Omega 57 (2015) 5–21

Contents lists available at ScienceDirect

Omega
journal homepage: www.elsevier.com/locate/omega

A multiple objective optimization based QFD approach for efficient

resilient strategies to mitigate supply chain
vulnerabilities: The case of garment industry of Bangladesh☆,☆$
Md. Maruf Hossan Chowdhury, Mohammed A. Quaddus n
School of Marketing, Curtin University, Perth, WA, Australia

art ic l e i nf o a b s t r a c t

Article history: With the upsurge of frequent disruptive events, organizations have become more vulnerable to the
Received 1 January 2014 consequences of these disruptive events. As a result, the need for more resilient supply chain (SC) to
Accepted 22 May 2015 mitigate the vulnerabilities has become paramount. Supply chain resilience (SCR) has been discussed in
Available online 25 June 2015
the literature and resilience index has been developed, but developing and selecting a portfolio of supply
Keywords: chain resilience capabilities in order to mitigate the vulnerabilities have not been studied. In this
Supply chain research we develop a 0–1 multi-objective optimization model based on QFD methodology. Our multi-
Resilience-efficiency objective method is interactive and interacts with the decision makers to choose the most satisfactory
Vulnerability efficient portfolio of supply chain resilience strategies. We apply our methodology to three large ready-
Multi-objective model
made garment (RMG) companies of Bangladesh. Results show that lack of materials (high dependence
AHP
on imported materials), disruptions in utility supply, increased competition (and hence competitive
QFD
pressure), impact of economic recession, and reputation loss are the top most vulnerabilities of
Bangladesh RMG industry. The most preferred resilience strategies to mitigate the vulnerabilities are:
back-up capacity, building relation with buyers and suppliers, quality control, skill and efficiency
development, ICT adoption, demand forecasting, responsiveness to customers, and security system
improvement. Theoretical and managerial implications of our study are included.
& 2015 Elsevier Ltd. All rights reserved.

Introduction
Ready Made Garment (RMG) industry contributes hugely to
Bangladesh’s economy. It creates more than four million direct
employment and several millions of indirect employment and acco-
unts for 78.6 percent of countries export earnings [5]. RMG sector also
immensely contributes in reducing the high rate of women unem-
ployment in the country as 80 percent of the garments workers are
women [5]. Thanks to the RMG sector, Bangladesh is also the second
largest apparel exporter in the world.
Despite its huge potentials the industry is struggling with numer-
ous Supply Chain (SC) disruptions [49,42]. The consequences of the
disruptions are huge, for example, RMG industry of Bangladesh loses
$26.15 million per day due to problems in SC functions caused by
political instability [1]. Moreover, the preferential access in U.S. market
is cancelled because of the poor safety standard in production plants
as building collapse in garment factory caused the death of more than
☆
It is an equally authored paper. Author’s names are listed in alphabetical order.
☆☆
This manuscript was processed by Associate Editor B. Lev.
n
Corresponding author.
E-mail addresses: marufhossan@gmail.com (Md.M.H. Chowdhury),
m.quaddus@curtin.edu.au (M.A. Quaddus).
eleven hundred workers [37]. These disruptions have chain effect to
all the members in SC network including the international buyers
(retail chains) and suppliers. In the wake of such a critical state in
RMG supply chain, developing resilience capabilities is vital, which is the
primary objective of this study.
Resilience has been defined by a number of authors in a related
manner. Vugrin et al. [99] define system resilience and resilience in
general. The authors highlight that resilience is the ability of a systems
to respond to a ‘disruption’ due to an event or set of events. Along the
same vein Christopher and Peck [24], Ponomarov and Holcomb [81]
and Jüttner and Maklan [53] define supply chain resilience as the
‘capability of the supply chain to responds to disruptions and recover
from them’. On the other hand Pettit et al. [79,80] developed a supply
chain resilience framework by identifying seven categories of vulner-
abilities and creating supply chain capabilities along 14 areas (sour-
cing, order fulfilment, capacity development; among others). The
authors surmise that current level of vulnerabilities and capabilities
must be assessed in order to ascertain the current level of resilience.
Literature emphasizes that developing resilience capability is vital for
organizations. It enables organizations to improve system perfor-
mance [80,99], achieve sustainable competitive advantage [81], gain
market share in competitive environments [90], and decreases
vulnerabilities [53,79,80]. However current literature lacks in

http://dx.doi.org/10.1016/j.omega.2015.05.016
0305-0483/& 2015 Elsevier Ltd. All rights reserved.
6 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21
proposing the ways and means to achieve supply chain resilience
capabilities. In this research we stress that resilience capabilities of RMG
supply chain of Bangladesh must be developed to mitigate organizations
vulnerabilities. In pursuing this research objective we introduce the
concept of ‘supply chain resilience efficiency (SCREF)’ which has
significant pedagogical importance.
It has been established that supply chain resilience capabilities has
multiple dimensions (objectives). For example Pettit et al. [79,80] in
their framework highlight 14 areas of supply chain resilience capabil-
ities to be developed from order fulfilment, capacity development
to financial strengths. From multiple objectives decision making
(MODM) perspectives the supply chain resilience capabilities must
be ‘efficient’ to mitigate vulnerabilities. While literature on efficiency
approach in MODM and its applications in production/operations
management area are plentiful (for example see, [63]; among many
others), the notion of ‘supply chain resilience efficiency (SCREF)’ from
multiple objectives perspective is novel.
In this research we define SCREF as follows:
(i) resilience capability must be resource efficient (e.g. minimum
cost of implementation), and
(ii) portfolio of chosen resilience capabilities must be efficient (or
non-dominated) from multiple objectives perspective [63].
We shall elaborate on (i) and (ii) later. It is observed that a number
of logistics and SC related capabilities are discussed in the literature
(for example; [80,38,81,90,24]; among many others) to develop SC
resilience but most of those are conceptual studies and fall short of
introducing the notion of resilience efficiency. Furthermore, in a state
of uncertainty, dynamic changes and resource limitation, selection of
optimal and efficient portfolio of resilience capabilities has not yet
been addressed adequately in the existing SC literature.
While a number of approaches could be undertaken to achieve the
research objective, this study has adopted Quality Function Deploy-
ment (QFD) [77,102] as a methodology to develop the resilience
capabilities of the RMG supply chain of Bangladesh and find the
optimal efficient portfolio of the resilience capabilities using a non-
linear 0–1 programming approach. Literature on QFD approach is
plentiful, which will be reviewed briefly in a later section. It is suffice
to say that QFD enables organizations to be proactive to vulnerabilities
mitigation rather than reactive and it is a proven technique for
designing supply chain mitigation capabilities in such situations
[34]. However we shall use Analytical Hierarchy Process (AHP) [86]
within QFD for the analysis of data. It is important to note that
methodologically our contribution lies as follows: we define supply
chain resilience efficiency (SCREF) and find portfolio of efficient
resilience capabilities for implementation using multiple objectives
based non-linear 0–1 mathematical program.
It is noted that overall domain of our present study is Enterprise
Risk management (ERM). Enterprise Risk management (ERM) has
been defined in many different ways. However, one common theme
of ERM is that it takes a ‘holistic and strategic’ approach to manage all
risks that an organization faces [32,74]. A recent literature review [20]
has found that Desheng Wu and David Olson are two of the most
dominant contributors on ERM and various aspects of enterprise risk.
One of their highly cited works is the application of ERM to assess
credit worthiness in bank [106]. Wu and Olson [104], Wu et al.
[107,108] have edited special issues of various journals on various
aspects of ERM, ranging from risk methods and tools in operations,
enterprise risk management in operations and business intelligence in
risk management. Various other applications of ERM are available
elsewhere [20] and hence will not be repeated here.
Our present research focusses on resilience capabilities of RMG
supply chain of Bangladesh to mitigate supply chain vulnerabilities.
Literature suggests that vulnerability is an ‘exposure to serious
disturbance arising from risks within supply chain’ [78]. Hence our
present study is highly relevant to the supply chain aspect of
enterprise risk management. There are applications of enterprise risk
management (ERM) in supply chain. For example, Wu and Olson [105]
developed a DEA based value at risk (VaR) model to manage supply
risks, specifically vendor selection problem. Olson and Wu [75]
presented a review of ERM in supply chain. Jiang et al. [50] developed
a LOGIT model of job satisfaction to reduce supply chain risks.
However, it has been mentioned before that we developed a new
approach to find an optimal portfolio of efficient resilience capabilities
to mitigate supply chain vulnerabilities.
In the next several sections we present the state of affairs of
RMG industry of Bangladesh, relevant literature, methodology
(QFD based mixed qualitative–quantitative approach), application
in RMG industry, followed by the results. The paper concludes
with the discussions & implications and conclusions.
Background
Bangladesh is one of the leading exporters of Ready Made
Garments (RMG) in the world. RMG industry is an economic propeller
of Bangladesh and apparel exports stood-up at 19.90 billion US dollar
in 2011 and marked Bangladesh as the second largest apparel exporter
in the world [5]. Because of enormous economic importance of RMG
in the economy of Bangladesh, smooth and efficient functioning of
supply chain activities is crucial. But, the RMG supply chain is facing a
climax situation owing to numerous challenges, such as, labour unrest
for violation of human rights, poor wages, poor and hazardous
working environment, political instability, interruption in utility
supply especially power shortage, inefficiency in customs and port
management, exchange rate fluctuation, disruption in timely supply of
fabrics and other accessories, increased competition, inefficiency in
operations, intensive competitive pressure, strict compliance code
regarding social and environmental issues; among many others
[49,22,42]. Furthermore, increased lead time and cost due to disrup-
tions in procurement and shipment of goods, lack of linkages and co-
ordination among related industries in the value chain, dependence
on imported inputs, limited variety of finished products [42], fall of
order because of global economic downturn are also issues of high
concern for the RMG supply chain of Bangladesh. As a result of these
disruptions the growth of RMG export from Bangladesh has fallen
from 23% in 2005–2006 to 15% in 2008-2009 [22]. In such a situation
it is crucial to find ways and means to make RMG supply chain
resilient and sustainable. Previous researchers focused mainly on RMG
competitiveness, the existing problems and challenges of the industry.
Table 1 summarizes these briefly. However, the issue of making RMG
supply chain resilient and efficient has not been investigated yet. This
study aims to fill this gap in the literature.
Literature review
Supply chain vulnerabilities
Maintaining an effective Supply Chain (SC) has become challenging
and difficult as the supply chains are inherently complex and in recent
times are overwhelmed with disruptive events. These disruptive
events make a supply chain vulnerable, as supply chain vulnerability
is the susceptibility of the supply chains to the consequences of
disruptive events [8,53]. Wagner and Neshat [101] posit that supply
chain vulnerability is determined by the vulnerability drivers arising
from demand side, supply side and supply chain design issues.
Similarly, supply chain vulnerability may also arise from a number
of factors such as, delay during transportation, port stoppages,
frequent occurrence of natural disasters, weak communication, supply
shortages, demand volatility, quality problem, operational issues and
 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21 7

Table 1
Vulnerability factors and the mitigation capabilities in RMG industry of Bangladesh.

Khondker et al. [60] Deals with important issues and challenges facing the RMG industry during the post-MFA regime. It also discusses the competitiveness issue in
relation to productivity, working environment and stakeholders.
Choudhury and Hossain The authors have taken step to define the challenges and opportunities in the post-MFA period in the RMG sector and indicated the
[21] government response to combat post-MFA challenges.
Karim [55] Discusses how the Bangladeshi RMG sector coped startegically with the post-MFA challenges.
Tewari [97] Emphasized on timely supply, short lead times, low inventories, innovation and the ability to contribute to design and full package supply.
Ferdousi and Ahmed Investigated the improvement of manufacturing performance through lean practice that helps to reduce lead time and cost and to improve
[36] productivity.
Hossan et al. [47] Discuss the recent unrest in the RMG industry which indicates political action in this industry.
Islam et al. [48] Operational disturbances, manufacturers are facing competition with respect to quality, cost and time to market.
Nuruzzaman et al. [73] Described how to create competitive advantage through SCM.
Chowdhury et al. [23] Addressed supply chain disruptions and mitigation processes
Chowdhury et al. [23] Dealt with upstream supply chain barriers and mitigation processes
Haider [42] Presents the challenges and the surface-level and deep-level competitiveness of the Bangladeshi RMG industry.

terrorism; among many others [25,61,7]. Schmitt et al. [88] study the
impact of supply disruptions on both centralized and the decentra-
lized multi-location system. Sawik [87] presents a bi-objective (mini-
mum cost and maximum service level) model to study the impact of
supplier disruptions Mizgier et al. [71] show the far-reaching effect of
disruptions in a SC network and the impact on the performance of
overall system. The studies by Hendricks and Singhal [45] show that
announcement of SC disruptions, such as, operational issue or delay in
shipment cause decrease in shareholder value significantly. Kleindor-
fer and Saad [61] identify three main sources of SC vulnerability:
firstly, operational factors which include equipment malfunctions and
systemic failures, abrupt discontinuity of supply, labour strikes, among
others; secondly, natural hazards which include earthquakes, hurri-
canes, storms; and thirdly, terrorism or political instability, among
others. Blos et al. [8] suggest four major sources of SC vulnerabilities
such as, financial vulnerability, strategic vulnerability, hazard vulner-
ability, and operations vulnerability. Similarly, a number of researchers
(such as [79,24,90,7,61]; among others) discussed SC vulnerability
factors which are summarized in Table 2 in terms of various vulner-
ability factors. However, these studies did not deal with strategies and
capabilities to mitigate SC vulnerabilities. Furthermore, most of the
studies are conceptual rather than empirical in nature. The research
reported in this paper prioritizes the existing vulnerabilities and finds
the efficient mitigation strategies and capabilities in the context of
RMG supply chain of Bangladesh by using AHP integrated QFD
approach. It thus addresses a specific gap in the existing literature.
Supply chain resilience capability
cohesion, control, connectedness to measure resilience [80,81,98].
Table 3 summarizes these capabilities. It is noted that two funda-
mental organizational resilience capabilities are buffering and bridging
[9,35]. All others are related to these two fundamental strategies.
Buffering is external to a current relationship with a supply chain
partner and acts as ‘shock absorbers’ to mitigate the detrimental
consequences [9]. For example the capability factors of ‘flexibility’,
‘reserve/backup capacity’ of Table 3 fall in this category. On the other
hand bridging is internal to a current relationship and refers to
strengthening the current relationship via formal structure [9]. The
capability factors of ‘integration’, ‘efficiency’ of Table 3 fall in this
category. The resilience capability needed by a system depends on
context, extent and type of vulnerabilities [16]. Therefore, in order to
deal with resilience it is important to identify the vulnerability factors
of the specific SC and the corresponding mitigation capabilities.
Along with developing resilience it is also important to measure
resilience to ensure a better resilience outcome. In the literature,
resilience is measured in a number of ways: (i) based on the extent of
systems departure from desired state [46,68], (ii) based on recovery
time after disaster or disruptions [92,90], (iii) based on reduction of
impact and consequences [85,67], (iv) based on time to respond [90]
and (v) based on cost of recovery [69,99].
Once the resilience capabilities are designed and measured, it is
also important to determine the efficiency of the resilience capability
for reducing the impact of vulnerabilities [99]. Literature lacks sign-
ificantly in conceptualizing resilience efficiency. In this paper, there-
fore, we offer a unique and elegant operationalization of resilience
efficiency in the methodology section.

Resilience is a multidisciplinary concept. Holling [46] was one
of the pioneers to conceptualize resilience ‘as the ability of system
to absorb changes’. Since then many authors echoed the concept of
resilience as system’s ability to recover and get back to the original
state [70,81,24]. Heckmann et al. [44] however mention that
supply chain resilience must have the ability to ‘overcome supply
chain vulnerability and to reduce supply chain risk’. In line with
extant literature [81,44], in this paper, we define supply chain
resilience (SCR) as the capability of a supply chain to reduce the
impact of vulnerabilities (due to disruptions) through developing
required level of readiness, quick response and recovery ability.
Vulnerabilities in the SC are sometimes beyond the direct control
of SC managers. However, SC managers need to be proactive to
predict the vulnerability factors in advance and develop resilience
capacity for mitigating the vulnerabilities [53]. Otherwise, the
consequence will be the discontinuity of SC operations which will
adversely affect both revenue and cost of the whole chain [81].
Researchers in supply chain management (SCM) emphasized on
capabilities, such as, adaptability, pro-activeness, diversity, flexibility,
efficiency, reserve capacity, integration, market development,
Quality function deployment (QFD)
QFD is a systematic process used by cross-functional teams to
identify and resolve the issues involved in providing products,
processes, services, and strategies that enhance customer satisfaction
[41]. The benefits of QFD model have been highlighted by various
researchers. For example Chan and Wu [17] in their review of QFD
theory and applications have noted wide range successful applications
of QFD from product development, customer needs analyses to
decision making. Carnevalli and Miguel [15] in another review of
QFD highlight QFD’s ability in adapting into various research methods
from modelling, theoretical-conceptual to action-research, experimen-
tal. Because of its wide applicability QFD has been used in various
fields, such as, determining customer needs [94], developing priorities
[43], manufacturing strategies [29,52], logistics and SCM [10,4]. QFD
has also been applied successfully for supply chain risk identification
and mitigation [82,34]. In line with previous literature we have
adapted QFD methodology to identify supply chain vulnerabilities
and mitigate the vulnerabilities with optimal and efficient resilience
capabilities.
8 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21

Table 2
Supply chain vulnerability factors.

Vulnerability factors Specifc vulnerabilities References

Hazard vulnerability Natural disaster (flood, cyclone) Christopher and Peck [24]; Sheffi and Rice [90]; Kleindorfer and Saad [61]; Wu et al.
[109]; Blackhurst et al. [7]
Political instability Kleindorfer and Saad [61]; Wu et al. [109]; Blackhurst et al. [7]; Blos et al. [8]
Fire and other accidental damage Blos et al. [8]
Labour unrest Sheffi and Rice [90]; Kleindorfer and Saad [62]; Wu et al. [109]; Blackhurst et al. [7]

Strategic vulnerability Increased competition Haider [42]; Blos et al. [8]

Non-compliance of social and environmental Islam and Deegan [49]
factors
Problem of relation with buyer (switching of buyer) Blos et al. [8]
Problem of integration and real-time information Gaudenzi and Borghesi [39]
Problem of relation with supplier Blos et al. [8]
Plant location problem Field study

Financial vulnerability Currency fluctuation Blos et al. [8]; Blackhurst et al. [7]
Economic recession Blos et al. [8]
Raw material price fluctuation Blos et al. [8]
High bank interest & fund shortage Blos et al. [8]
Bankruptcy or credit default of any supply chain Blos et al. [8]; Blackhurst et al. [7]
member

Operational Shortage of skilled worker Haider [42]

vulnerability Switching and absenteeism of workers Chowdhury et al. [22].
Fault in production planning and inventory Chowdhury et al. [22]; Wu et al. [109]
management
Failure of IT system and machineries Blos et al. [8]
Disruption in utility supply Blos et al. [8]
Product quality defection (poor quality) Blos et al. [8]
Illiteracy of workers and supervisors Chowdhury et al. [22]

Infrastructure Delay in custom clearance Colicchia [25]

vulnerability Delay for Congestion and inefficiency in port Colicchia [25]; Blackhurst et al. [7]
Strike by port workers Colicchia [25]; Blos et al. [8]
Delay in transportation for Poor infrastructure Blackhurst et al. [7]

Demand & supply Suppliers’ delay Blackhurst et al. [7]

vulnerability Dependence on imported material and lack of Haider [42]; Nuruzzaman and Rafiq [73]; Craighead et al. [26]
backward linkage
Lack of alternative for some critical items Craighead et al. [26]
Defection or nonconformity of material Blackhurst et al. [7]
Opportunism of buyers (expect discount) Ponomarov and Holcomb [81]
Demand fluctuation/uncertainaty Wu et al. [109]
Suppliers opportunism Ponomarov and Holcomb [81]

In QFD modelling, ‘customer requirements’ or existing problems of
the organizations (for example vulnerabilities) are referred to as
WHATs and ‘how to fulfil the customer’s requirements’ or organiza-
tional problems are referred to as HOWs (resilience capabilities). The
basic QFD framework is shown in Fig. 1 where CRi and DRj are the
HOWs and WHATS respectively. The process of using appropriate
HOWs to meet the given WHATs is represented in the relationship
matrix (Rij in Fig. 1). Different researchers build different QFD models
involving various elements but the most widely used QFD model
contains at least the requirements/problems (WHATs) and their
relative importance (Wi in Fig. 1), technical measures or design
requirements (HOWs) and their relationships with the WHATs, and
the correlation between the HOWs (see Fig. 1).
In our case row elements CRi (WHATs) of Fig. 1 represent the
vulnerabilities that RMG supply chains are facing currently. The
column elements DRj (HOWs) are the resilience strategies or capabil-
ities to mitigate the vulnerabilities. We define the elements of the
relationship matrix Rij (see Fig. 1) as the ‘extent of mitigating the
specific vulnerabilities by specific resiliency strategies’. In line with
QFD literature [17,82,34] we measure Rij using the scale of 9 (strong
mitigation), 3 (moderate mitigation), 1 (little mitigation) and 0 (no
mitigation). The AI and RI in Fig. 1 are the absolute and relative
importances of the HOWs (resilience strategies) which are found as
follows [77]:
X
m Integration with Analytical Hierarchy Process (AHP)
AI j ¼ wi Rij 8 j ; j ¼ 1; …; n
i¼1
where AIj is the absolute importance of jth design requirement (DR)
(or, resilience strategy), wi is the weight of the ith supply chain
vulnerability. Rij is the relationship value; extent of mitigating ith
vulnerability by jth resilience strategy (9, 3, 1, or 0), n is the number of
design requirements (resilience strategies); m is the number of supply
chain vulnerabilities.
It is noted that in our case AIj is interpreted as ‘total resilience’
of the jth resilience strategy to mitigate the vulnerabilities.
The relative importance (resilience) of the resilience strategy j
is
AIj
RIj ¼ Pn ð2Þ
j¼1 AI j
The correlation between the DRjs (HOWs) (see Fig. 1) plays a
significant part in many QFD applications including ours. It represents
the extent of correlation (similarities) when two HOWs are imple-
mented. Literature suggests that there is some degree of dependencies
among the HOWs in real applications [103,77]. If HOWi and HOWj are
correlated then there is cost savings sij in their implementation [77].
These sij’s need to be estimated from the decision makers. We argue
that in most business, management and social science applications
(including our present case) some HOWs will be highly correlated.
ð1Þ In basic QFD three fundamental data are needed: (i) relative
importance of WHATs, (ii) the relationships between the WHATs
 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21 9

Table 3
Supply chain resilience capabilities.

Capability Factors Specific capabilities References

Flexibility Flexibility in production (different volume of order, flexible production schedule) Duclos et al. [33]; Braunscheidel and
Suresh [11]
Ability to modify a wide varity of product as per buyer requirement (mix flexibility) Braunscheidel and Suresh [11]
Flexibility in contract with SC partners (Partial order and payment, partial shipment) Duclos et al. [33]
Efficient and effective logistics and supply chain functions (e.g. sourcing, producing, Duclos et al. [33]
distribution)
Ability to respond to additional order or sudden demand Jüttner and Maklan [53]
Ability to supply new and different products to different customer groups (mix flexibility) Braunscheidel and Suresh [11]

Reserve/backup capacity Alternative and reserve capacity (logistical options) Pettit et al. [79]
Buffer stock Pettit et al. [79]
Backup energy source Pettit et al. [79]

Integration Sharing information with supply chain partners Braunscheidel and Suresh [11]; Blackhurst
et al. [7]
Communication and information flow with different departments (e.g. supply chain and Braunscheidel and Suresh [11]
other departments)
Joint or collaborative planning (e.g product development) Braunscheidel and Suresh [11]
Communication with supply chain partners Braunscheidel and Suresh [11]
ICT supported planning and integration Narasimhan and Kim [72]

Efficiency Waste elimination (efficient use of resource) Pettit et al. [79]; Fiksel [38]; Sheffi and
Rice [90]
Efficient and hardworking employees
Quality control and less defection Pettit et al. [79]; Kleindorfer and Saad [61]

Customer satisfaction & market Buyer satisfaction (product quality and service) Pettit et al. [79]
position Customer loyalty and Preference Pettit et al. [79]
Good relation with buyers and suppliers Zsidisin et al. [111]; Pettit et al. [79]

Financial strength Fund availability Pettit et al. [79]; Tang [95]

Profitability Pettit et al. [79]
Insurance Pettit et al. [79]

and HOWs, and (iii) the correlation between the HOWs. In finding the
relative importance of WHATs (i.e.wi’s; see Fig. 1) we plan to use AHP.
AHP was originally developed by Saaty [86] which is an well-
established multi-criteria decision making approach that employs a
unique method of hierarchical structuring of a problem and subse-
quent ranking of alternative solutions by a paired comparison
technique. For brevity full description of AHP process will not be
presented in this paper, which is available elsewhere in the literature
[86]. AHP is frequently used in QFD process, for instance see, Kamvysi
et al. [54], Park and Kim [77], Bhattacharya et al. [6], Chan and Wu
[18]; among others. Methodologically QFD has been frequently
combined with other tools to increase its robustness and applicability.
For example Ramanathan and Yunfeng [84] combined QFD with Data
Envelopment Analysis (DEA) and applied it to design security fasten-
ers in a Chinese company. Lin et al. [66] combined QFD with DEA and
AHP and applied it to evaluate the economic performance of local
governments in China.
functions (customer satisfaction, cost and technicial difficulties)
are converted into goals. Other multiple objectives optimization
based QFD applications also include various variations of goal
program, for example see Karsak and Özogul [58], Chen and Weng
[19], Lee et al. [65], Büyüközkan and Berkol [13]; among others.
Karsak [56,57] on the other hand has applied multiple objective
optimization in QFD and has determined the non-dominated
(efficient) solutions of QFD design requirements (HOWs). The
author however has not gone far enough and developed any
procedure of obtaining efficient design requirements in QFD
applications. In our research we contribute in this aspect and
develop a procedure to obtain efficient design requirements
(efficient resilience capabilities in our case).
Proposed methodology to determine efficient resilience
capabilities in QFD

The concept of efficiency and generating efficient solutions is

QFD optimization
It is noted that in QFD an optimization method is always needed to
determine the most desirable HOWs to satsify the WHATs [31,57]
under certain constraints. For example, Park and Kim [77] optimize
total absolute importance of the HOWs by formulating a 0–1 linear
and quadratic program to find the most desirable HOWs under budget
constraint. Wasserman [103] formulates a linear program to find the
most desirable design requirements, HOWs, under budget constraints.
Zhou [110] proposes a mixed integer linear program that maximizes
an utility function under budget, technological feasibiity and competi-
tion constraints to determine the HOWs.
Multiple objective optimization approaches [62,30] have also
been used in determining the HOWs. For example, Karsak et al.
[59] formulated a 0–1 goal program combining analytical network
process in QFD application in product planning. Delice and Güngör
[31] proposes a mixed integer goal program where the objectives
prevalent in multiple objective decision making domain [62,30]. A
general multiple objective decision problem is represented as
follows:
)
MaxðMinÞ f i ðXÞ ¼ C i ðXÞ; i ¼ 1…p
ð3Þ
Subject to : g j ðX Þ r bj ; j ¼ 1…q
where X¼(x1, x2, …, xn) are n-dimensional decision variables; fi (.)
represents p linear conflicting objective functions and gj(.) are q
different constraints inequalities. A feasible solution Xn to problem
(3) is said to be efficient (for a maximizing problem) if there does
not exist any other feasible solution X such that for all i¼1, …, p,
fi(X)Zfi(Xn), and fi(X)4fi(Xn) for at least one i. In other words Xn is
not dominated by any other solution in terms of achievement in
the objective function. As will be shown in a later section in our
application we have three objective functions in QFD optimization
10 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21

formulated as follows:
P 9
Max f 1 ðX Þ ¼ j A n REj xj >
>
P >
>
Max f 2 ðX Þ ¼ k A n;k a j REk xk >
>
>
>
>
>
… >
>
P =
Max f p ðX Þ ¼ l A n;l a k a j RE l xl ð4Þ
>
>
X
n Xn X >
>
Subject to : c j xj 
n
r >
>
s ij x i x j B >
>
j4i >
>
j¼1 i¼1
> >
;
xAX

where n is the number of resilience startegies; REj is the resilience

Fig. 1. QFD Framework. Note: CRi ¼customer requirements; Wi ¼ degree of impor-
tance of CRi’s; DRj ¼ design requirements; Rij ¼relationship matrix (i.e. degree to
which CRi is met by DRj). A.I. ¼absolute importance of DRj’s; R.I. ¼relative impor-
tance of DRj’s.
problem as: maximize procurement strategies, maximize proces-
sing strategies, and maximize distribution strategies, to mitigate
the vulnerabilities.
As presented earlier we define ‘supply chain resilience effi-
ciency (SCREF)’ as follows:
(i) resilience capability must be resource efficient (e.g. minimum
cost of implementation), and
(ii) portfolio of chosen resilience capabilities must be efficient (or
non-dominated) from multiple objectives perspective [63].
Vugrin et al. [99] touched on the efficiency of system resilience.
According to the authors system resilience must use ‘lowest
possible amount of resources’ to be efficient. Inline with  Vugrin
et al. [99] we define ‘Resilience Efficiency’ REj as AIj =C j , where C j
is the cost of implementing jth resilience strategy. It is noted that
parameters in the objective functions in our QFD optimization
problem are the REj. This satisfies our condition (i) of SCREF.
To satisfy condition (ii) of SCREF we ascertain that the solution
of the QFD multiple objective optimization problem is efficient
(non-dominated). We develop a multiple objective 0–1 optimiza-
tion problem that will find the efficient portfolio of resilience
strategies to mitigate the vulnerabilities. Pettit et al. [79] recently
have mentioned portfolio approach to resilience capabilities.
However, the authors did not offer any methodology of achieving
that. Our 0–1 QFD multiple objective optimization problem is
efficiency (described earlier) of resilience strategy j, xj is one or
zero depending on if the corresponding resilience strategy j is
selected or not; cj is the expected cost of implementing resilience
strategy j; sij is the savings if resilience strategies i and j are
implemented together; B is the available budget.
It is noted that our constraints are non-linear and similar to
Park and Kim [77]. But depending on the application the con-
straint set could be formulated in an extensive way including the
budget, technological feasibility and competition constraints [110].
We argue that there are p different conflciting objectives among
the REj’s which need to be optimized simultaneously, hence
satisfactory effcient solution of problem (4) need to be found out
by interacting with the decision maker. It is noted that any
solution to problem (4) will offer a portfolio of resilience strategies
to mitigate the vulnerabilities. To find the efficient portfolios of
strategies we need to reformulate problem (4) as follows:
9
X p
>
λi f i ðXÞ >
>
Max >
>
>
>
i¼1 =
Xn X n X n ð5Þ
Subject to : c j xj  sij xi xj r B >
>
>
>
>
j¼1 i ¼ 1j 4 i >
>
;
x AX
where λi (i¼1,…,p) are positive numbers representing the weights
(importance) attached to the objective function f i ðXÞ by the decision
maker. Theorems from multiobjective optimization domain suggest
that any solution of problem (5) above is an efficient (non-dominated)
solution to problem (4) [63]. It is noted that the important weights λi
are only needed to find the first efficient solution to problem (4). It is
noted that Karsak [56,57] has also determined efficient solution for
multiple objective optimization in QFD. However our multi-objective
model formulation and solution approach are different from Karsak
[56,57] in the following ways: (i) Karasak deals with fuzzy multiple
objectives, while we deal with non-fuzzy multiple objective formula-
tion, (ii) Karasak finds fuzzy priorities of the objectives and use those
priorities to find a single efficient solution, while we provide an
interactive method which finds an initial efficient solution then
explores other efficient solutions by changing the weights as par the
likings of the decision makers. It is noted that in the domain of
multiple objective decision making interactive approaches are pre-
ferred than the non-interactive approaches [91, p. 98].
We now offer an interactive procedure to find satisfactory portfolio
of efficient resilience strategies to mitigate the vulnerabilities.
Step 1: Optimize each objective function of problem (4). There
will be p such solutions. Offer them to the decision maker.
These will act as maximum goal of each individual objective.
Any efficient solution will be a compromise solution from
these goals.
Step 2: Formulate problem (5) where each λi ¼ 1 (i¼ 1,…,p).
Solve problem (5). The solution will be efficient (non-domi-
nated) for problem (4). Offer it to the decision maker.
Step 3: If the decision maker is satisfied with this solution (after
comparing it with the solutions found in step 1), Stop. This will
 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21 11

be the satisfactory portfolio of resilience strategies to mitigate
the vulnerabilities.
If the decision maker is not satisfied, go to step 4.
Step 4: Interact with the decision maker to find new values of
λi’s which represent his/her preferences for the objective
functions.
Step 5: Formulate and solve problem (5) with the new values of
λi’s. Offer it to the decision maker. Go to Step 3.
Application in RMG industry of Bangladesh
Methodology
There are fundamentally two research paradigms: positivist and
interpretivist [76]. Positivist paradigm is associated with the quanti-
tative research method based on specific research questions and
hypotheses testing [51,27]. Whereas the interpretivist paradigm relies
on the qualitative method and there is subjective interpretation of the
researcher involved [27]. However, in recent times research based on
mixed methods, a combination of qualitative and quantitative meth-
ods, has gained popularity [12], because it assists in increasing the
quality, accuracy, validity and reliability of data [27,2].
It is noted that the primary objective of this study is to ‘develop
efficient resilience capabilities of RMG supply chain of Bangladesh to
mitigate organizations vulnerabilities’. To effectively conduct this study
we have adopted mixed methods of qualitative and quantitative
approaches under positivist paradigm. This uniquely fits with the
Quality Function Deployment (QFD) as a research methodology which
has aspects of both qualitative and quantitative methods [77,102] and
which we have embraced in our current study.
Our applications in RMG industry of Bangladesh are conducted
in three studies and in three stages as follows.
Study 1
Our study 1 is conducted on one of the largest apparel manufac-
turer in Bangladesh. Since its establishment in 1984, the company has
been attaining specific experience in designing and manufacturing
different types of apparels. It specializes in high quality apparel
production, and is one of the leading apparel exporters in Bangladesh.
It has 30,000 employees in 28 apparel production units in different
parts of Bangladesh as well in Cambodia and Vietnam. It produces
Bottom, Shirt, Sportswear, Polo knit and Sweater compliant with
various quality requirements. It exports its products to North America,
South America and Western Europe. The total export volume of the
company is 120,000 dozen/month which accounts for an aggregate
turnover of US$100 million. Sears, C & A, PVH, GAP, Wal-mart, JCPanny
and H&M are the major buyers of its products. Paired with the apparel
production, it has developed the backward linkage facilities as it has
its own textile (waiving, cotton yarn spinning) with dying facilities of
cotton & synthetic, poly, label, button, zipper, thread and carton
factories. The apparel export of the company is increasing every year
however, it is still bogged with multiple disruptive events like labour
unrest, political instability, interruption in utility supply especially
power shortage; among others. These disruptive events expose the
company to various types of vulnerabilities. Although the conglomer-
ate has 30 years of experience in combatting various vulnerabilities, a
formal resilient approach to mitigate the vulnerabilities is needed.
Consequently, the management of this group of companies has agreed
to take part in QFD approach proposed in this paper to explicate the
vulnerabilities and resilience strategies to overcome the vulne-
rabilities.
Stage 1
In stage 1 the supply chain vulnerabilities (WHATs) and corre-
sponding resilience strategies (HOWs) to mitigate these vulnerabilities
are found. To accomplish this, data have been collected from three
RMG manufacturing companies and two accessory production com-
panies of the parent Group. Although under the same conglomerate,
these five companies have unique features, and face and try to resolve
problems independently. It also allows us to collect data from multiple
sources (five decision makers) and thus, enhances the reliability of our
data [3]. The data have been collected via semi-structured detailed
interviews from five respondents. Table 4 provides the profiles of the
five companies and the respondents. Each interview lasted between
60 and 80 min.
Table 5 presents the explicated supply chain vulnerabilities and
resilience strategies. There are 26 vulnerabilities. Out of these, 17
vulnerabilities have been supported by majority of the respondents.
These 17 vulnerabilities are considered for further analysis. Corre-
sponding to the 17 vulnerabilities the respondents identified 13
resilience strategies to mitigate the vulnerabilities. It is observed that
most of these vulnerabilities and resilience strategies are consistent
with the literature (as per Tables 2 and 3). This adds further validity to
our collected data [3].
Stage 2
In stage 2 we collect the quantitative data (wi, Rij) to find the AIj
and RIj (see Table 2 and Eqs. (1) and (2)). We also collect the data on
costs (cj) of implementing the resilience strategies in order to find the
resilience efficiency REj (see earlier discussion) and the savings sij’s
when two resilience strategies i and j are implemented together. As

Table 4
Profile of five companies and Participants.

Participant Position Company type Product(s) Company size (number of Age of Production and
employees) company sales

D1 Manager RMG Bottom item (shorts, trousers) More than 5000 10 years 60–80,000 dozens/
Merchandising manufacturing year
unit
D2 Deputy general Accessory Cartoon and poly bags which are used for packing Less than 1000 5 years 14–1,500,000
manager supplying unit products during shipment. cartoons each year
D3 Manager RMG Main product is shirt. 2000–3000 10 years 25–30,000 dozens/
Merchandising manufacturing year
unit
D4 Supply chain RMG Main product is shirt. More than 5000 25 years 40–50,000 dozen/
manager manufacturing year
unit
D5 Deputy General Accessory Process raw cotton to yarn and then spin the yarn for Less than 1000 15 years 40–5,000,000 yard/
manager supplying unit knitting. year
12 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21

Table 5
Vulnerabilities and corresponding resilience capabilities.

Vulnerability factors Specific vulnerabilities Enterprises AHP Weights Adjusted weights

(WHATs)
1 2 3 4 5

Hazard (.132) Natural disaster (flood, cyclone) (HV1) y y y .169 .022

Political instability and labour unrest (HV2) y y y y y .534 .071
Fire and other accidental damage (HV3) y y y .297 .039
Sabotage y y
Piracy and theft y

Strategic (.223) Increased competition (SV1) y y y y y .445 .099

Failure to comply with social-environmental factors (SV2) y y y y .383 .085
Problem of relation with buyer & suppliers (SV3) y y .172 .04
Problem of integration y y

Financial (.154) Currency & raw material price fluctuation (FV1) y y y y .253 .039
Economic recession (FV2) y y y y y .602 .093
Bankruptcy of supply chain members (FV3) y y .145 .022
Higher rate of bank intest y

Operational (.195) Worker (skill, absenteeism, illeteracy) (OV1) y y y y .282 .055

Disruption in utility supply (OV2) y y y y y .543 .106
Product quality defection (poor quality) (OV3) y y y .175 .034
Machinery breakdown and failure y y
IT system failure
Production planning problem y y

Infrastructural (.115) Delay in port and customs (IV1) y y y y .578 .067

Delay in transportation for poor infrastructure and port facilities (IV2) y y .422 .049
Strike by port workers y

Demand & supply (.181) Suppliers’ disruptions (DSV1) y y y .242 .043

Lack of material (dependence on imported material and lack of backward linkage) y y y y y .597 .108
(DSV2)
Buyers’ disruptions (expect discount) (DSV3) y y y .161 .029
Unpredictability of demand y

Resience strategies
(HOWs) Product differentiation and customization (St1) y y y
Multiple sources of supply (St2) y y y y y
Channel rerouting, reconfiguration (St3) y y
Back up capacity (St4) y y y y
Quality control and reducing defection (St5) y y y y
Skill and efficiency development through traing and counselling (St6) y y y y y
Product and process improvement for efficiency and waste reduction (St7) y y y y
Forecasting and predictive analysis (St8) y y y y
Forward and backward linkage (St9) y y y
Responsiveness to customer (St10) y y y y y
Compliance of social and environmental issues (St11) y y y y y
ICT adoption and information intergartion (St12) y y y y
Cooperation, communication and building relation with buyers and suppliers y y y y y
(St13)

Note: Factors with low responses have not been considered for importance rating.

these data are extremely demanding to collect, we select one RMG
manufacturing company from the parent group of companies (D1 in
Table 4). Table 4 shows that it is a large company both in terms of
number of employees and sales volume. It also produces more than
one product type and hence susceptible to more vulnerabilities.
Barratt et al. [3] mention that single case company allows to collect
much deeper data, which is the situation in our study. However, we
collect data from multiple sources (three decision makers) of the case
company which enhances reliability of the collected data [3].
To find the wi we use Saaty’s [86] AHP method in a hierarchical
setting. The wi’s are averaged for three respondents and are shown
in the last column of Table 5. To find Rij’s we ask the respondents to
indicate (in their opinions) the ‘extent of mitigating the vulner-
ability i by resiliency strategy j’ using the widely used scale of 9, 3,
1 and 0 [17,82,34]. The Rij’s are also averaged for three respondents.
Fig. 2 shows the wiRij values ( in the main body of the matrix) and
the AIj’s and the RIj’s for different resilence strategies. It is observed
that resilience strategies 13 (building relations with with buyers
and suppliers) and 4 (back-up capacity) have the highest AIs of 4.11
and 3.74 respectively.
The cost (cj) of implementing the resilience strategies is found
in a more elaborate way. Each respondent is asked to give their
most likely, optimistic and pessimistic estimates of cj. Then the
expected cost is found by the formula Ce ¼(4Cm þ Co þCp)/6, where
Ce, Cm, Co and Cp are the expected, most likely, optimistic and
pessimistic cost estimates. These costs are then averaged for three
respondents. Fig. 2 shows the costs cj’s and the resilience efficien-
cies REj’s. It is noted that resilience strategy 13 (building relations
with buyers and suppliers) has the highest RE of .19 followed by
resilience strategy 8 (forecasting and prediction) of .15. The cost
figures are in millions of Taka.1 To find the savings sij the
respondents were asked to indicate which resilience strategies
could be implemented simultaneously and what could be the
estimated savings. The roof of Fig. 2 shows these savings data. For
example, resilience strategies 1 and 10 can be implemented sim-
ultaneously and the estimated savings would be Taka 3.6 million.
1
Taka is Bangladeshi currency. At the time of this study the exchange rate was
1 US$ ¼ 77 Taka.
 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21 13

3.6 6.4
7
5.6
3.3 5.8
4
8 3.7
9.7 4.6
8.3 5.2
4.3
5.5 2.9 5.6
7.2
4 7.9 4.8

SCVs Weights St1 St2 St3 St4 St5 St6 St7 St8 St9 St10 St11 St12 St13 A.I
HV1 0.022 0 0.066 0.066 0.98 0 0 0 0.037 0.022 0 0 0.051 0 1.222
HV2 0.071 0 0.165 0.497 0 0 0.071 0.071 0 0 0.497 0.165 0 1.466
HV3 0.039 0 0.039 0.117 0.351 0 0 0 0 0 0 0.351 0.117 0 0.975
SV1 0.099 0.891 0.165 0 0.099 0.693 0.693 0.891 0.099 0.231 0.693 0.495 0.231 0.891 6.072
SV2 0.085 0 0 0 0.255 0 0.198 0 0 0.085 0.595 0 0 1.133
SV3 0.04 0.067 0.12 0 0.12 0.04 0 0.2 0 0.12 0.28 0.04 0.04 0.36 1.387
FV1 0.039 0 0.117 0 0.195 0 0 0 0.351 0.039 0 0 0 0.195 0.897
FV2 0.093 0.279 0.279 0 0 0.22 0.651 0.837 0.837 0.155 0.279 0.093 0.22 0.837 4.687
FV3 0.023 0 0.161 0 0.023 0 0 0 0.069 0.038 0.023 0 0.38 0.54 1.234
OV1 0.055 0 0 0 0 0 0.495 0.055 0 0 0 0.165 0 0 0.715
OV2 0.106 0 0 0 0.247 0.106 0.106 0.106 0.106 0 0.106 0 0.106 0.247 1.13
OV3 0.034 0.034 0.08 0 0 0.306 0.238 0.306 0 0.067 0.102 0.057 0.034 0.102 1.326
IV1 0.067 0 0.56 0.18 0.112 0 0 0 0.18 0.18 0.067 0 0.067 0.112 1.458
IV2 0.049 0.114 0 0.147 0 0 0 0 0 0 0 0 0 0.261
DSV1 0.043 0 0.387 0.72 0.129 0.043 0 0 0.1 0.215 0.129 0 0.129 0.387 2.239
DSV2 0.108 0.324 0.972 0.252 0.756 0 0.108 0.18 0.252 0.756 0 0 0.108 0.18 3.888
DSV3 0.029 0.029 0 0.029 0.087 0.203 0.261 0.203 0.029 0.087 0.068 0.145 0.087 0.261 1.489
A.I 1.624 3.225 1.364 3.743 1.866 2.552 3.047 2.131 1.91 1.832 2.438 1.735 4.112
Cost 51.6 36.4 25.2 27.3 22.7 18.8 37.2 14.3 71.8 22.4 56.7 15.6 21.6
RE 0.031473 0.088599 0.054127 0.137106 0.082203 0.135745 0.081909 0.149021 0.026602 0.081786 0.042998 0.111218 0.19037

Fig. 2. Supply Chain resilience model: Study 1. Note: A.I.¼absolute importance; Stj ¼resilience strategy j; HV, SV, FV, OV, IV, DSV¼ various vulnerabilities, RE ¼ resilience
efficiency.

Stage 3
In stage 3 we develop the 0–1 multiple objective problem (as in
(4)) and apply the stepwise procedure to find the satisfactory
portfolio of efficient resilience strategies. In order to find the
multiple objectives among the resilience strategies we interacted
with the three decision makers of the case company and came up
with three objectives to be maximized as follows:
f1(X)¼ ‘Procurement’ related strategies which includes strate-
gies ST2, ST4, ST9 and ST13.
f2(X)¼ ‘Processing’ related strategies which includes ST5, ST6,
ST7, ST11 and ST12.
f3(X)¼ ‘Distribution’ related strategies which includes ST3, ST8,
ST1, and ST10.
Problem (4) now becomes:
Max f1(X)¼ RE2x2 þRE4x4 þ RE9x9 þRE13x13
Max f2(X)¼ RE5x5 þRE6x6 þ RE7x7 þRE11x11 þ RE12x12
Max f3(X)¼ RE3x3 þRE8x8 þ RE1x1 þRE10x10
Subject to:
c1 x1 þ c2 x2 þ c3 x3 þ c4 x4 þ c5 x5 þ c6 x6 þ c7 x7
þ c8 x8 þ c9 x9 þ c10 x10 þ c11 x11 þ c12 x12
þ c13 x13  S1;6 x1 x6  S1;10 x1 x10  S2;4 x2 x4
 S3;8 x3 x8  S3;9 x3 x9  S3;10 x3 x10 S4;8 x4 x8
 S4;10 x4 x10  S4;12 x4 x12  S4;13 x4 x13 S5;6 x5 x6
 S5;7 x5 x7  S6;7 x6 x7  S7;10 x7 x10
 S7;11 x7 x11  S7;13 x7 x13  S8;13 x8 x13 S9;13 x9 x13  S10;11 x10 x11
 S10;12 x10 x12  S11;13 x11 x13 rB where xj ¼ 0 or 1
The values of REj and cj are obtained from Fig. 2. According to
the decision makers the budget can be set aside as 110 million
Taka. Hence the budget B is set at 110 million Taka.
We now follow the stepwise procedure to find the satisfactory
portfolio of efficient resilience strategies. As per step 1 of the stepwise
procedure, each objective is optimized separately. We use EXCEL
Solver as the optimization software. The optimal solutions and
corresponding portfolio of resilience strategies are shown in Table 6.
It is observed that optimal values of the resilience efficiencies (REs) of
the procurement (f1), processing (f2) and distribution (f3) strategies are
.416, .411 and .316 respectively. If f1 is optimized alone then ‘multiple
sources of supply’ (St2), ‘back up capacity’ (St4) and ‘cooperation and
communication with buyers and suppliers’ (St13) should be imple-
mented (see Table 5) for a total cost of 85.3 million Taka leaving 24.7
million Taka unspent. However, since (St2, St4) and (St4, St13) are
implemented together there is potential savings of 13.6 million Taka
(see roof of Fig. 2). Thus, there is still 38.3 million Taka available for
implementing other strategies. Since ‘forward and backward linkages’
(St9; another component of f1) costs 71.8 million Taka to implement
the management can look into other strategies to be implemented for
the remaining budget of 38.3 million Taka. Other optimized solution
for f2 and f3 can be analysed similarly.
In our case the three decision makers wanted to explore more
efficient solutions in order to obtain a compromise among the
three objectives. We then applied step 3 of the stepwise procedure
using equal weighting for the three objectives. This produces an
efficient solution as shown in Table 7 (first row). It is noted that
the objective function values (resilience efficiencies) are .327 (for
f1; a deviation of 21.65% from the optimal value of f1), .329 (for f2; a
deviation of 19.95% from the optimal value of f2) and .231 (for f3; a
deviation of 26.89% from the optimal value of f3). The decision
makers now wanted to weight the objectives according to their
preferences and explore further compromise and efficient solu-
tions. The steps 3 and 4 come into action now. After some
14 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21

Table 6
Aspiration levels of the objectives for study 1.

Objective Function Aspiration level Procurement strategy (f1) Processing strategy (f2) Distribution strategy (f3) Budget

X2 X4 X9 X13 X5 X6 X7 X11 X12 X1 X3 X8 X10

f1 .416 1 1 0 1 0 0 0 0 0 0 0 0 0 110
f2 .411 0 0 0 0 1 1 1 0 1 0 0 0 0
f3 .316 0 0 0 0 0 0 0 0 0 1 1 1 1

Note: X2, X4, X9, X13 ¼ procurement strategy (f1); X5, X6, X7, X11, X12 ¼processing strategy (f2); X1, X3, X8 and X10 ¼ distribution strategy (f3).

Table 7
Efficient resilient portfolios of study 1.

Objective function Procurement strategies Processing strategies Distribution strategies Budget

f1 f2 f3 X2 X4 X9 X13 X5 X6 X7 X11 X12 X1 X3 X8 X10

Weighted (1,1,1) .327 .329 .231 0 1 0 1 1 1 0 0 1 0 0 1 1 110

Weighted (.2, .5, .3) .327 .411 .149 0 1 0 1 1 1 1 0 1 0 0 1 0

Note: .2 weight for f1; .5 weight for f2; .3 weight for f3.

deliberations the weights of (.2, .5, and .3) were settled with for
objectives f1, f2 and f3 respectively. It shows that the decision
makers prefer ‘processing’ objective (f2) more compared to other
two objectives. Applying step 5 we now obtain a second efficient
solution as shown in Table 7 (second row). The objective function
values of this solution are .327 (for f1; a deviation of 21.65% from
the optimal value of f1), .411 (for f2; a deviation of 0% from the
optimal value of f2), and .149 (for f3; a deviation of 52.8% from the
optimal value of f3). It is interesting to note that this solution
produces optimal value for objective f2 (processing), as more
weight (.5) was given to this objective. However, the deviation
from the optimal solution for f3 was on the high side. The decision
maker therefore chose the earlier (equal weighting) solution with
objective function values of .327, .329, and .231. It is observed from
Table 7 that this solution selects strategies St4 (back-up capacity),
St13 (building relation with buyers and suppliers), St5 (quality
control), St6 (skill and efficiency development), St12 (ICT adop-
tion), St8 (forecasting), and St10 (responsiveness to customers) to
be implemented which require a total budget of 142.7 million
Taka. However there is a savings of 41.1 million Taka as (St4, St8),
(St4, St10), (St4, St12), (St4, St13), (St5, St6), (St8, St12), and (St10,
St12) are implemented together. Thus net required budget is 101.6
million Taka which is well within the budget constraint of 110
million Taka.
Study 2
The study 2 is conducted on one of the leading manufacturer of
jeans in Bangladesh to ensure the transferability of our findings and
applicability of our method in another setting. From its modest
beginning in 1984 the company has come a long way. It has modern
research and development facility. Within its six production units the
company now employs over 22,000 employees, produces over 30
million jeans per year and exports to more than 25 countries. We
conducted the case study on one of the units of the company which
produces casual wears for both men and women. Two executives of
the unit took part in our study.
Stages 1 and 2
Tashakkori and Teddlie [96] mention that external validity of
quantitative research and transferability of qualitative research are
similar in nature, which refer to the degree to which results of one
context can be applicable to other situation. Since our study 2 is also
on apparel industry, we offered the vulnerabilities of study 1 to the
two executives of study 2 and asked them to delve into them and
come up with vulnerabilities specific to their situation. In the end the
executives came up with 25 vulnerabilities of which 20 are similar to
studies 1 and 5 are new. Table 8 shows these vulnerabilities. The
bolded ones are the new vulnerabilities. It is noted that for the next
stage of the analysis (i.e. developing the supply chain resilience
model) the highest three weighted vulnerabilities from each group
were considered. For example, for hazard vulnerability group HV1,
HV2, and HV3 were selected which had the highest weights among
the five hazard vulnerabilities (see Fig. 3).
Next, resilience strategies to mitigate these vulnerabilities were
sought. Again the strategies of study 1 were given to the executives.
They explored these and at the end came up with 14 strategies of
which 13 are similar to study 1 and one is new (see Table 8). This
acceptability of the outcomes from study 1 to study 2 ensures the
transferability of our results.
In the next stage of the analysis, the quantitative data (wi, Rij)
were collected from the two executives. The data on cost cj and
savings sij were also estimated. Finally the supply chain resilience
model was developed as shown in Fig. 3. Unlike study 1 the
resilience strategy 9 (Demand forecasting) has the highest RE of
.108 followed by resilience strategy 5 (Maintaining reserve capa-
city). However it is noted that demand forecasting was ranked 2 in
study 1. This highlights the need for accurate demand forecasting
in the apparel industry of Bangladesh to mitigate the demand–
supply related vulnerabilities.
Stage 3
Like study 1 in stage 3 we develop and solve the 0–1 multiple
objective model and find the satisfactory portfolio of efficient
resilience strategies. After interacting with the two executives the
following three objectives were formulated:
f1(X)¼‘Procurement’ related strategies comprising ST3, ST5,
ST10 and ST14.
f2(X)¼‘Processing’ related strategies comprising ST1, ST6, ST7,
ST8, ST12, ST13.
f3(X)¼‘Distribution’ related strategies comprising ST4, ST9, ST2
and ST11.
For brevity the multi-objective model is not presented here which
is similar to study 1. The stepwise procedure is now followed to find
 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21 15

Table 8
Vulnerabilities and strategies for study 2.

Vulenarabilities Strategies

HV1¼ Sabotage St1 ¼ Security system improvement

HV2¼ Political instability St2¼Product customization
HV3¼ Factory fire St3¼Alternative sources of supply
HV4¼ Natural disaster (flood, cyclone) St4¼Alternative transportation routing
HV5 ¼ Foreign government policy change St5¼Maintaining reserve capacity
SV1¼Competitive pressure St6¼Quality control
SV2¼Reputation loss St7¼Skill development training
SV3¼Problem of integration with supply chain members St8¼Improving process technology
SV4¼Failure to comply with socio-environmental issues St9¼Demand forecasting
FV1¼ Increasing raw material price St10 ¼Backward linkage development
FV2¼ Impact of Economic recession St11 ¼Quick response to customers’ requirements
FV3 ¼Cost of Finance St12¼ Improving social and environmental performance
FV4¼ Exchnage rate fluctuation St13 ¼Information integration
OV1 ¼ Shortage of skilled worker St14 ¼Cooperation and collaboration with suppliers.
OV2 ¼ Impact of Power crisis
OV3¼Concentrated production location
OV4 ¼ Machine breakdown & failure
OV5 ¼ IT system failure
OV6 ¼ Production planning problem
IV1¼ Delay in port
IV2¼ Poor transportation infrastructure
IV3¼ Delay due to export import document processing
DSV1¼ Suppliers’ disruption
DSV2¼ High dependence on imported materials
DSV3¼ Buyers’ disruption

4.5
9
9.5
4.5
7
3.5

7.5 3.7
8
5 8.5

SCVs Weights St1 St2 St3 St4 St5 St6 St7 St8 St9 St10 St11 St12 St13 St14 AI
HV1 0.02 0.27 0 0 0.1 0 0 0 0 0 0 0 0 0 0.37
HV2 0.063 0 0 0.147 0 0.567 0 0 0 0.119 0 0 0 0.119 0.084 1.036
HV3 0.034 0.238 0 0.039 0 0.273 0 0 0 0 0 0 0.351 0.065 0.102 1.068
SV1 0.082 0 0.891 0.165 0 0 0.693 0.891 0.693 0.297 0.231 0.891 0.231 0.231 0.297 5.511
SV2 0.089 0 0 0 0 0 0.623 0 0 0 0 0.801 0 0.41 1.834
SV3 0.07 0 0.067 1.2 0.093 0.04 0 0.12 0 0.12 0.28 0 0.04 0.36 2.32
FV1 0.041 0 0 0.091 0 0.117 0 0 0 0.351 0.039 0 0 0 0.117 0.715
FV2 0.081 0 0.465 0.279 0 0 0.093 0.837 0.651 0.837 0.093 0.217 0.093 0.093 0.837 4.495
FV3 0.023 0 0 0.207 0 0.054 0 0 0 0.069 0.023 0.054 0 0.054 0.069 0.53
OV1 0.048 0 0 0 0 0 0 0.495 0.55 0 0 0 0.09 0 0 1.135
OV2 0.104 0 0 0 0 0.954 0 0 0 0.177 0 0 0 0 0.106 1.237
OV3 0.033 0 0 0 0 0.231 0 0 0 0 0 0 0 0 0 0.231
IV1 0.061 0 0 0.335 0.469 0.18 0 0 0 0.18 0.156 0 0 0 0.18 1.5
IV2 0.045 0 0.08 0 0.245 0 0 0 0 0 0 0 0 0 0.325
IV3 0.036 0 0 0 0 0.18 0 0 0 0.08 0.108 0 0 0 0 0.368
DSV1 0.046 0 0 0.387 0 0.215 0 0 0 0.215 0.245 0.129 0 0.072 0.129 1.392
DSV2 0.106 0 0 0.972 0 0.54 0 0.252 0 0.324 0.972 0 0 0 0.54 3.6
DSV3 0.025 0 0.075 0 0.175 0.058 0.175 0.058 0.125 0.087 0.068 0.225 0.125 0.058 0 1.229
AI 1.007 0.508 1.498 2.702 1.844 3.807 1.624 2.533 2.139 2.736 2.055 1.796 1.691 0.732 3.231
Cost 11.5 82.2 45.7 37.4 41.6 43.6 36.3 75.7 25.2 120.4 46.8 102.3 33.6 45.4
RE 0.044 0.018224 0.059125 0.049305 0.091514 0.037248 0.06978 0.028256 0.108571 0.017068 0.038376 0.01653 0.021786 0.071167

Fig. 3. Supply Chain resilience model: Study 2.

the satisfactory efficient portfolio for study 2. First, each objective
function is optimized separately. The optimal solutions and corre-
sponding portfolio of resilience strategies are shown in Table 9. It is
observed that optimal values of the resilience efficiencies (REs) of the
procurement (f1), processing (f2) and distribution (f3) strategies are
.163, .173 and 0. 196 respectively. It is noted that the optimal values of
the objectives are quite different from that of study 1. When f1 is
optimized alone ‘maintaining reserve capacity’ (St5), and ‘cooperation
and collaboration with suppliers’ (St14) should be implemented. It is
noted that these two strategies were also selected in study 1.
Having seen the aspiration levels of three optimized objectives
the two executives wanted to explore further efficient portfolios of
resilience strategies. Thus step 3 of our stepwise procedure came
into action at this stage. We first found an efficient solution by
equal weighting of the objectives. This solution is shown in
Table 10 (first row). Next the executives wanted to weight the
16 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21

Table 9
Aspiration levels of the objectives for study 2.

Objective function Aspiration level Procurement strategy (f1) Processing strategy (f2) Distribution strategy (f3) Budget

X3 X5 X10 X14 X1 X6 X7 X8 X12 X13 X2 X4 X9 X11

f1 .163 0 1 0 1 0 0 0 0 0 0 0 0 0 130
f2 .173 0 0 0 0 1 1 1 0 0 1 0 0 0 0
f3 .196 0 0 0 0 0 0 0 0 0 0 1 1 1

Note: X3, X5, X10, X14 ¼ procurement strategy (f1); X1, X6, X7, X8, X12, X13 ¼ processing strategy (f2); X2, X4, X9, and X11 ¼ distribution strategy (f3).

Table 10
Efficient resilient portfolios of study 2.

Objective Function Procurement strategies Processing strategies Distribution strategies Budget

f1 f2 f3 X3 X5 X10 X14 X1 X6 X7 X8 X12 X13 X2 X4 X9 X11

Weighted (1,1,1) .092 .114 .158 0 1 0 0 1 0 1 0 0 0 0 1 1 0 130

Weighted (.25, .45, .3) .092 .151 .109 0 1 0 0 1 1 1 0 0 0 0 0 1 0

Note: .25 weight for f1; .45 weight for f2; .3 weight for f3.

objectives now and a weighting scheme of (.25, .45, and .3) was
settled with for objectives f1, f2 and f3 respectively. With this, the
second efficient solution is obtained which is shown in second row
of Table 10. Compared to the equal weighting solution this solution
still offers the optimal value for objective f1, with some improve-
ment in objective f2 at the sacrifice of objective f3. The two
executives preferred this solution compared to the equal weight-
ing solution. It is observed from Table 10 that this compromise
solution selected five out of fourteen resilience strategies as
follows: St1 (security system improvement), St5 (maintaining
reserve capacity), St6 (quality control), St7 (skill development
training) and St9 (demand forecasting). The study 2 also selects
one of the resilience strategies ‘security system improvement’
(St1) which was not even one of the resilience strategies of study 1.
It is noted that while in study 1 the preferred optimal efficient
solution was spread out among the three objectives (procurement,
processing, and distribution), in study 2 the preferred solution
concentrated more on the processing strategies. This highlights
the difference in management attitude of the two giant garment
companies in Bangladesh.
Study 3
Study 2 was primarily conducted to ensure the transferability
[96] of results and findings from one context to another context (i.
e. transferring vulnerabilities and resilience strategies from study
1 to study 2). To ensure further external validity [14] of our
method we have conducted study 3 on another garment manu-
facturer from Bangladesh. It is noted that data from this company
were independently collected without any reference to studies
1 and 2. Two executives from the company took part in the study.
Study 3 company is a family owned business, situated in
Chittagong, Bangladesh and was established in 1983. It has four
manufacturing factories with work force of more than 3000. The
products of the company are casual and dress pants, shirts and
men’s shorts. The company is certified by the buyers for complying
with the social and environmental sustainability factors. It markets
its products mainly in the USA, the UK and Australia. It has its own
washing plant, screen printing unit, embroidery unit and in-house
clearing unit. This group also has their own transportation and
logistical services. The company is growing steadily but the
external uncertainties, specifically the political and economic
factors, pose threat to the smooth operation and growth of its
business. A resilience approach is therefore essential for the
survival and long term growth of the company.
Stages 1 and 2
The vulnerabilities and corresponding resilience strategies
were collected first from the two executives of the company.
Table 11 shows these vulnerabilities and resilience strategies along
with their commonality with studies 1 and 2. There are three
hazard, five strategic, three financial, five operational, two infra-
structural, and five demand–supply vulnerabilities. Among the 23
vulnerabilities of study 3, nine vulnerabilities are common to both
studies 1 and 2. These are: HV1—Political unrest, HV2—Fire and
other accident, HV3—Natural disaster, SV2—Increased competition,
SV4—Problem of relation with buyers & suppliers, OV1—Disruption
of utility, IV1—Inefficient port facility, IV2—Inefficient customs
process, and DSV1—Dependence of imported material. It can be
thus reasonably assumed that these are the most common
vulnerabilities of Bangladesh garment industry. It is noted that
for the next stage of the analysis (i.e. developing the supply chain
model) the highest three weighted vulnerabilities from each group
were considered.
Table 11 also shows the 14 resilience strategies. It is noted that
six of them are common to both studies 1 and 2. These are: St1—
Back up capacity, St3—Focusing on sustainability practise, St5—
Customer relationship development, St10—Multiple suppliers,
St12—Developing relationship with suppliers, and St13—Product
differentiation.
In the next stage of the analysis the supply chain resilience
model was developed as shown in Fig. 4. It is noted that the
weight (wi) of the vulnerabilities, relationship value (Rij), cost (cj)
and savings (sij) were also collected from the two executives of the
company which are shown in Fig. 4. It is observed that resilience
strategy 5 (customer relationship development) has the highest RE
value of .332 followed by resilience strategy 6 (developing new
buyers and markets) and resilience strategy 8 (strict quality
control at different stage). Interestingly these rankings of strate-
gies are quite different from studies 1 and 2.
Stage 3
Like studies 1 and 2 we then develop the 0–1 multiple
objective model and find the satisfactory efficient resilient
 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21 17

Table 11
Vulnerabilities and strategies for Study 3.

Vulnerabilities Study 1 Study 2 Strategies Study 1 Study 2

HV1—Political unrest ✓ ✓ St1—Back up capacity ✓ ✓

HV2—Fire and other accident ✓ ✓ St2—Risk management team
HV3—Natural disaster ✓ ✓ St3—Focusing on sustainability practise ✓ ✓
SV1—Reputation risk ✓ St4—Using updated technology ✓
SV2—Increased competition ✓ ✓ St5—Customer relationship development ✓ ✓
SV3—Selecting wrong production site St6—Developing new buyers & markets
SV4—Problem of relation with buyers & suppliers ✓ ✓ St7—Flexibility in production
SV5—Lack of sustainability planning & standard St8—Strict quality control at different stage ✓
FV1—Increased production cost St9—Training & development ✓
FV2—Lack of order St10—Multiple suppliers ✓ ✓
FV3—Loss due to rejection of shipment St11—Supplier selection & evaluation
OV1—Disruption of utility ✓ ✓ St12—Developing relationship with suppliers ✓ ✓
OV2—Switching of workers St13—Product differentiation ✓ ✓
OV3—lead time failure St14—Alternative transportation (e.g. air shipment) ✓
OV4—Lack of efficiency of workers ✓
OV5—Fault in Quality control ✓
IV1—Inefficient port facility ✓ ✓
IV2—Inefficient customs process ✓ ✓
DSV1—Dependence of imported material ✓ ✓
DSV2—Switching of buyers
DSV3—Lack of commitment of suppliers (late delivery, quality )
DSV4—Supply shortage
DSV5—Buyers' opportunism

strategies. As per the two executives of the company there are
three objectives to be optimized as follows:
f1(X)¼ ‘Procurement’ related strategies comprising ST1, ST10,
ST11 and ST12.
f2(X)¼ ‘Processing’ related strategies comprising ST2, ST3, ST4,
ST7, ST8, ST9.
f3(X)¼ ‘Distribution’ related strategies comprising ST5, ST6,
ST13 and ST14.
The stepwise procedure is now followed to find the satisfactory
efficient portfolio of strategies for study 3. First, each objective is
optimized separately. Table 12 shows the aspiration level of
objectives for the budgetary restriction of 80 million Taka. Steps
2 and 3 now comes into play to interact with the decision makers
for satisfactory efficient solution. The executives settled with a
weighting scheme of (.2, .4, .35) for objectives f1, f2 and f3
respectively. The corresponding satisfactory solution is shown in
Table 13 (2nd row). It is noted that the satisfactory solution is a
compromise solution among the three objectives. Table 13 shows
that the selected portfolio of strategies is: St10 (Multiple suppli-
ers), St11 (Supplier selection and evaluation), St12 (Developing
relationship with suppliers), St2 (Risk management team), St7
(Flexibility in production), St8 (Strict quality control at different
stage), St9 (Training and development), St5 (Customer relationship
development), and St6 (Developing new buyers and markets).
It is noted from the above findings of study 3 that our method
possesses good external validity. In spite of the fact that data on
study 3 has been collected independently following the structured
procedure of our method, the results have similarities with studies
1 and 2. This was expected as all three companies belong to the
readymade garment industry in Bangladesh and they operate in
similar competitive environment. The differences in results can be
attributed to the differences in management attitudes of three
different companies and their tangible and intangible resources.
Discussions and implications
This research aimed to develop resilience capabilities of Ban-
gladesh readymade garment (RMG) supply chain in order to
mitigate various vulnerabilities that RMG supply chain face. In
doing so, we adopted mixed-method research design with quali-
tative and quantitative approaches [27]. We defined supply chain
resilience efficiency (SCREF) and develop an interactive methodol-
ogy to determine efficient resilience capabilities based on QFD
approach. The proposed methodology is then applied on three
large RMG companies in Bangladesh. The three studies were
conducted to ensure external validity (and transferability) [14,96]
of our proposed methodology. Our experience indicates that the
proposed methodology can be successfully applied to determ-
ine efficient portfolio of resilience strategies to mitigate vulne-
rabilities.
While three companies face somewhat similar vulnerabilities
and their resilience strategies are also similar, there are some
differences in quantitative results that deserve attention. Table 11
shows the similarities and differences of vulnerabilities and
resilience strategies among the three companies. It is noted that
company 3 has come up with 11 new vulnerabilities which are
different from both companies 1 and 2. In terms of importance of
these vulnerabilities study 1 has identified four top vulnerabilities
as DSV2 (lack of materials), OV2 (disruptions in utility supply), SV1
(increased competition) and FV2 (impact of economic recession)
(see Fig. 3 and Table 6). While top four important vulnerabilities of
study 2 are DSV2 (high dependence on imported materials), OV2
(impact of power crisis), SV2 (reputation loss), SV1 (competitive
pressure), and FV2 (impact of economic recession) with equal
weighting with SV1 (see Fig. 3 and Table 8). It is noted that
‘reputation loss’ is a significant vulnerability of study 2, while this
was not even considered as a possible vulnerability by study 1.
Both companies are export oriented. However study 2 company
prides in its R & D and takes quality issue very seriously, which
might explain why this company takes ‘reputation loss’ very
seriously. The top four vulnerabilities of study 3 are FV2 (lack of
order), OV1 (disruption of utility), FV1 (increased production cost),
and HV1 (political unrest) (see Fig. 4 and Table 11). It is noted that
‘lack of order’ and ‘increased production cost’ are significant
vulnerabilities of study 3, while these two were not even con-
sidered as possible vulnerabilities of studies 1 and 2 (see Table 11).
Being family owned, company 3 works in an extremely competi-
tive environment. Hence these two vulnerabilities are very sig-
nificant for the company.
18 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21

1.5

2 2

5 2 2

SCVs Weights St1 St2 St3 St4 St5 St6 St7 St8 St9 St10 St11 St12 St13 St14
HV1 0.113 0.339 1.017 0 0 0 0 0.339 0 0 0 0 0 0 0.339
HV2 0.028 0.252 0.252 0.252 0 0 0 0 0 0.084 0 0 0 0 0
HV3 0.011 0.099 0.099 0 0 0 0 0 0 0 0 0 0 0 0
SV1 0.069 0 0 0.207 0.621 0.207 0.621 0.207 0.207 0.207 0 0 0 0.621 0
SV2 0.021 0 0 0 0 0 0 0 0 0.063 0 0 0 0 0
SV3 0.038 0 0.114 0.342 0 0 0 0 0.342 0 0 0.342 0.114 0 0
FV1 0.113 0 0 0.339 1.107 0 0 0.339 0.339 1.107 0.339 0 0.339 0.339 0
FV2 0.225 0 0 0.675 0 2.025 2.025 0.675 0.675 0 0 0 0 0.675 0
FV3 0.056 0 0 0 0 0.168 0 0 0.504 0 0 0 0 0 0
OV1 0.13 1.17 0 0 0 0 0 0.39 0 0 0 0 0 0 0
OV2 0.041 0 0 0.123 0 0 0 0 0 0.123 0 0 0 0 0
OV3 0.017 0 0 0 0.051 0 0 0.153 0 0.051 0 0 0.051 0 0.153
IV1 0.024 0 0 0 0 0 0 0 0 0 0 0 0 0 0.072
IV2 0.012 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DSV1 0.058 0.174 0 0 0 0 0 0 0 0 0.522 0 0 0 0.174
DSV2 0.029 0 0 0.087 0.087 0.261 0 0 0 0 0 0 0 0.087 0
DSV3 0.015 0.045 0 0 0 0 0 0 0 0 0.135 0.135 0.135 0 0
A.I 2.079 1.482 2.025 1.866 2.661 2.646 2.103 2.067 1.635 0.996 0.477 0.639 1.722 0.738
Cost 30 8 40 35 8 10 15 10 10 8 5 8 25 22
RE 0.069 0.185 0.051 0.053 0.332 0.265 0.14 0.207 0.164 0.125 0.095 0.08 0.067 0.034

Fig. 4. Supply Chain resilience model: Study 3.

Table 12
Aspiration levels of the objectives for study 3.

Objective function Aspiration level Procurement strategy (f1) Processing strategy (f2) Distribution strategy (f3) Budget

X1 X10 X11 X12 X2 X3 X4 X7 X8 X9 X5 X6 X13 X14

f1 .369 1 1 1 1 0 0 0 0 0 0 0 0 0 0 80
f2 .749 0 0 0 0 1 0 1 1 1 1 0 0 0 0
f3 .698 0 0 0 0 0 0 0 0 0 0 1 1 1 1

Note: x1,x10,x11,x12 ¼ procurement strategy (f1); x2,x3,x4,x7,x8, x9 ¼processing strategy (f2); x5,x6,x13 and x14 ¼distribution strategy (f3).

Table 13
Efficient Resilient Portfolios of study 3.

Objective Function Procurement strategies Processing strategies Distribution strategies Budget

f1 f2 f3 X1 X10 X11 X12 X2 X3 X4 X7 X8 X9 X5 X6 X13 X14

Weighted (1,1,1) .30 .696 .597 0 1 1 1 1 0 0 1 1 1 1 1 0 0 80

Weighted (.25, .40,.35) .30 .696 .597 0 1 1 1 1 0 0 1 1 1 1 1 0 0

Note: .25 weight for f1; .40 weight for f2; .35 weight for f3.

As pointed out earlier the three companies have different
management attitudes while mitigating vulnerabilities via resili-
ence strategies. This is reflected while finding the satisfactory
efficient resilience strategies via our interactive method. In study
1 the decision makers settled with efficient resilience strategies
which assigned equal weights to the ‘procurement’ (f1), ‘proces-
sing’ (f2) and ‘distribution’ (f3) strategies. Their preferred portfolio
of resilience strategies included seven out of possible 13 strategies
with two from procurement objective, three from processing
objective and two from distribution objective. This is a balanced
management approach practised by this company in order to
mitigate the vulnerabilities. On the other hand, in study 2 the
decision makers settled with efficient resilience strategies that
assigned more weight to the ‘processing’ strategy. Their preferred
solution included five out of possible 14 strategies with one from
‘procurement’ objective, three from ‘processing’ objective and one
from ‘distribution’ objective. The management of this company is
more concerned with satisfying the ‘processing’ objective in order
to mitigate the vulnerabilities. It is observed that in terms of
choosing the portfolio of resilience strategies study 3 is similar to
study 1, where both companies chose a compromise solution
spread out among the three objectives. Study 3 thus chose nine
 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21
strategies (among 14 strategies) with three from procurement
objective, four from processing objective and two from distribu-
tion objective.
Theoretical implications
The most important theoretical contribution of our research is that
we propose an interactive methodology to obtain satisfactory portfolio
of resilience efficient strategies. This extends traditional QFD based
optimization method to determine the most desirable HOWs [31,57],
which are non-interactive and do not guarantee to find efficient
solutions. Our method interacts with the decision makers in a
systematic way and generates efficient solutions (portfolios) which
satisfy the decision makers. We therefore take the ‘satisficing’
approach to decision making in our proposed methodology [93,89].
It is worthwhile to compare our methodology with existing
methodologies which deal with vulnerabilities in supply chains.
Wagner and Neshat [100] proposed a methodology to mitigate supply
chain vulnerabilities. The authors first found various vulnerability
drivers and then used graph theory to quantify vulnerability index.
The directed graph shows the relationship among the vulnerability
drivers, the knowledge of which is useful to mitigate vulnerabilities.
The authors mention that supply chain managers can use ‘risk
management methods and implement mitigation strategies’ to ease
vulnerability drivers. However, the authors do not offer any guide to
explicate the mitigation strategies nor do they show any relationship
between the vulnerabilities and mitigation strategies. In contrast in
our proposed method, although we do not develop any index for
vulnerability, we prioritize the vulnerabilities which are grouped as
hazard, strategic, financial, operational, infrastructure, and demand–
supply vulnerabilities [8]. We also develop specific resilience strategies
to mitigate the vulnerabilities and help the decision makers in an
interactive way to find the satisfactory efficient portfolio of resilience
strategies. Goh et al. [40] also developed a quantitative stochastic
model and methodology to deal with supply chains and vulnerabil-
ities. However their theoretical model development was for interna-
tional facility location and distribution problem for a company with
one product. Their primary objective was to find the optimal open–
shut decision of plants and the corresponding shipment quantities
from various plants to various markets in order to maximize profit
and minimize risks. It is noted that although along the same vein
Goh et al.’s problem definition and solution methodology is quite
different. It needs rich and elaborate data. We are not yet aware of its
application in any real world problem.
Managerial implications
Organizations need to develop resilience capabilities in order to
mitigate supply chain vulnerabilities [101]. In this vein our research
can help managers, primarily in the RMG sector in Bangladesh,
answer three fundamental questions: (i) what are the supply chain
vulnerabilities that RMG sector currently facing? (ii) What are the
resilience strategies to mitigate these vulnerabilities? (iii) What is the
efficient resilient portfolio of strategies to mitigate vulnerabilities
subject to budget constraint? Our study has identified a number of
vulnerabilities across three studies on three large RMG companies in
Bangladesh. With respect to question (i) it has been shown earlier that
there are lot of similarities among these vulnerabilities of three
companies. Other RMG companies in Bangladesh (and for that matter
elsewhere) can start with these vulnerabilities and contextualize them
for their specific situation. It has been shown in our study that the
vulnerabilities have good external validity. With respect to question
(ii) it has been observed that resilience strategies are also similar for
three companies. Once again, other RMG companies (in Bangladesh
and elsewhere) can start with these strategies and contextualize them
for their own use. For both questions (i) and (ii) starting with the
19
findings of our study will help the managers of RMG companies to be
one-step ahead instead of developing the vulnerabilities and resilience
strategies from clean slate.
For question (iii) we used our methodology to find the portfolio
of efficient resilient strategies subject to budget constraint. We
found that the three companies had different approaches to
managing the resilient strategies. Our findings can help the RMG
managers as an eye opener to find satisfactory efficient resilient
strategies in their own situation. It is highlighted here that our
interactive methodology is targeted for interaction with the
managers (decision makers) where they can see the results when
they input various weight for various objectives.
Conclusions
In this research we address the problem of mitigating vulnerabil-
ities of the ready-made garment industry of Bangladesh. Our basic
methodology is QFD [17,41], which we use to find the vulnerabilities
and their resilience strategies. We then develop an interactive multi-
objective methodology to find the satisfactory (as per the decision
makers) efficient portfolio of resilience strategies to mitigate the
vulnerabilities. We argue that this approach is novel and has a number
of advantages over existing approaches [31,57]. First, our method
ensures efficient portfolio of strategies and second, it interacts with
the decision makers and thus finds the satisfactory efficient solutions
in an interactive way.
We applied our method to three large garment companies in
Bangladesh. Using the stepwise procedure of QFD we found the
vulnerabilities, and corresponding resilience strategies in study 1. We
observed that these vulnerabilities and resilience strategies are
transferrable to study 2, although study 2 has few specific vulner-
abilities and resilience strategies of its own. Data for study 3 has been
collected independently of studies 1 and 2. Even then there are some
similarities among the vulnerabilities and resilient strategies of studies
1, 2 and 3. Combining the top prioritized vulnerabilities of the three
studies we conclude that most important vulnerabilities of the RMG
industry of Bangladesh are: lack of materials (high dependence on
imported materials), disruptions in utility supply, increased competi-
tion (and hence competitive pressure), impact of economic recession,
and reputation loss. Similarly, combining the preferred resilience
strategies of the two studies we conclude that the most preferred
strategies to mitigate the vulnerabilities are: back-up capacity, build-
ing relation with buyers and suppliers, quality control, skill and
efficiency development, ICT adoption, demand forecasting, respon-
siveness to customers, and security system improvement.
Our research is not free from limitations. First, for practical
application any QFD based method requires rich and detailed data.
Hence data collection is the major limitation. In our case one of the
researchers had good contacts with the RMG industry of Bangladesh.
It helped us to make good connections with three of the largest RMG
companies in Bangladesh. The management of these companies
understood the value of our analysis and were deeply motivated in
participating in the study. Second, in our data collection process we
used qualitative methods to collect various data. Hence reliability of
the data could be an issue. However, we dealt with real decision
makers and as such all the qualitative data (vulnerabilities, resilience
strategies) reflect their perception as they see fit for the companies.
We surmise that despite the limitations we have offered a detailed
methodology and its applications in RMG industry of Bangladesh in a
systematic manner. The methodology can be effectively applied in
other industrial settings and in other applications elsewhere. How-
ever, in any future applications the decision makers of the targeted
company need to be convinced about the value of the study. We have
found that once the value of the research is disseminated well the
decision makers are happy to be involved in the study.
20 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21
Future research can be directed along the following ways. First,
instead of multi-objective optimization multi-attribute methods can
be used based on the resilience efficiency values (RE) to find the
portfolio of resilience strategies [28]. Second, to cater for uncertainties
in input data stochastic methodology can be used instead of fuzzy
methods and thus develop the stochastic efficient resilience strategies
[64]. Third, the concept of dynamics of vulnerability [83] can be
introduced in our methodology and thus develop dynamic portfolio of
resilience strategies. Lastly, a decision support system based on our
methodology can be developed and make it more comprehensive to
include a variety of constraints and objective functions.
Acknowledgements
We are indebted to Prof. Ben Lev, the special issue editor and
three anonymous reviewers for their extensive feedbacks which
greatly improved the paper.
References
[1] Asia News Network, A. Garment sector in B'desh hit hard by instability,
〈http://www.asianewsnet.net/Garment-sector-in-Bdesh-hit-hard-by-instabil
ity-45844.html〉; 2013 [accessed 18.06.13].
[2] Babbie E. The practice of social research. 11th ed. Belmont, CA, USA: Thomson
Wadsworth; 2007.
[3] Barratt M, Choi TY, Li M. Qualitative case studies in operations management:
trends, research outcomes, and future research implications. Journal of
Operations Management 2011;29(4):329–42.
[4] Bevilacqua M, Ciarapica FE, Giacchetta G. A fuzzy-QFD approach to supplier
selection. Journal of Purchasing and Supply Management 2006;12(1):14–27.
[5] BGMEA. Bangladesh apparel and textiles exposition, 〈http://www.bgmea.
com.bd/batexpo/index.htm〉; 2012 [accessed 04.04.13].
[6] Bhattacharya A, Sarkar B, Mukherjee SK. Integrating AHP with QFD for robot
selection under requirement perspective. International Journal of Production
Research 2005;43(17):3671–85.
[7] Blackhurst JV, Scheibe KP, Johnson DJ. Supplier risk assessment and mon-
itoring for the automotive industry. International Journal of Physical Dis-
tribution & Logistics Management 2008;38(2):143–65.
[8] Blos MF, Quaddus M, Wee HM, Watanabe K. Supply chain risk management
(SCRM): a case study on the automotive and electronic industries in Brazil.
Supply Chain Management: An International Journal 2009;14(4):247–52.
[9] Bode C, Wagner SM, Petersen KJ, Ellram LM. Understanding responses to
supply chain disruptions: insights from information processing and resource
dependence perspectives. Academy of Management Journal 2011;54
(4):833–56.
[10] Bottani E, Rizzi A. Strategic management of logistics service: a fuzzy QFD
approach. International Journal of Production Economics 2006;103
(2):585–99.
[11] Braunscheidel MJ, Suresh NC. The organizational antecedents of a firm’s
supply chain agility for risk mitigation and response. Journal of Operations
Management 2009;27(2):119–40.
[12] Bryman A. Integrating quantitative and qualitative research: how is it done?
Qualitative Research 2006;6(1):97–113.
[13] Büyüközkan G, Berkol C. Designing a sustainable supply chain using an
integrated analytic network process and goal programming approach in
quality function deployment. Expert Systems with Applications 2011;38
(11):13731–48.
[14] Calder BJ, Phillips LW, Tybout AM. The concept of external validity. Journal of
Consumer Research 1982;9(3):240–4.
[15] Carnevalli JA, Miguel PC. Review, analysis and classification of the literature
on QFD – types of research, difficulties and benefits. International Journal of
Production Economics 2008;114(2):737–54.
[16] Carpenter S, Walker B, Anderies JM, Abel N. From metaphor to measurement:
resilience of what to what? Ecosystems 2001;4(8):765–81.
[17] Chan LK, Wu ML. Quality function deployment: a literature review. European
Journal of Operational Research 2002;143(3):463–97.
[18] Chan LK, Wu ML. Prioritizing the technical measures in Quality Function
Deployment. Quality Engineering 1998;10(3):467–79.
[19] Chen LH, Weng MC. An evaluation approach to engineering design in QFD
processes using fuzzy goal programming models. European Journal of
Operational Research 2006;172(1):230–48.
[20] Choi Y, Ye X, Zhao L, Luo AC. Optimizing enterprise risk management: a
literature review and critical analysis of the work of Wu and Olson. Annals of
Operations Research 2015:1–20.
[21] Choudhury SR, Hossain MG. Post-MFA issues and challenges: social dimen-
sion. Dhaka, Bangladesh: Ministry of Commerce; 2005.
[22] Chowdhury GH, Sarker MAR, Afroze R. Recent unrest in the RMG sector of
Bangladesh: is this an outcome of poor labour practices? International
Journal of Business and Management 2012;7(3):206–18.
[23] Chowdhury MH, Dewan MNA, Quaddus MA. Supply chain resilience to
mitigate disruptions: a QFD approach. In: Pacific Asia Conference on
Information Systems (PACIS). Ho Chi Minh city, Vietnam: AIS; 2012.
[24] Christopher M, Peck H. Building the resilient supply chain. International
Journal of Logistics Management 2004;15(2):1–13.
[25] Colicchia C, Dallaria F, Melacini M. Increasing supply chain resilience in a
global sourcing context. Production Planning & Control 2010;21(7):680–94.
[26] Craighead CW, Blackhurst J, Rungtusanatham MJ, Handfield RB. The severity
of supply chain disruptions: design characteristics and mitigation capabil-
ities. Decision Sciences 2007;38(1):131–56.
[27] Creswell JW, Clark VLP. Designing and conducting mixed methods research.
Thousand Oaks, CA: Sage Publications; 2007.
[28] Cristiano JJ, White III CC, Liker JK. Application of multiattribute decision
analysis to quality function deployment for target setting. IEEE Transactions
on Systems, Man, and Cybernetics, Part C: Applications and Reviews 2001;31
(3):366–82.
[29] Crowe TJ, Cheng C-C. Using quality function deployment in manufacturing
strategic planning. International Journal of Operations & Production Man-
agement 1996;16(4):35–48.
[30] Deb K. Multi-objective optimization. In: Burke EK, Kendall G, editors. Search
methodologies: introductory tutorials in optimization and decision support
techniques. New York: Springer; 2014. p. 403–49.
[31] Delice EK, Güngör Z. A mixed integer goal programming model for discrete
values of design requirements in QFD. International Journal of Production
Research 2011;49(10):2941–57.
[32] Dickinson G. Enterprise risk management: its origins and conceptual
foundation. Geneva Papers on Risk and Insurance – Issues and Practice
2001;26(3):360–6.
[33] Duclos LK, Vokurka RJ, Lummus RR. Delphi study on supply chain flexibility.
International Journal of Production Research 2005;43(13):2687–708.
[34] Faisal MN. Managing risk in small and medium enterprises (SMEs) supply
chains' using quality function deployment approach. International Journal of
Operations Research and Information Systems 2013;4(1):64–83.
[35] Fennell ML, Alexander JA. Organizational boundary spanning in institutio-
nalized environments. Academy of Management Journal 1987;30(3):456–76.
[36] Ferdousi F, Ahmed A. An investigation of manufacturing performance
improvement through lean production: a study on Bangladeshi Garment
Firms. International Journal of Business and Management 2009;4(9):106–16.
[37] Fibre2fashion News Desk. US suspends GSP facility to Bangladesh, 〈http://
www.fibre2fashion.com/news/apparel-news/newsdetails.aspx?news_
id=147897〉; 2013 [accessed 04.06.13].
[38] Fiksel J. Designing resilient, sustainable systems. Environmental Science &
Technology 2003;37(23):5330–9.
[39] Gaudenzi B, Borghesi A. Managing risks in the supply chain using the AHP
method. The International Journal of Logistics Management 2006;17
(1):114–36.
[40] Goh M, Lim JYS, Meng F. A stochastic model for risk management in global
supply chain networks. European Journal of Operations Research
2007;182:164–73.
[41] González ME, Quesada G, Picado F, Eckelman CA. Customer satisfaction using
QFD: an e-banking case. Managing Service Quality 2004;14(4):317–30.
[42] Haider MZ. Competitiveness of the Bangladesh Ready-made Garment
Industry in major international markets. Asia-Pacific Trade and Investment
Review 2007;3(1):3–27.
[43] Han SB, Chen SK, Ebrahimpour M, Sodhi MS. A conceptual QFD planning
model. The International Journal of Quality & Reliability Management
2001;18(8):796–812.
[44] Heckmann I, Comes T, Nickel S. A critical review on supply chain risk –
definition, measure and modeling. Omega 2015;52:119–32.
[45] Hendricks KB, Singhal VR. The effect of supply chain glitches on shareholder
wealth. Journal of Operations Management 2003;21:501–22.
[46] Holling CS. Resilience and stability of ecological systems. Annual Review of
Ecology and Systematics 1973;4:1–23.
[47] Hossan CG, Sarker MAR, Afroze R. Recent unrest in the RMG Sector of
Bangladesh: is this an outcome of poor labour practices? International
Journal of Business and Management 2012;7(3):206.
[48] Islam MA, Begum MN, Rashed CAA. Operational disturbances and their
impact on the manufacturing business—an empirical study in the RMG
Sector of Bangladesh. International Journal of Research in Management &
Technology 2012;2(2):184–91.
[49] Islam MA, Deegan C. Motivations for an organisation within a developing
country to report social responsibility information: evidence from Bangla-
desh. Accounting, Auditing & Accountability Journal 2008;21(6):850–74.
[50] Jiang B, Baker RC, Frazier GV. An analysis of job dissatisfaction and turnover
to reduce global supply chain risk: evidence from China. Journal of Opera-
tions Management 2009;27(2):169–84.
[51] Johnson RB, Onwuegbuzie AJ. Mixed methods research: a research paradigm
whose time has come. Educational Researcher 2004;33(7):14–26.
[52] Jugulum R, Sefik M. Building a robust manufacturing strategy. Computers &
Industrial Engineering 1998;35(1):225–8.
[53] Jüttner U, Maklan S. Supply chain resilience in the global financial crisis: an
empirical study. Supply Chain Management: An International Journal
2011;16(4):246–59.
 Md.M.H. Chowdhury, M.A. Quaddus / Omega 57 (2015) 5–21
[54] Kamvysi K, Gotzamani K, Andronikidis A, Georgiou A. Capturing and
prioritizing students’ requirements for course design by embedding Fuzzy-
AHP and linear programming in QFD. European Journal of Operational
Research 2014;237:1083–94.
[55] Karim S. Coping with post-MFA challenges: strategic response for Bangla-
desh RMG sector. Dhaka: CPD; 2003.
[56] Karsak EE. Fuzzy multiple objective decision making approach to prioritize
design requirements in quality function deployment. International Journal of
Production Research 2004;42(18):3957–74.
[57] Karsak EE. Fuzzy multiple objective programming framework to prioritize
design requirements in quality function deployment. Computers & Industrial
Engineering 2004;47(2):149–63.
[58] Karsak EE, Özogul CO. An integrated decision making approach for ERP
system selection. Expert Systems with Applications 2009;36(1):660–7.
[59] Karsak EE, Sozer S, Alptekin SE. Product planning in quality function
deployment using a combined analytic network process and goal program-
ming approach. Computers & industrial engineering 2002;44(1):171–90.
[60] Khondker BH, Razzak A, Ahmed N. Exports, employment and working
conditions: emerging issues in the post-MFA RMG industry(revised draft).
International Labour Office; 2005.
[61] Kleindorfer PR, Saad GH. Managing disruption risks in supply chains.
Production & Operations Management 2005;14(1):53–68.
[62] Köksalan MM, Wallenius J, Zionts S. Multiple criteria decision making: from
early history to the 21st century. World Scientific; 2011.
[63] Larbani M, Aouni B. A new approach for generating efficient solutions within
the goal programming model. Journal of the Operational Research Society
2011;62(1):175–82.
[64] Lee DE, Lim TK, Arditi D. Automated stochastic quality function deployment
system for measuring the quality performance of design/build contractors.
Automation in Construction 2009;18(3):348–56.
[65] Lee AH, Kang HY, Yang CY, Lin CY. An evaluation framework for product
planning using FANP, QFD and multi-choice goal programming. International
Journal of Production Research 2010;48(13):3977–97.
[66] Lin CM, Lee Y, Ho T. Applying integrated DEA/AHP to evaluate the economic
performance of local governments in China. European Journal of Operational
Research 2011;209:129–40.
[67] Lockwood JL, Pimm SL. Biological diversity: species: would any of them be
missed? Current Biology 1994;4(5):455–7.
[68] Ludwig D, Walker B, Holling CS. Sustainability, stability and resilience.
Conservation Ecology 1996;1:1–27.
[69] Martin S. The cost of restoration as a way of defining resilience: a viability
approach applied to a model of lake eutrophication. Ecology and Society
2004;9(2):8.
[70] Mitroff IL, Alpaslan MC. Preparing for evil. Harvard Business Review 2003;81
(4):109–15.
[71] Mizgier KJ, Wagner SM, Holyst JA. Modeling defaults of companies in multi-
stage supply chain networks. International Journal of Production Economics
2012;135(1):14–23.
[72] Narasimhan R, Kim SW. Information system utilization strategy for supply
chain integration. Journal of Business Logistics 2001;22(2):51–75.
[73] Nuruzzaman A Haque, Rafiq A. Is Bangladeshi RMG Sector fit in the Global
Apparel Business? Analyses the supply chain management The South East
Asian Journal of Management 2010;4(1):53–72.
[74] Olson DL, Wu D. Enterprise risk management models. Heidelberg: Springer;
2010.
[75] Olson DL, Wu D. A review of enterprise risk management in supply chain.
Kybernetes 2010;39(5):694–706.
[76] Onwuegbuzie AJ, Leech NL. On becoming a pragmatic researcher: the
importance of combining quantitative and qualitative research methodolo-
gies. International Journal of Social Research Methodology 2005;8(5):
375–387.
[77] Park T, Kim KJ. Determination of an optimal set of design requirements using
house of quality. Journal of Operations Management 1998;16(5):569–81.
[78] Peck H. Reconciling supply chain vulnerability, risk and supply chain
management. International Journal of Logistics Research and Applications
2006;9(2):127–42.
[79] Pettit TJ, Croxton KL, Fiksel J. Ensuring supply chain resilience: development
and implementation of an assessment tool. Journal of Business Logistics
2013;34(1):46–76.
[80] Pettit TJ, Fiksel J, Croxton KL. Ensuring supply chain resilience: development
of a conceptual framework. Journal of Business Logistics 2010;31(1):1–21.
[81] Ponomarov SY, Holcomb MC. Understanding the concept of supply chain
resilience. International Journal of Logistics Management 2009;20(1):
124–139.
21
[82] Pujawan IN, Geraldin LH. House of risk: a model for proactive supply chain
risk management. Business Process Management Journal 2009;15
(6):953–67.
[83] Raharjo H, Xie M, Brombacher AC. A systematic methodology to deal with
the dynamics of customer needs in Quality Function Deployment. Expert
Systems with Applications 2011;38(4):3653–62.
[84] Ramanathan R, Yunfeng J. Incorporating cost and environmental factors in
quality function deployment using data envelopment analysis. Omega
2009;37:711–23.
[85] Rose A. Defining and measuring economic resilience to disaster. Disaster
Prevention and Management 2004;13(4):307–14.
[86] Saaty TL. AHP: the analytic hierarchy process. New York: McGraw-Hill; 1980.
[87] Sawik T. On the fair optimization of cost and customer service level in a
supply chain under disruption risks. Omega 2015;53:58–66.
[88] Schmitt A, Sun S, Snyder L, Shen Z. Centralization versus decentralization:
risk pooling, risk diversification, and supply chain disruptions. Omega
2015;52:201–12.
[89] Schwartz B, Ben-Haim Y, Dacso C. What makes a good decision? Robust
satisficing as a normative standard of rational decision making Journal for
the Theory of Social Behaviour 2011;41(2):209–27.
[90] Sheffi Y, Rice JB. A supply Chain View of the resilient Enterprise. MIT Sloan
management 2005;47(1):47–8.
[91] Shin WS, Ravindran A. Interactive multiple objective optimization: Survey I—
continuous case. Computers & Operations Research 1991;18(1):97–114.
[92] Simchi-Levi D, Schmidt W, Wei Y. From superstorms to factory fires
managing unpredictable supply-chain disruptions. Harvard Business Review
2014;92(1–2):96–101.
[93] Simon H. Theories of bounded rationality. In: McGuire CB, Radner R, editors.
Decisions and organizations. North Holland; 1972.
[94] Stratton B. The refined focus of automotive quality. Quality Progress 1989;22
(10):47–50.
[95] Tang CS. Robust strategies for mitigating supply chain disruptions. Interna-
tional Journal of Logistics Research and Applications 2006;9(1):33–45.
[96] Tashakkori A, C Teddlie C, editors. Sage handbook of mixed methods in social
& behavioral research. Sage; 2010.
[97] Tewari M. Is price and cost competitiveness enough for apparel firms to gain
market share in the world after quotas? A review Global Economy Journal
2006;6(4):1–46.
[98] Tomlin B. On the value of mitigation and contingency strategies for
managing supply chain disruption risks. Management Science 2006;52
(5):639–57.
[99] Vugrin ED, Warren DE, Ehlen MA. A resilience assessment framework for
infrastructure and economic systems: quantitative and qualitative resilience
analysis of petrochemical supply chains to a hurricane. Process Safety
Progress 2011;30(3):280–90.
[100] Wagner SM, Neshat N. Assessing the vulnerability of supply chains using
graph theory. International Journal of Production Economics 2010;126
(1):121–9.
[101] Wagner SM, Neshat N. A comparison of supply chain vulnerability indices for
different categories of firms. International Journal of Production Research
2012;50(11):2877–91.
[102] Wang HF, Hong WK. An integrated service strategy by QFD approach: a case
of a telecom company in Taiwan. International Journal of Management and
Decision Making 2007;8(2–4):251–67.
[103] Wasserman GS. On how to prioritize design requirements during the QFD
planning process. IIE Transactions 1993;25(3):59–65.
[104] Wu D, Olson DL. On the special issue of risk issues in operations: methods
and tools. Production Planning and Control 2009;20(4):293–4.
[105] Wu DD, Olson D. Enterprise risk management: a DEA VaR approach in vendor
selection. International Journal of Production research 2010;48(16):4919–32.
[106] Wu DD, Olson D. Enterprise risk management: coping with model risk in a
large bank. Journal of the Operational Research Society 2010;61(2):179–90.
[107] Wu D, Olson DL, Birge JR. Introduction to special issue on "Enterprise risk
management in operations”. International Journal of Production Economics
2011;134:1–2.
[108] Wu D, Chen SH, Olson DL. Business intelligence in risk management: some
recent progress. Information Sciences 2014;256:1–7.
[109] Wu T, Blackhurst J, Chidambaram V. A model for inbound supply risk
analysis. Computers in Industry 2006;57(4):350–65.
[110] Zhou M. Fuzzy logic and optimization models for implementing QFD.
Computers & Industrial Engineering 1998;35(1):237–40.
[111] Zsidisin GA, Melnyk SA, Ragatz GL. An institutional theory perspective of
business continuity planning for purchasing and supply management. Inter-
national Journal of Production Research 2005;43(16):3401–20.