problems caused by manufacturing the raw materials
Productivity: A supplier usually cannot manufacture
customized raw materials for the case company
Service: A supplier usually fails to handle the case
companys complaint within a reasonable time
Table 3
Scales for detection.
Detection Rating Probability of detection (%)
for all six criteria
Remote 1 76100
Low 2 5175
Moderate 3 2650
High 4 025
Table 4
Pairwise comparison matrix of the main criteria of supplier selection.
Criterion Cost Quality Deliverability Technology Productivity Service Weight
Cost 1 3 9 1 5 5 0.31
Quality 1/3 1 9 1/5 5 7 0.18
Deliverability 1/9 1/9 1 1/9 1/5 1/5 0.02
Technology 1 5 9 1 5 9 0.38
Productivity 1/5 1/5 5 1/5 1 1 0.06
Service 1/5 1/7 5 1/9 1 1 0.05
k
max
= 6.61 C.I. = 0.122 R.I. for 6 = 1.24 C.R. = 0.098 < 0.1
Table 5
Pairwise comparison matrix of the sub-criteria with respect to quality.
Sub-criterion Input
Quality
control
Manufacturing
capability
Reliability High
yield
control
Weight
Input quality
control
1 1/7 1/7 1/9 0.04
Manufacturing
capability
7 1 1 1/3 0.22
Reliability 7 1 1 1/3 0.22
High yield
control
9 3 3 1 0.53
k
max
= 4.09 C.I. = 0.030 R.I. for 4 = 0.9 C.R. = 0.03 < 0.1
Table 6
Pairwise comparison matrix of the sub-criteria with respect to deliverability.
Sub-criterion Production
cycle
On time
delivery
Delivery lead
time
Weight
Production
cycle
1 1/5 1 0.13
On time
delivery
5 1 7 0.75
Delivery lead
time
1 1/7 1 0.12
k
max
= 3.013 C.I. = 0.006 R.I. for
3 = 0.58
C.R. = 0.01 < 0.1
638 P.-S. Chen, M.-T. Wu/ Computers & Industrial Engineering 66 (2013) 634642
C:I:
k
max
n
n 1
4
7. Calculating the RPN: The RPN is the product of severity, occur-
rence, and detection. For the products total cost of Supplier A,
the severity is 4, the occurrence is 2, the detection is 1, and
RPN is 4

1 = 8. According to the column Weight of Sub-cri-

teria in Table 9, the weight RPN (R
i
) is then calculated by the
RPN multiplied by the weight of sub-criteria. For example, the
weight RPNs of products total cost for Supplier A is
8

0.2573 = 2.0584. The results of each weight RPNs for Suppli-

ers A, B, and C are shown in Tables 1012, respectively.
8. Sorting each criterion by its RPNs: For all sub-criteria, a higher
RPN means that the higher priority of the criterion needs to be
improved. If the RPN of a criterion is greater than or equal to 6,
the specic failure of the candidate supplier requires improving.
Further, the project team proposes some suggestions for the
candidate supplier to x failures, as illustrated in Table 13.
Step 5: Categorizing qualied suppliers.
For evaluating a new supplier, the case company adopts the
average weight RPNs method. The average weight RPNs of Suppli-
ers A, B, and C are 1.03, 0.62, and 1.13. This means that there is less
risk for the case company to purchase Supplier Bs lead frames than
those of the other suppliers. Through step 4, there may be several
qualied suppliers who meet the requirement of the case com-
pany, but their performances are not equal to each other. There-
fore, qualied suppliers have to be categorized according to the
supply risk. These discussions will be described in Section 5.1.
5. Discussions and implications
5.1. Discussions
5.1.1. Analysis on each criterion of suppliers
According to Table 14, the case company can have a good
understanding of each suppliers advantages. Supplier Bs quality,
technology, and productivity scores are less than those of the other
two suppliers. This indicates that Supplier B can provide better
quality, technology, and productivity strategies for the case com-
pany and help the case company gain some competitive advantage
through this supply chain collaboration. Supplier C has the ability
to offer some information and suggestions on cost, deliverability,
and service strategies. Therefore, if the case company adopts the
cost reduction strategy, Supplier C will be the best choice. How-
ever, Supplier C is the weakest in its quality, technology, and pro-
ductivity. In conclusion, the case company should evaluate the six
criteria of suppliers instead of the four criteria in order to nd the
right supplier for its supply chain and to reduce its supply chain
risk.
5.1.2. Categorization of qualied suppliers
Based on the above analysis (Table 14), the case company se-
lects Supplier B which has the lowest supply chain risk. Although
Supplier Bs service score is higher than Supplier As, and Supplier
Table 7
Pairwise comparison matrix of the sub-criteria with respect to technology.
Sub-criterion Design capability Problem solving capability Continuous improvement capability Weight
Design capability 1 5 7 0.73
Problem solving capability 1/5 1 3 0.19
Continuous improvement capability 1/7 1/3 1 0.08
k
max
= 3.065 C.I. = 0.032 R.I for 3 = 0.58 C.R. = 0.056 < 0.1
Table 8
Pairwise comparison matrix of the sub-criteria with respect to Service.
Sub-criterion Complaint processing Response to demands Report generation Weight
Complaint processing 1 5 7 0.73
Response to demands 1/5 1 3 0.19
Report generation 1/7 1/3 1 0.08
k
max
= 3.065 C.I. = 0.032 R.I for 3 = 0.58 C.R. = 0.056 < 0.1
Table 9
Weights of criteria and sub-criteria of supplier selection.
Main
criterion
Weights of
main
criterion
(1)
Sub-criterion Weight of
sub-
criterions
(2)
Weight of
sub-criterion
(3) = (1)

(2)
Cost 0.31 Products total
cost
0.83 0.2573
Cost reduction
plan
0.17 0.0527
Quality 0.18 Input quality
control
0.04 0.0072
Manufacturing
capability
0.22 0.0396
Reliability 0.22 0.0396
High yield
control
0.53 0.0954
Deliverability 0.02 Production
cycle
0.13 0.0026
On time
delivery
0.75 0.0150
Delivery lead
time
0.12 0.0024
Technology 0.38 Design
capability
0.73 0.2774
Problem
solving
capability
0.19 0.0722
Continuous
improvement
capability
0.08 0.0304
Productivity 0.06 Productivity
exibility
0.75 0.0450
Amount of
production
0.25 0.0150
Service 0.05 Complaint
processing
0.73 0.0365
Response to
demands
0.19 0.0095
Report
generation
0.08 0.0040
P.-S. Chen, M.-T. Wu/ Computers & Industrial Engineering 66 (2013) 634642 639
Bs scores on cost, deliverability, and service are higher than Sup-
plier Cs, the average weight RPNs of Supplier B is still the lowest.
This indicates that collaborating with Supplier B enables the case
company to operate in the low-risk supply chain environment.
As a result, this research recommends that the case company coop-
erate with Supplier B.
In addition, considering the needs of the case company, this pro-
ject team determines the following selection criteria. The average
weight RPNs of qualied suppliers must be lower than 0.75. Suppli-
ers with scores between 0.76 and 1.25 are categorized as suppliers
that needimprovement withina limitedtime. Suppliers whoseaver-
age weight RPNs are more than 1.26 will be disqualied. Therefore,
Suppliers A, B, and C are categorized as the needs-improvement,
qualied, and needs-improvement, respectively.
Meanwhile, from the suppliers perspective, suppliers can rec-
ognize the needs of their customers and improve their competitive
advantage from those with higher RPNs. In order to solve problems
which cause the higher RPNs, suppliers need to adjust their strat-
egies. After doing so, candidate suppliers will prevent their cus-
tomers from entering the high-risk supply chain environment.
5.2. Managerial implications
The following are some implications pertaining to the supplier
selection framework, supplier selection management, and practical
supplier procurement and management.
5.2.1. Supplier selection framework
In the literature, most of the traditional supplier selection sys-
tems focus on cost, quality, deliverability, and service. Based on
Table 4, the weight of six criteria cost, quality, deliverability,
technology, productivity, and service are 0.31, 0.18, 0.02, 0.38,
Table 11
Weighted RPNs of Supplier B.
Supplier B Risk assessment
Criterion Sub-criterion S O D RPN Sub-criterions weight (W
i
) Weighted RPN (R
i
)
Cost Products total cost 2 2 1 4 0.2573 1.0292
Cost reduction plan 3 1 1 3 0.0527 0.1581
Quality Input quality control 2 2 1 4 0.0072 0.0288
Manufacturing capability 1 1 1 1 0.0396 0.0396
Reliability 1 2 1 2 0.0396 0.0792
High yield control 2 2 1 4 0.0954 0.3816
Deliverability Production cycle 2 1 1 2 0.0026 0.0052
On time delivery 1 3 1 3 0.0150 0.0450
Delivery lead time 2 2 1 4 0.0024 0.0096
Technology Design capability 1 3 1 3 0.2774 0.8322
Problem solving capability 2 1 1 2 0.0722 0.1444
Continuous improvement capability 3 4 1 12 0.0304 0.3648
Productivity Productivity exibility 1 3 1 3 0.0450 0.1350
Amount of production 3 2 1 6 0.0150 0.0900
Service Complaint processing 3 3 1 9 0.0365 0.3285
Response to demands 2 1 1 2 0.0095 0.0190
Report generation 2 1 1 2 0.0040 0.0080
Total = 1 Total = 3.70
Average = 0.62
Table 10
Weighted RPNs of Supplier A.
Supplier A Risk assessment
Criterion Sub-criterion S O D RPN Sub-criterions weight (W
i
) Weight RPN (R
i
)
Cost Products total cost 4 2 1 8 0.2573 2.0584
Cost reduction plan 2 2 1 4 0.0527 0.2108
Quality Input quality control 2 1 1 2 0.0072 0.0144
Manufacturing capability 2 2 1 4 0.0396 0.1584
Reliability 1 2 1 2 0.0396 0.0792
High yield control 2 2 1 4 0.0954 0.3816
Deliverability Production cycle 1 1 1 1 0.0026 0.0026
On time delivery 2 2 1 4 0.0150 0.0600
Delivery lead time 3 2 1 6 0.0024 0.0144
Technology Design capability 3 3 1 9 0.2774 2.4966
Problem solving capability 1 1 1 1 0.0722 0.0722
Continuous improvement capability 2 1 1 2 0.0304 0.0608
Productivity Productivity exibility 4 2 1 8 0.0450 0.3600
Amount of production 1 1 1 1 0.0150 0.0150
Service Complaint processing 2 2 1 4 0.0365 0.1460
Response to demands 2 2 1 4 0.0095 0.0380
Report generation 2 4 1 8 0.0040 0.0320
Total = 1 Total = 6.20
Average = 1.03
Note: S means Severity; O means Occurrence; and D means Detection.
640 P.-S. Chen, M.-T. Wu/ Computers & Industrial Engineering 66 (2013) 634642
0.06, and 0.05. The important criterion for the case company is
technology, followed by cost, quality, productivity, service, and
deliverability. Therefore, the cost, quality, and technology are
essential criteria for the case company, which is in the high-tech
industry, to select the right supplier. Further, since the case com-
pany has high demands for lead frames, suppliers of lead-frames
usually cooperate with the case company and make their deliveries
on time. As a result, deliverability does not have a signicant effect
on supplier selection.
However, this study considers two additional criteria: produc-
tivity and technology. This raises the following question: are there
any differences between using four and six criteria? After the cal-
culation in Table 14, the average weight RPNs of four criteria for
Suppliers A, B, and C are 0.80, 0.53, and 0.45, respectively. There
is no difference between Suppliers A (needs-improvement within
a limited time) and B (a qualied supplier). Nonetheless, Supplier
C is categorized as the best-qualied supplier when using the tra-
ditional four criteria.
If this research adopts the traditional four criteria for evaluating
suppliers, some suppliers, such as Supplier C, will have been cate-
gorized as qualied. Although Supplier C has a strong competitive
advantage on the cost strategy, it has adverse effects on quality and
technology strategies. Nonetheless, quality always has a critical ef-
fect on supplier selection and customer satisfaction. Selecting Sup-
plier C as a lead-frame supplier will draw the case company and
supply chain partners into a high-risk supply chain environment.
Hence, the challenges for the case company are to balance trade-
offs among the three key criteria cost, quality, and technology
in order to decrease supply chain risk.
5.2.2. Supplier selection management
When the case company selects a supplier, ve criteria (except
for deliverability) need to be considered. Since each criterion is re-
lated to the others, the traditional FMEA cannot predict the total
supply chain risk. Therefore, this study integrates the risks of each
criterion and sub-criterion into the MFMEA model. More speci-
cally, this supplier selection system considers each elements RPN
(each criterion has several elements to be measured), each crite-
rions RPN, as well as weight of criteria and sub-criteria.
In this research, Supplier B is categorized as a qualied supplier.
However, Supplier Bs performance shouldbe reviewedperiodically.
Suppliers with good performance need to be encouraged, and sup-
pliers with poor performance must be warned or eliminated.
Moreover, Suppliers A and C are rated as a needs-improvement
supplier within a limited time. This supplier has to submit an
improving proposal to the case company within the dened time.
The report will be reviewed by the supplier selection team. If the
report meets the case companys requirements, Supplier A or C
can become a qualied supplier. Otherwise, Supplier A or C will
be disqualied.
Table 13
Recommendations for Supplier A.
Sub-criterion RPN P6 Recommendations for Sub-criterion
with RPN P6
Design capability 9 Increase the number of the R&D
personnel and strengthen R&D on
developing diversied products
Products total cost 8 The products total cost of Supplier A is
higher than the industry average.
Therefore, Supplier A should have better
cost management on its products
Production exibility 8 Increase the diversication of products
Report generation 8 Collect more data on production and
quality
Delivery lead time 6 Shorten order processing time
Table 14
The comparisons among three suppliers.
Weight RPNs of Each Criterion
Criterion Supplier A Supplier B Supplier C
Cost

2.27 1.19 0.73

Quality

0.63 0.53 0.82

Deliverability

0.08 0.06 0.05

Technology 2.63 1.34 4.38
Productivity 0.38 0.23 0.60
Service

0.22 0.36 0.20

The average RPNs of six criteria 1.03 0.62 1.13
The average RPNs of four criteria (

) 0.80 0.53 0.45

Note: The bold and italic score represents the lowest score among three suppliers.
Table 12
Weighted RPNs of Supplier C.
Supplier C Risk assessment
Criterion Sub-criterion S O D RPN Sub-criterions weight (W
i
) Weighted RPN (R
i
)
Cost Products total cost 2 1 1 2 0.2573 0.5146
Cost reduction plan 2 2 1 4 0.0527 0.2108
Quality Input quality control 3 2 1 6 0.0072 0.0432
Manufacturing capability 2 2 1 4 0.0396 0.1584
Reliability 2 3 1 6 0.0396 0.2376
High yield control 2 2 1 4 0.0954 0.3816
Deliverability Production cycle 2 2 1 4 0.0026 0.0104
On time delivery 2 1 1 2 0.0150 0.0300
Delivery lead time 2 2 1 4 0.0024 0.0096
Technology Design capability 4 3 1 12 0.2774 3.3288
Problem solving capability 4 3 1 12 0.0722 0.8664
Continuous improvement capability 3 2 1 6 0.0304 0.1824
Productivity Productivity exibility 3 4 1 12 0.0450 0.5400
Amount of production 2 2 1 4 0.0150 0.0600
Service Complaint processing 2 2 1 4 0.0365 0.1460
Response to demands 2 2 1 4 0.0095 0.0380
Report generation 2 2 1 4 0.0040 0.0160
Total = 1 Total = 6.77
Average = 1.13
P.-S. Chen, M.-T. Wu/ Computers & Industrial Engineering 66 (2013) 634642 641
Finally, for an unqualied supplier, a supplier can improve its
high-risk elements and adjust its strategies based on the needs of
the case company. The report that the supplier has been improved
is sent to the supplier selection team for review. If the report meets
the case companys requirement, the unqualied supplier can be
reviewed during the supplier selection process.
5.2.3. Practical supplier procurement and management
Minimizing cost is always the main consideration of the pro-
curement department in selecting a supplier. This may lead suppli-
ers to reduce costs and at the same time decrease their product
quality and R&D capability. Thus, it is very common to hear that
when the quality and assurance personnel requests the suppliers
to provide better quality products, or when the engineers ask the
suppliers for better technology, they usually fail to do so.
In brief, this study recommends the case company to take the
supplier with the lowest average RPN value into account. If a com-
pany adopts the minimumcost strategy as the most important sup-
plier criterion, it may bring the holistic supply chain members into a
riskier environment. Therefore, the concept of the minimal cost
strategy should be replaced by the minimal supplier-risk strategy.
6. Conclusions
This research makes three contributions. First, it develops a
supplier-selection framework (Fig. 1), which incorporates the
MFMEA and AHP methods. The proposed MFMEA method selects
six criteria and 17 sub-criteria to assess the RPNs of three lead-
frame suppliers. Then, this study applies the AHP method to deter-
mine the weight of each criterion and sub-criterion. As a result,
suppliers can be categorized as needs-improvement, qualied,
and unqualied based on their average weight RPNs. For the
needs-improvement suppliers, the case company makes valuable
suggestions so that they can turn their weaknesses into strengths.
The MFMEA method can help companies select the right supplier
with the lowest risk in order to reduce the holistic supply chain
risk.
Second, for the criteria of supplier selection, the four essential
criteria in the literature are cost, quality, deliverability, and service.
According to these criteria, a supplier with the lower cost strategy
and less advanced technology strategy may be viewed as the pre-
ferred supplier. For the high-tech industry, this will create a high-
risk supply chain. After adding technology and productivity, the
case company can select the right supplier with the lowest supply
risk. The results of this research suggest that the case company
needs to consider technology for supplier selection, but omit
deliverability.
Finally, this research uses a case study to verify the feasibility of
the proposed supplier-selection framework and MFMEA method.
Section 5.1 summarizes the discussions of supplier selection. Fur-
ther, the managerial implications of the case study in Section 5.2
can help companies in making their future supplier-selection
decisions.
Despite the contributions, this study has several limitations.
First, when different enterprises select suppliers, their viewpoints
may depend on their supply chain environment. Even though this
research studies the IC assembly industry, other similar industries
may not consider it in the same way. In addition, this MFMEA is re-
stricted by the information provided by the case company, and
therefore there are six criteria for supplier selection, which may
not be sufcient for other companies. Other considerations may
need to be factored into future analyses.
Finally, this research suggests two directions for future
research:
1. Suppliers evaluative criteria play important roles in the supplier
selectionprocess. Howtochooseappropriate suppliers criteriais
complicated. Future researchers should pay more attention to
the criteria of supplier selection for different industries.
2. Other than the FMEA, other approaches, such as the Delphi
method, fault tree analysis, and stress testing are often used
for supplier selection. Other approaches for the risk assessment
can be applied to see if they can generate similar results in the
future.
References
Almannai, B., Greenough, R., & Kay, J. (2008). A decision support tool based on QFD
and FMEA for the selection of manufacturing automation technologies. Robotics
and Computer Integrated Manufacturing, 24, 501507.
Che, Z. H., & Wang, H. S. (2008). Supplier selection and supply quantity allocation of
common and non-common parts with multiple criteria under multiple
products. Computers & Industrial Engineering, 55(1), 110133.
Chen, L.-H., & Ko, W.-C. (2009). Fuzzy linear programming models for new product
design using QFD with FMEA. Applied Mathematical Modelling, 33(2), 633647.
Chin, K.-S., Wang, Y.-M., Poon, G. K. K., & Yang, J.-B. (2009). Failure mode and effects
analysis by data envelopment analysis. Decision Support Systems, 48(1),
246256.
Dickson, G. (1966). An analysis of supplier selection systems and decisions. Journal
of Purchasing, 2(1), 517.
Ekmekcioglu, M., & Kutlu, A. C. (2012). A fuzzy hybrid approach for fuzzy process
FMEA: An application to a spindle manufacturing process. International Journal
of Computational Intelligence Systems, 5(4), 611626.
Hallikas, J., Karvonen, I., Pulkkinen, U., Virolainen, V.-M., & Tuominen, M. (2004).
Risk management processes in supplier networks. International Journal of
Production Economics, 90(1), 4758.
Ho, W., Xu, X., & Dey, P. K. (2010). Multi-criteria decision making approaches for
supplier evaluation and selection: A literature review. European Journal of
Operational Research, 202(1), 1624.
Katsikeas, C. S., Paparoidamis, N. G., & Katsikea, E. (2004). Supply source selection
criteria: The impact of supplier performance on distributor performance.
Industrial Marketing Management, 33(8), 755764.
Ko, W. C. (2013). Exploiting 2-tuple linguistic representational model for
constructing HOQ-based failure modes and effects analysis. Computers &
Industrial Engineering, 64(3), 858865.
Liu, H. C., Liu, L., & Liu, N. (2013). Risk evaluation approaches in failure mode and
effects analysis: A literature review. Expert Systems with Applications, 40(2),
828838.
Neiger, D., Rotaru, K., & Churilov, L. (2009). Supply chain risk identication with
value-focused process engineering. Journal of Operations Management, 27(2),
154168.
Ng, W. L. (2008). An efcient and simple model for multiple criteria supplier
selection problem. European Journal of Operational Research, 186(3), 10591067.
Pillay, A., & Wang, J. (2003). Modied failure mode and effects analysis using
approximate reasoning. Reliability Engineering & System Safety, 79(1), 6985.
Saaty, T. L. (1980). The analytic hierarchy process. New York: McGraw Hill.
Sawik, T. (2011). Supplier selection in make-to-order environment with risks.
Mathematical and Computer Modelling, 53(910), 16701679.
Schoenherr, T., Rao Tummala, V. M., & Harrison, T. P. (2008). Assessing supply chain
risks with the analytic hierarchy process: Providing decision support for the
offshoring decision by a US manufacturing company. Journal of Purchasing and
Supply Management, 14(2), 100111.
Shu, M. H., & Wu, H. C. (2009). Quality-based supplier selection and evaluation
using fuzzy data. Computers & Industrial Engineering, 57(3), 10721079.
Srinivasan, M., Mukherjee, D., & Gaur, A. S. (2011). Buyer-supplier partnership
quality and supply chain performance: Moderating role of risks, and
environmental uncertainty. European Management Journal, 29(4), 260271.
Tang, C. S. (2006). Perspectives in supply chain risk management. International
Journal of Production Economics, 103(2), 451488.
Thun, J.-H., & Hoenig, D. (2011). An empirical analysis of supply chain risk
management in the German automotive industry. International Journal of
Production Economics, 131(1), 242249.
Tracey, M., & Tan, C. (2001). Empirical analysis of supplier selection and
involvement, customer satisfaction, and rm performance. Supply Chain
Management: An International Journal, 6(4), 174188.
Tseng, M.-L., Chiang, J. H., & Lan, L. W. (2009). Selection of optimal supplier in supply
chain management strategy with analytic network process and choquet
integral. Computers & Industrial Engineering, 57(1), 330340.
Weber, C. A., Current, J. R., & Benton, W. C. (1991). Vendor selection criteria and
methods. European Journal of Operational Research, 50(1), 218.
Wu, D., & Olson, D. L. (2008). Supply chain risk, simulation, and vendor selection.
International Journal of Production Economics, 114(2), 646655.
Wu, T., Blackhurst, J., & Chidambaram, V. (2006). A model for inbound supply risk
analysis. Computers in Industry, 57(4), 350365.
642 P.-S. Chen, M.-T. Wu/ Computers & Industrial Engineering 66 (2013) 634642