Author's Accepted Manuscript

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

www.elsevier.com/locate/omega

PII:DOI:Reference:

S0305-0483(15)00133-4 http://dx.doi.org/10.1016/j.omega.2015.05.016 OME1557

To appear in: Omega

Received date: 1 January 2014Accepted date: 22 May 2015

Cite this article as: Md. Maruf Hossan Chowdhury, Mohammed A. Quaddus, A multiple objective optimization based QFD approach for efficient resilient strategies to mitigate supply chain vulnerabilities: The case of garment industry of Bangladesh, Omega, http://dx.doi.org/10.1016/j.omega.2015.05.016

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

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

School of Marketing, Curtin University, Perth, WA, Australia. marufhossan@gmail.com

Mohammed A. Quaddus1

School of Marketing, Curtin University, Perth, WA, Australia, m.quaddus@curtin.edu.au

* It is an equally authored paper. Author’s names are listed in alphabetical order.

1 Corresponding author

1

A multiple objective optimization based QFD approach for efficient resilient strategies to mitigate supply chain vulnerabilities: The case of garment industry of Bangladesh

Abstract

With the upsurge of frequent disruptive events, organizations have become more

vulnerable to the consequences of these disruptive events. As a result, the need for more

resilient supply chain (SC) to mitigate the vulnerabilities has become paramount. Supply

chain resilience (SCR) has been discussed in the literature and resilience index has been

developed, but developing and selecting a portfolio of supply chain resilience capabilities

in order to mitigate the vulnerabilities have not been studied. In this research we develop

a 0-1 multi-objective optimization model based on QFD methodology. Our multi-objective

method is interactive and interacts with the decision makers to choose the most satisfactory efficient

portfolio of supply chain resilience strategies. We apply our methodology to three large ready-made

garment (RMG) companies of Bangladesh. Results show that lack of materials (high dependence on

imported materials), disruptions in utility supply, increased competition (and hence competitive

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.

Keywords: Supply chain, resilience-efficiency, vulnerability, multi-objective model, AHP, QFD.

1. 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 accounts for 78.6 percent of countries export earnings (BGMEA 2012).

RMG sector also immensely contributes in reducing the high rate of women

2

unemployment in the country as eighty percent of the garments workers are

women (BGMEA 2012). 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 numerous Supply Chain (SC)

disruptions (Islam and Deegan 2008; Haider 2007). 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 (Asia News Network 2013). 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 eleven

hundred workers (Fibre2fashion News Desk 2013). 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

(2011) 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 (2004), Ponomarov and Holcomb

(2009) and Juttner and Maklan (2011) 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 (2010, 2013) developed a supply chain resilience framework by identifying

seven categories of vulnerabilities and creating supply chain capabilities along fourteen

areas (sourcing, 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

3

performance (Vurgin et al. 2011, Pettit et al. 2010), achieve sustainable competitive

advantage (Ponomarov and Holcomb (2009), gain market share in competitive

environments (Sheffi and Rice (2005), and decreases vulnerabilities (Pettit et al. 2010,

2013, Juttner and Maklan (2011). However current literature lacks in 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. (2010, 2013) in their framework highlight fourteen

areas of supply chain resilience capabilities 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,

Larbani and Aouni 2011; 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 (eg. minimum cost of implementation), and

(ii) portfolio of chosen resilience capabilities must be efficient (or non-dominated) from

multiple objectives perspective (Larbani and Aouni 2011).

4

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; Pettit et al. 2010; Fiksel

2003; Ponomarov and Holcomb 2009; Sheffi and Rice 2005; Christopher and Peck

2004; 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 Deployment (QFD) (Park and Kim

1998; Wang and Hong 2007) 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

(Faisal 2013). However we shall use Analytical Hierarchy Process (AHP) (Saaty 1980)

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

5

to manage all risks that an organization faces (Dickinson 2001, Olson and Wu 2010a).

A recent literature review (Choi et al. 2015) 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 (Wu and Olson 2010b). Wu and Olson (2009), Wu et al

(2011, 2014) 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 (Choi et al. 2015) 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” (Peck 2006).

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 (2010a) developed a DEA based value at risk (VaR) model to

manage supply risks, specifically vendor selection problem. Olson and Wu (2010b)

presented a review of ERM in supply chain. Jian et al (2009) 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.

6

1. 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 (BGMEA 2012). 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 (Islam and Deegan

2008; Chowdhury, Sarker, and Afroze 2012; Haider 2007). Furthermore, increased lead

time and cost due to disruptions 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 (Haider 2007), 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 2005-06 to 15% in 2008-09 (Chowdhury, Sarker, and Afroze 2012).

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.

7

(Insert Table 1 about here)

3. Literature Review

3.1 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 (Blos et al. 2009; Juttner 2011). Wagner and Neshat (2012) 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 terrorism; among many others (Colicchia, Dallaria,

and Melacini 2010; Kleindorfer and Saad 2005; Blackhurst, Scheibe, and Johnson 2008).

Schmitt et al. (2015) study the impact of supply disruptions on both centralized and the

decentralized multi-location system. Sawik (2015) presents a bi-objective (minimum cost

and maximum service level) model to study the impact of supplier disruptions Mizgier et al.

(2012) 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 (2003) show that

announcement of SC disruptions, such as, operational issue or delay in shipment cause

decrease in shareholder value significantly. Kleindorfer and Saad (2005) 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,

8

natural hazards which include earthquakes, hurricanes, storms; and thirdly, terrorism

or political instability, among others. Blos et al. (2009) suggest four major sources of

SC vulnerabilities such as, financial vulnerability, strategic vulnerability, hazard

vulnerability, and operations vulnerability. Similarly, a number of researchers (such

as Pettit et al. 2013; Christopher and Peck 2004; Sheffi and Rice 2005; Blackhurst et

al. 2008; Kleindorfer and Saad 2005 ; among others) discussed SC vulnerability

factors which are summarized in Table 2 in terms of various vulnerability 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.

(Insert Table 2 about here)

3.2 Supply chain resilience capability

Resilience is a multidisciplinary concept. Holling (1973) 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 (Mitroff and Alpasan 2003; Ponomarov and Holcomb 2009; Christopher and

Peck 2004). Heckmann et al. (2015) 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 (Ponomarov and Holcomb 2009, Heckmann et al. 2015), 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.

9

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 (Juttner

and Maklan 2011). Otherwise, the consequence will be the discontinuity of SC

operations which will adversely affect both revenue and cost of the whole chain

(Ponomarov and Holcomb 2009).

Researchers in supply chain management (SCM) emphasized on capabilities,

such as, adaptability, pro-activeness, diversity, flexibility, efficiency, reserve capacity,

integration, market development, cohesion, control, connectedness to measure

resilience (Pettit et al. 2010; Ponomarov and Holcomb 2009; Tomlin 2006). Table 3

summarizes these capabilities. It is noted that two fundamental organizational resilience

capabilities are buffering and bridging (Bode et al. 2011, Fennell et al. 1987). 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 (Bode et al. 2011). 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 (Bode et al. 2011). 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 (Carpenter et

al. 2001). Therefore, in order to deal with resilience it is important to identify the

vulnerability factors of the specific SC and the corresponding mitigation capabilities.

(Insert Table 3 about here)

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

10

ways: (i) based on the extent of systems departure from desired state (Holling 1973;

Ludwig, Walker, and Holling 1996), (ii) based on recovery time after disaster or

disruptions (Simchi-levi et al. 2014; Sheffi and Rice 2005), (iii) based on reduction of

impact and consequences (Rose 2004; Lockwood and Pimm 1994), (iv) based on time

to respond (Sheffi and Rice 2005) and (v) based on cost of recovery (Martin 2004;

Vugrin et al. 2011).

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 (Vugrin et al. 2011). Literature lacks significantly in conceptualizing

resilience efficiency. In this paper, therefore, we offer a unique and elegant

operationalization of resilience efficiency in the methodology section.

3.3 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 (González et al. 2004). The benefits of QFD model have been

highlighted by various researchers. For example Chan and Wu (2002) 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

(2008) in another review of QFD highlight QFD’s ability in adapting into various research

methods from modelling, theoretical-conceptual to action-research, experimental. Because

of its wide applicability QFD has been used in various fields, such as, determining

customer needs (Stratton 1989), developing priorities (Han et al. 1998), manufacturing

strategies (Crowe and Cheng 1996; Jugulum and Sefik 1998), logistics and SCM (Bottani

and Rizzi 2006; Bevilacqua, Ciarapica, and Giacchetta 2006).

11

QFD has also been applied successfully for supply chain risk identification and

mitigation (Pujawan and Geraldin 2009; Faisal 2013). 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.

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 organizational problems are referred to as HOWs

(resilience capabilities). The basic QFD framework is shown in Figure 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

Figure 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 Figure 1), technical measures or design

requirements (HOWs) and their relationships with the WHATs, and the correlation

between the HOWs (see Figure 1).

(Insert Figure 1 about here)

In our case row elements CRi (WHATs) of figure 1 represent the vulnerabilities

that RMG supply chains are facing currently. The column elements DRj (HOWs) are

the resilience strategies or capabilities to mitigate the vulnerabilities. We define the

elements of the relationship matrix Rij (see figure 1) as the “extent of mitigating the

specific vulnerabilities by specific resiliency strategies”. In line with QFD literature

(Chan and Wu 2002, Pujawan and Geraldin 2009; Faisal 2013) we measure Rij using

the sacle of 9 (strong mitigation), 3 (moderate mitigation), 1 (little mitigation) and 0

(no mitigation). The AI and RI in figure 1 are the absolute and relative importances of

the HOWs (resilience strategies) which are found as follows (Park and Kim 1998):

12

m

AI j = ∑wi Rij∀ j , j = 1,……, n (1)i =1

where,

AIj = absolute importance of jth design requirement (DR) (or, resilience strategy) wi = weight of the ith supply chain vulnerability.

Rij = relationship value; extent of mitigating ith vulnerability by jth resilience strategy

(9, 3, 1, or 0)

n = number of design requirements (resilience strategies);

m = number of supply chain vulnerabilities.

It is noted that in our case AI j is interpreted as “total resilience” of the jth resilience

strategy to mitigate the vulnerabilities.

The relative importance (resilience) of the resilience strategy j is:

RI j =

AI j

(2)∑ n

AI jj =1

The correlation between the DRjs (HOWs) (see Figure 1) plays a significant part in many

QFD applications including ours. It represents the extent of correlation (similarities) when

two HOWs are implemented. Literature suggests that there is some degree of dependencies

among the HOWs in real applications (Wasserman 1993, Park and Kim 1998). If HOWi

and HOWj are correlated then there is cost savings sij in their

implementation (Park and Kim 1998). These sij’s need to be estimated from the decision

13

makers. We argue that in most business, management and social science applications

(inclduing our present case) some HOWs will be highly correlated.

3.3.1 Integration with Analytical Hierarchy Process (AHP)

In basic QFD three fundamental data are needed: (i) relative importance of WHATs,

(ii) the relationships between the WHATs and HOWs, and (iii) the correlation between the

HOWs. In finding the relative importance of WHATs (i.e.wi’s; see Figure 1) we plan to use

AHP. AHP was originally developed by Saaty (1980) which is an well-established multi-

criteria decision making approach that employs a unique method of hierarchical structuring

of a problem and subsequent 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 (Saaty 1980). AHP is frequently used in QFD

process, for instance see, Kamvysi et al. (2014), Park and Kim (1998), Bhattacharya et al.

(2005), Chan and Wu (1998); among others. Methodologically QFD has been frequently

combined with other tools to increase its robustness and applicability. For example

Ramanathan and Yunfeng (2009) combined QFD with Data Envelopment Analysis (DEA)

and applied it to design security fasteners in a Chinese company. Lin et al (2011) combined

QFD with DEA and AHP and applied it to evaluate the economic performance of local

governments in China.

3.3.2 QFD Optimization

It is noted that in QFD an optimization method is always needed to determine the most

desirable HOWs to satsify the WHATs (Delice and Gunger 2011, Karsak 2004b) under

14

certain constarints. For example, Park and Kim (1998) optimize total absolute importance

of the HOWs by formulating a 0-1 linear and quadratic program to find the most deriable

HOWs under budget constraint. Wasserman (1993) formulates a linear program to find the

most desrirable design requirements, HOWs, under budget constraints. Zhou (1998)

proposes a mixed integer linear program that maximizes an unitily function under budget,

technological feasibiity and competition constraints to determine the HOWs.

Multiple objective optimization approaches (Koksalan et al. 2011, Deb 2014) have

also been used in determining the HOWs. For example, Karsak et al (2002) formulated a 0-

1 goal program combining analytical network process in QFD application in product

planning. Delic and Gunger (2011) proposes a mixed integer goal program where the

objectives 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 Ozogul (2009), Chen and

Weng (2006), Lee et al. (2010), Buyukozkan and Berkol (2011); among others. Karsak

(2004a and 2004b) 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).

4. Proposed Methodology to Determine Efficient Resilience Capabilities in QFD

15

The concept of efficiency and generating efficient solutions is prevalent in multiple

objective decision making domain (Koksalan et al. 2011, Deb 2014). A general multiple

objective decision problem is represented as follows:

Max (Min) fi(X) = Ci(X); i= … p…………….. (3)

Subject to: gj(X) ≤ bj; j = 1 ….…q

Where, X = (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 X* 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 i = 1, …, p, fi(X) ≥

fi(X*), and fi(X) > fi(X

*) for at least one i. In other words X* 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

problem as: maximize procurement strategies, maximize processing strategies, and

maximize distribution strategies, to mitigate the vulnerabilities.

As presented earlier we define “supply chain resilience efficiency (SCREF)” as

follows:

(i) resilience capability must be resource efficient (eg. minimum cost of implementation),

and

(ii) portfolio of chosen resilience capabilities must be efficient (or non-dominated) from

multiple objectives perspective (Larbani and Aouni 2011)

Vurgrin et al (2011) touched on the efficiency of system resilience. According to the

authors system resilience must use “lowest possible amount of resources” to be efficient.

In line with Vurgrin et al. (2011) we define “Resilience Efficiency” REj as ( AI j / C j ) ,

16

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 optimization problem that will find the efficient portfolio of

resilience strategies to mitigate the vulnerabilities. Pettit et al. (2013) 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 formulated as follows:

Max f1 (X) = ∑RE j x jj∈n

Max f2 (X) = ∑ RE k xk

k ∈n , k ≠ j

…..(4)

Max fp (X) = ∑ REl xl

l∈n ,l ≠ k ≠ j

n nn

Subject to: ∑c j x j − ∑∑sij xi x j ≤ Bj =1 i =1 j >i

x ∈ X

where, n is the number of resilience startegies; RE j is the resilience 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.

17

It is noted that our constraints are non-linear and similar to Park and Kim (1998). But

depending on the application the constraint set could be formulated in an extensive way

including the budget, technological feasibility and competition constraints (Zhou 1998).

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:

p

Max ∑λi f i ( X )(5)i =1

n n n

Subject to: ∑c j x j − ∑∑sij xi x j ≤ Bj =1 i =1 j >i

x ∈ X

where λi (i = 1, .. p) are positive numbers representing the weights (importance)

attached to the objective function f i ( X ) 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) (Larbani and Aouni 2011). It is noted

that the important weights λi are only needed to find the first efficient solution to

problem (4). It is noted that Karasak (2004a, 2004b) has also determined efficient solution

for multiple objective optimization in QFD. However our multi-objective model

formulation and solution approach are different from Karasak (2004a, 2004b) in the

following ways: (i) Karasak deals with fuzzy multiple objectives, while we deal with non-

fuzzy multiple objective formulation, (ii) Karasak finds fuzzy priorities of the

18

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

preferred than the non-interactive approaches (Shin and Ravindran 1991, pp. 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 = 1, (i = 1, .. p). Solve problem (5). The

solution will be efficient (non-dominated) 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 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.

19

5. Application in RMG Industry of Bangladesh

5.1. Methodology

There are fundamentally two research paradigms: positivist and interpretivist

(Onwuegbuzie and Leech 2005). Positivist paradigm is associated with the quantitative

research method based on specific research questions and hypotheses testing (Johnson

and Onwuegbuzie 2004; Creswell and Clark 2007). Whereas the interpretivist paradigm

relies on the qualitative method and there is subjective interpretation of the researcher

involved (Creswell and Clark 2007). However, in recent times research based on mixed

methods, a combination of qualitative and quantitative methods, has gained popularity

(Bryman 2006), because it assists in increasing the quality, accuracy, validity and

reliability of data (Creswell and Clark 2007; Babbie 2007).

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 (Park and Kim 1998, Wang

and Hong 2007) and which we have embraced in our current study.

Our applications in RMG industry of Bangladesh are conducted in two studies and in

three stages as follows.

5.2. Study 1

20

Our study 1 is conducted on one of the largest apparel manufacturer 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 & 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 conglomerate 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 vulnerabilities.

21

5.2.1. Stage 1

In stage 1 the supply chain vulnerabilities (WHATs) and corresponding 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

companies 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 (Barratt et al. 2011). 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 to 80 minutes.

(Insert Table 4 about here)

Table 5 presents the explicated supply chain vulnerabilities and resilience strategies.

There are twenty-six vulnerabilities. Out of these, seventeen vulnerabilities have been

supported by majority of the respondents. These seventeen vulnerabilities are considered

for further analysis. Corresponding to the seventeen vulnerabilities the respondents

identified thirteen resilience strategies to mitigate the vulnerabilities. It is observed that

most of these vulnerabilities and resilience strategies are consistent with the literature (as

par Table 2 and 3). This adds further validity to our collected data (Barratt et al. 2011).

(Insert Table 5 about here)

5.2.2. Stage 2

In stage 2 we collect the quantitative data (wi, Rij) to find the AIj and RIj (see table 2 and

equations 1 and 2). We also collect the data on costs (cj) of implementing the resilience

22

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 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. (2011) 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 (Barratt et al. 2011).

To find the wi we use Saaty’s (1980) 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 vulnerability i by resiliency strategy j” using the widely used scale of 9, 3, 1 and 0

(Chan and Wu 2002, Pujawan and Geraldin 2009; Faisal 2013). The Rij’s are also

averaged for three respondents. Figure 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. (Insert Figure 2 about here)

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= (4Cm+ Co+ Cp)/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. Figure 2 shows the costs

23

cj’s and the resilience efficiencies REj’s. It is noted that resilience strategy 13 (building

relations with buyers and suppliers) has the highest RE of 0.19 followed by resilience

strategy 8 (forecasting and prediction) of 0.15. The cost figures are in Millions of Taka1.

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 figure 2 shows these savings data. For example, resilience strategies 1 and 10 can be

implemented simultaneously and the estimated savings would be Taka 3.6 Million.

5.2.3. 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) = ‘Procurement’ related strategies which includes strategies ST2, ST4, ST9

and ST13.

f2(X) = ‘Processing’ related strategies which includes ST5, ST6, ST7, ST11 and

ST12.

f3(X) = ‘Distribution’ related strategies which includes ST3, ST8, ST1, and ST10.

Problem (4) now becomes:

Max f1(X) = RE2x2 + RE4x4 + RE9x9 + RE13x13

Max f2(X) = RE5x5 + RE6x6 + RE7x7 + RE11x11 + RE12x12

Max f3(X) = RE3x3 + RE8x8 + RE1x1 + RE10x10

Subject to:

1 Taka is Bangladeshi currency. At the time of this study the exchange rate was 1 US$ = 77 Taka.

24

c1 x1 + c2 x2 + c3 x3 + c4 x4 + c5 x5 + c6 x6 + c7 x7 + c8 x8 + c9 x9 + c10 x10 + c11 x11 + c12 x12 + c13 x13 -

S1,6

x1 x

6 -

S

1,10 x

1 x

10 −

S

2,4 x

2 x

4 -

S

3,8 x

3 x

8 -

S

3,9 x

3 x

9 -

S

3,10 x

3 x

10 -

S

4,8 x

4 x

8 -,-,-

S 4,13 x

4 x

13 - S

5,6 x

5 x

6 - S

5,7 x

5 x

7 - S

6,7 x

6 x

7 - S

7,10 x

7 x

10 - S

7,11 x

7 x

11 - S

7,13 x

7 x

13 - S

8,13 x

8 x

13 - S

9,13 x

9 x

13 -

S10,11

x10

x11

- S

10,12 x

10 x

12 -

S

11,13 x

11 x

13 ≤

B

where xj = 0 or 1.

The values of REj and cj are obtained from figure 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 par 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 0.416, 0.411 and 0.316 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 implemented (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 figure 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.

(Insert Table 6 about here)

25

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 0.327 (for f1; a deviation of 21.65% from the optimal

value of f1), 0.329 (for f2; a deviation of 19.95% from the optimal value of f2) and 0.231

(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 solutions. The steps 3 and 4 come into action now. After some

deliberations the weights of (0.2, 0.5, and 0.3) 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

0.327 (for f1; a deviation of 21.65% from the optimal value of f1), 0.411 (for f2; a deviation

of 0% from the optimal value of f2), and 0.149 (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 (0.5) 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

(0.327, 0.329, and 0.231). 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 adoption), 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),

26

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.

(Insert Table 7 about here)

5.3 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.

5.3.1 Stages 1 and 2

Tashakkori and Teddlie (2010) 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 study 1 and 5 are new. Table 8 shows these vulnerabilities. The red coloured and bolded

ones are the new vulnerabilities. It is noted that for the next stage of the analysis (ie.

developing the supply chain resilience model) the highest three weighted vulnerabilities

from each group were considered. For example, for hazard vulnerability

27

group HV1, HV2, and HV3 were selected which had the highest weights among the five

hazard vulnerabilities (see figure 3).

(Insert Table 8 about here)

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 Figure 3. Unlike study 1 the

resilience strategy 9 (Demand forecasting) has the highest RE of 0.108 followed by

resilience strategy 5 (Maintaining reserve capacity). 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.

(Insert Figure 3 about here)

5.3.2 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:

28

f1(X) = ‘Procurement’ related strategies comprising ST3, ST5, ST10 and ST14.

f2(X) = ‘Processing’ related strategies comprising ST1, ST6, ST7, ST8, ST12, ST13.

f3(X) = ‘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 the satisfactory efficient portfolio

for study 2. First, each objective function is optimized separately. The optimal solutions

and corresponding 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 0.163, 0.173 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.

(Insert Table 9 about here)

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 objectives now and a weighting scheme of (0.25, 0.45, and

0.3) 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 stills offers the optimal value for objective f1, with some

improvement in objective f2 at the sacrifice of objective f3. The two executives preferred

this solution compared to the equal weighting solution. It is observed from

29

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.

(Insert Table 10 about here)

5.4 Study 3

Study 2 was primarily conducted to ensure the transferability (Tashakkori and Teddlie

2010) of results and findings from one context to another context (ie. transferring

vulnerabilities and resilience strategies from study 1 to study 2). To ensure further

external validity (Calder et al. 1982) of our method we have conducted study 3 on another

garment manufacturer 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, UK and Australia. It has

30

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.

5.4.1 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 study 1 and 2. There are three hazard, five strategic,

three financial, five operational, two infrastructural, and five demand-supply

vulnerabilities. Among the twenty three vulnerabilities of study 3, nine vulnerabilities are

common to both study 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 (ie. developing the supply chain model) the

highest three weighted vulnerabilities from each group were considered.

Table 11 also shows the fourteen resilience strategies. It is noted that six of them are

common to both study 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.

31

(Insert Table 11 about here)

In the next stage of the analysis the supply chain resilience model was developed as shown

in Figure 4. It is noted that the weight (wi) of the vulnerabilities, relationship value (Rij),

cost (cj) and savings (sij) were also collected form the two executives of the company

which are shown in Figure 4. It is observed that resilience strategy 5 (customer relationship

development) has the highest RE value of 0.332 followed by resilience strategy 6

(developing new buyers and markets) and resilience strategy 8 (strict quality control at

different stage). Interestingly these rankings of strategies are quite different from study 1

and 2.

5.4.2 Stage 3

Like study 1 and 2 we then develop the 0-1 multiple objective model and find the

satisfactory efficient resilient strategies. As par the two executives of the company there

are three objectives to be optimized as follows:

f1(X) = ‘Procurement’ related strategies comprising ST1, ST10, ST11 and ST12.

f2(X) = ‘Processing’ related strategies comprising ST2, ST3, ST4, ST7, ST8, ST9.

f3(X) = ‘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. Step 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 (0.2, 0.4, 0.35) for objectives

32

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

suppliers), St11 (Supplier selection & evaluation), St12 (Developing relationship with

suppliers), St2 (Risk management team), St7 (Flexibility in production), St8 (Strict quality

control at different stage), St9 (Training & development), St5 (Customer relationship

development), and St6 (Developing new buyers & markets).

(Insert tables 12 and 13 about here).

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 study 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.

6. Discussions and Implications

This research aimed to develop resilience capabilities of Bangladesh 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 qualitative and

quantitative approaches (Creswell and Clark 2007). We defined supply chain resilience

efficiency (SCREF) and develop an interactive methodology to determine efficient

resilience capabilities based on QFD approach. The proposed methodology is then

33

applied on three large RMG companies in Bangladesh. The three studies were conducted

to ensure external validity (and transferability) (Calder et al. 1982, Tashakkori and Teddlie

2010) of our proposed methodology. Our experience indicates that the proposed

methodology can be successfully applied to determine efficient portfolio of resilience

strategies to mitigate vulnerabilities.

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 figure 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 figure 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 figure 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 considered as possible vulnerabilities of studies 1

and 2 (see Table 11). Being family owned, company 3 works

34

in an extreme competitive environment. Hence these two vulnerabilities are very

significant for the company.

As pointed out earlier the three companies have different management attitude while

mitigating vulnerabilities via resilience 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), ‘processing’ (f2) and ‘distribution’ (f3) strategies. Their preferred

portfolio of resilience strategies included seven out of possible thirteen 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 fourteen 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 strategies (among 14 strategies) with three from procurement objective,

four from processing objective and two from distribution objective.

6.1 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.

35

This extends traditional QFD based optimization method to determine the most desirable

HOWs (Delice and Gunger 2011, Karsak 2004b), 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 (Simon 1972, Schwartz et al 2011).

It is worthwhile to compare our methodology with existing methodologies which

deal with vulnerabilities in supply chains. Wagner and Neshat (2010) 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 (Bloss et

al. 2009). 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. (2007) also developed a quantitative stochastic model and

methodology to deal with supply chains and vulnerabilities. However their theoretical

model development was for international 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

36

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.

6.2 Managerial Implications

Organizations need to develop resilience capabilities in order to mitigate supply chain

vulnerabilities (Wagner and Neshat 2012). 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

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

37

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.

7. Conclusions

In this research we address the problem of mitigating vulnerabilities of the ready-made

garment industry of Bangladesh. Our basic methodology is QFD (Chan and Wu 2002,

Gonzalez et al. 2004), which we use to find the vulnerabilities and their resilience

strategies. We then develop an interactive multi-objective methodology to find the

satisfactory (as par 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 (Delice and Gunger 2011, Karsack 2004b). 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 vulnerabilities and resilience

strategies of its own. Data for study 3 has been collected independently of study 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

38

of materials (high dependence on imported materials), disruptions in utility supply,

increased competition (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, building relation with buyers and suppliers, quality control, skill and

efficiency development, ICT adoption, demand forecasting, responsiveness 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 two 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. However, 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.

39

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 (Chrstiano et al 2001).

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 (Lee et al

2009). Third, the concept of dynamics of vulnerability (Raharjo et al. 2011) 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.

40

REFERENCES

Akao, Y. 1990. Quality Function Deployment (QFD): Integrating customer requirementsinto Product Design. Cambridge, MA: Productivity Press.

Asia News Network, A. 2013. Garment sector in B'desh hit hard by instability. http://www.asianewsnet.net/Garment-sector-in-Bdesh-hit-hard-by-instability-45844.html (accessed 18/06/2013).

Babbie, E. 2007. The practice of social research. 11th ed. Belmont, CA, USA: Thomson Wadsworth.

Barratt, M., T.Y. Choi, and M. Li. 2011. Qualitative case studies in operations management: trends, research outcomes, and future research implications. Journal of Operations Management 29(4): 329-342.

Bevilacqua, M., F. E. Ciarapica, and G. Giacchetta. 2006. A fuzzy-QFD approach to supplier selection. Journal of Purchasing and Supply Management 12 (1): 14-27.

BGMEA. 2012. Bangladesh Apparel and Textiles Exposition.http://www.bgmea.com.bd/batexpo/index.htm (accessed 04/04/2013).

Bhattacharya, A., B. Sarkar, and S. K. Mukherjee. 2005. Integrating AHP with QFD for robot selection under requirement perspective. International journal of production research 43 (17): 3671-3685.

Blackhurst, J. V., K. P. Scheibe, and D. J. Johnson. 2008. Supplier risk assessment and monitoring for the automotive industry. International Journal of Physical Distribution & Logistics Management 38 (2): 143 - 165.

Blos, M. F., M. Quaddus, H. M. Wee, and K. Watanabe. 2009. Supply chain risk management (SCRM): a case study on the automotive and electronic industries in Brazil. Supply Chain Management:An International Journal 14 (4): 247 - 252.

Bode, C., Wagner, S. M., Petersen, K. J., & Ellram, L. M. (2011). Understanding responses to supply chain disruptions: Insights from information processing and resource dependence perspectives. Academy of Management Journal, 54(4), 833-856.

Bottani, E., and A. Rizzi. 2006. Strategic management of logistics service: A fuzzy QFDapproach. International Journal of Production Economics 103 (2): 585-599.

Braunscheidel, M. J., and N.C. Suresh. 2009. The organizational antecedents of a firm’s supply chain agility for risk mitigation and response. Journal of Operations Management 27(2): 119-140.

Bryman, A. 2006. Integrating Quantitative and Qualitative Research: How is it done? .Qualitative Research 6 (1): 97-113.

Büyüközkan, G., and C. Berkol. 2011. Designing a sustainable supply chain using an integrated analytic network process and goal programming approach in quality function deployment. Expert Systems with Applications 38(11):13731-13748.

Calder, B.J., Phillips, L.W and Tybout, A. M (1982). The Concept of External Validity, Journal of Consumer Research, 9(3): 240-244.

Carnevalli, J. A., and P.C. Miguel.2008. Review, analysis and classification of the literature on QFD -Types of research, difficulties and benefits. International Journal of Production Economics 114(2): 737-754.

Carpenter, S., B. Walker, J. M. Anderies, and N. Abel. 2001. From metaphor to measurement: resilience of what to what? Ecosystems 4 (8): 765 - 781.

Chan, L.K., and M.L. Wu. 2002. Quality function deployment: A literature review.European Journal of Operational Research 143 (3): 463-497.

41

Chan, L. K., and M. L. Wu. 1998. Prioritizing the technical measures in Quality Function Deployment. Quality Engineering 10 (3): 467-479 .

Chen, L. H., and M.C. Weng. 2006. An evaluation approach to engineering design in QFD processes using fuzzy goal programming models. European Journal ofOperational Research 172(1): 230-248.

Choi, Y., Ye, X., Zhao, L. and Luo, A. C. (2015), Optimizing enterprise risk management: a literature review and critical analysis of the work of Wu and Olson, Annals of Opr Res, 1-20.

Choudhury, S. R., and M. G. Hossain. 2005. Post-MFA Issues and Challenges: SocialDimension. Dhaka, Bangladesh: Ministry of Commerce.

Chowdhury, G. H., M. A. R. Sarker, and R. Afroze. 2012. recent unrest in the RMG sector of Bangladesh: is this an outcome of poor labour practices? International Journal of Business and management 7 (3): 206-218.

Chowdhury, M. H., Dewan, M. N. A., & Quaddus, M. A. (2012). Supply Chain Resilience To Mitigate Disruptions: A QFD Approach. In Pacific Asia Conference on Information Systems (PACIS), Ho Chi Minh city, Vietnam: AIS

Christopher, M., and H. Peck. 2004. Building the resilient supply chain. International Journal of Logistics Management 15 (2): 1-13.

Colicchia, C., F. Dallaria, and M. Melacini. 2010. Increasing supply chain resilience in aglobal sourcing context. Production Planning & Control 21 (7): 680-694.

Craighead, C. W., J. Blackhurst, M. J. Rungtusanatham, and R. B. Handfield. 2007. TheSeverity of Supply Chain Disruptions: Design Characteristics and MitigationCapabilities. Decision Sciences 38 (1): 131-156.

Creswell, J. W., and V. L. P. Clark. 2007. Designing and conducting mixed methods research. Thousand Oaks, CA: Sage publications.

Cristiano, J. J., White III, C. C., & Liker, J. K. (2001). Application of multiattribute decision analysis to quality function deployment for target setting.Systems, Man, and Cybernetics, Part C: Applications and Reviews, IEEE Transactions on, 31(3), 366-382.

Crowe, T. J., and C.-C. Cheng. 1996. Using quality function deployment in manufacturing strategic planning. International Journal of Operations & Production Management 16 (4): 35-48.

Deb, K. 2014. Multi-objective optimization. In Search Methodologies: IntroductoryTutorials in Optimization and Decision Support Techniques, E. K. Burke and G.Kendall (Eds.):403-449. Springer New York.

Delice, E. K., and Z. Günger. 2011. A mixed integer goal programming model for discrete values of design requirements in QFD. International journal of production research 49(10): 2941-2957.

Dickinson, G. (2001). Enterprise risk management: Its origins and conceptual foundation. Geneva Papers on Risk and Insurance- Issues and Practice, 26(3): 360-366.

Duclos, L. K., R. J. Vokurka, and R. R. Lummus. 2005. Delphi study on supply chain flexibility. International Journal of Production Research 43 (13): 2687-2708.

Faisal, M. N. 2013. Managing risk in small and medium enterprises (SMEs) supply chains' using quality function deployment approach. International journal of Operations Research and Information Systems 4 (1): 64-83.

Fennell, M. L., & Alexander, J. A. (1987). Organizational boundary spanning in institutionalized environments. Academy of Management Journal, 30(3), 456-476.

42

Ferdousi, F., and A. Ahmed. 2009. An Investigation of Manufacturing Performance Improvement through Lean Production: A Study on Bangladeshi Garment Firms. International Journal of Business and Management 4 (9): 106-116.

Fibre2fashion News Desk. 2013. US suspends GSP facility to Bangladesh. http://www.fibre2fashion.com/news/apparel-news/newsdetails.aspx?news_id=147897 (accessed 4-06-2013).

Fiksel, J. 2003. Designing resilient, sustainable systems. Environmental Science & Technology 37 (23): 5330 - 5339.

Goh, M., J.Y.S Lim, and F. Meng. 2007. A stochastic model for risk management in global supply chain networks, European Journal of Operations Research, 182: 164-173.

González, M. E., G. Quesada, F. Picado, and C. A. Eckelman. 2004. Customer satisfaction using QFD: an e-banking case. Managing Service Quality 14 (4): 317-330.

Haider, M. Z. 2007. Competitiveness of the Bangladesh Ready-made Garment Industry in Major International Markets. Asia-Pacific Trade and Investment Review 3 (1):3-27.

Han, S. B., S. K. Chen, M. Ebrahimpour, and M. S. Sodhi. 2001. A conceptual QFD planning model. The International Journal of Quality & Reliability Management 18 (8): 796 - 812.

Heckmann, I, Comes, T and Nickel, S. (2015). A critical review on supply chain risk– Definition, measure and modeling, Omega 52: 119–132.

Hendricks, K. B., and V. R. Singhal. 2003. The effect of supply chain glitches on shareholder wealth. Journal of Operations Management 21: 501–522.

Holling, C. S. 1973. Resilience and Stability of Ecological Systems. Annual Review of Ecology and Systematics 4 : 1-23.

Hossan, C. G., M. A. R. Sarker, and R. Afroze. 2012. Recent Unrest in the RMG Sector of Bangladesh: Is this an Outcome of Poor Labour Practices? International Journal of Business and Management 7 (3): p206.

Islam, M. A., M. N. Begum, and C. A. A. Rashed. 2012. 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 2 (2): 184-191.

Islam, M. A., and C. Deegan. 2008. Motivations for an organisation within a developing country to report social responsibility information: Evidence from Bangladesh. Accounting, Auditing & Accountability Journal 21 (6): 850 - 874.

Jiang, B., Baker, R. C., & Frazier, G. V. (2009). An analysis of job dissatisfaction and turnover to reduce global supply chain risk: Evidence from China. Journal of Operations Management, 27(2), 169-184.

Johnson, R. B., and A.J. Onwuegbuzie. 2004. Mixed methods research: A research paradigm whose time has come. Educational Researcher 33(7):14-26.

Jugulum, R., and M. Sefik. 1998. Building a robust manufacturing strategy. Computers & Industrial Engineering 35 (1): 225-228.

Jüttner, U., and S. Maklan. 2011. Supply chain resilience in the global financial crisis: an empirical study. Supply Chain Management: An International Journal 16(4): 246-259.

Kamvysi, K., Gotzamani, K., Andronikidis, A., and Georgiou, A. (21014), Capturing and prioritizing students’ requirements for course design by embedding Fuzzy-AHP and linear programming in QFD, European Journal of Operational Research 237: 1083–1094

43

Karim, S. 2003. Coping With Post-MFA Challenges: Strategic Response For BangladeshRMG Sector. Dhaka: CPD.

Karsak, E. E. 2004a. Fuzzy multiple objective decision making approach to prioritize design requirements in quality function deployment. International Journal of

Production Research 42(18): 3957-3974.Karsak, E. E. 2004b. Fuzzy multiple objective programming framework to prioritize

design requirements in quality function deployment. Computers & Industrial Engineering 47(2):149-163.

Karsak, E. E., and C.O. Özogul. 2009. An integrated decision making approach for ERP system selection. Expert Systems with Applications 36(1): 660-667.

Khondker , B. H., A. Razzak, and N. Ahmed. 2005. Exports, Employment and Working Conditions: Emerging Issues in the Post-MFA RMG Industry( Revised draft). International Labour Office.

Kleindorfer, P. R., and G. H. Saad. 2005. Managing Disruption Risks in Supply Chains.Production & Operations Management 14 (1): 53-68.

Köksalan, M. M., J. Wallenius, and S. Zionts. 2011. Multiple criteria decision making:from early history to the 21st century. World Scientific.

Larbani, M., and B. Aouni. 2011. A new approach for generating efficient solutions within the goal programming model. Journal of the Operational Research Society 62(1): 175-182.

Lee, D. E., Lim, T. K., & Arditi, D. (2009). Automated stochastic quality function deployment system for measuring the quality performance of design/build contractors. Automation in Construction, 18(3), 348-356.

Lee, A. H., H.Y. Kang, C. Y. Yang, and C. Y.Lin. 2010. An evaluation framework for product planning using FANP, QFD and multi-choice goalprogramming. International Journal of Production Research 48(13): 3977-3997.

Lin, C. M., Lee, Y. and Ho, T. (2011), Applying integrated DEA/AHP to evaluate theeconomic performance of local governments in China, European Journal of Operational Research 209: 129–140

Lockwood, J. L., and S. L. Pimm. 1994. Biological Diversity: Species: would any ofthem be missed? Current Biology 4 (5): 455-457.

Ludwig D, Walker B, and CS. Holling 1996. Sustainability, stability and resilience.Conserv. Ecol 1: 1-27.

Martin, S. 2004. The cost of restoration as a way of defining resilience: a viabilityapproach applied to a model of lake eutrophication. Ecology and Society 9 (2): 8.

Mitroff, I. L. and M. C. Alpaslan. 2003. Preparing for evil. Harvard Business Review 81(4): 109-115.

Mizgier, K. J., S. M. Wagner, and J. A. Holyst 2012. Modeling defaults of companies in multi-stage supply chain networks. International Journal of Production Economics 135(1): 14-23.

Narasimhan, R., and S. W. Kim. 2001. Information system utilization strategy for supply chain integration. Journal of Business Logistics 22 (2): 51-75.

Nuruzzaman, A. Haque, and A. Rafiq. 2010. Is Bangladeshi RMG Sector fit in the Global Apparel Business? Analyses the supply Chain Management. The South East Asian Journal of Management 4 (1): 53-72.

Olson, D L and Wu, D (2010a), Enterprise Risk Management Models, Springer, Heidelberg.

Olson, D L and Wu, D (2010b),A Review of enterprise risk management in supply chain, Kybernetes, Vol 39, No 5, pp. 694-706.

44

Onwuegbuzie, A. J., and N.L. Leech. 2005. On becoming a pragmatic researcher: Theimportance of combining quantitative and qualitative researchmethodologies. International Journal of Social Research Methodology 8(5): 375-387.

Park, T., and K.J. Kim. 1998. Determination of an optimal set of design requirements using house of quality. Journal of Operations Management 16 (5): 569-581.

Peck, H. (2006), Reconciling supply chain vulnerability, risk and supply chain management, International J. of Logistics Research and Applications, 9:2 127-142.

Pettit, T. J., K. L. Croxton, and J. Fiksel. 2013. Ensuring Supply Chain Resilience: Development and Implementation of an Assessment Tool. Journal of Business Logistics 34 (1): 46-76.

Pettit, T. J., J. Fiksel, and K. L. Croxton. 2010. Ensuring supply chain resilience: Development of a conceptual framework. Journal of Business Logistics 31 (1): 1-21.

Ponomarov, S. Y., and M. C. Holcomb. 2009. Understanding the concept of supply chain resilience. International Journal of Logistics Management 20 (1): 124-139.

Pujawan, I.N. and L. H. Geraldin. 2009. House of risk: a model for proactive supplychain risk management. Business Process Management Journal 15 (6): 953 - 967.

Raharjo, H., Xie, M., & Brombacher, A. C. (2011). A systematic methodology to deal with the dynamics of customer needs in Quality Function Deployment.Expert Systems with Applications, 38(4), 3653-3662.

Ramanathan, R and Yunfeng, J. (2009) Incorporating cost and environmental factors in quality function deployment using data envelopment analysis, Omega 37: 711-723.

Rose, A. 2004. Defining and measuring economic resilience to disaster. Disaster Prevention and Management 13 (4): 307-314.

Saaty, T. L. 1980. AHP: The Analytic Hierarchy Process. New York: McGraw-Hill. Sawik, T. (2015) On the fair optimization of cost and customer service level in a supply

chain under disruption risks, Omega 53: 58-66.Schmitt, A., Sun, S., Snyder, L., and Shen, Z. (2015) Centralization versus

decentralization: Risk pooling, risk diversification, and supply chain disruptions, Omega 52: 201-212.

Schwartz,B., Ben-Haim, Y., and Dacso, C. 2011. What makes a good decision? Robust satisficing as a normative standdard of rational decision making, Journal for the Theory of Social Behaviour, 41 (2): 209-227.

Sheffi, Y., and Rice. 2005. A supply chain view of the resilient enterprise. MIT Sloan Management Review 47 (1): 41-48.

Shin, W. S., & Ravindran, A. (1991). Interactive multiple objective optimization: Survey I—Continuous case. Computers & Operations Research, 18(1), 97-114.

Simchi-Levi, D., W. Schmidt, and Y. Wei. 2014. From superstorms to factory fires managing unpredictable supply-chain disruptions. Harvard Business Review 92(1-2): 96-101.

Simon, H. 1972, Theories of Bounded Rationality, In C. B. McGuire and R. Radner (eds.), Decisions and Organizations, North Holland.

Stratton, B. 1989. The refined focus of automotive quality. Quality Progress 22 (10): 47-50.

Tang, C. S. 2006. Robust strategies for mitigating supply chain disruptions. International Journal of Logistics Research and Applications 9 (1): 33-45.

45

Tashakkori, A., and C. Teddlie, C. (Eds.). 2010. Sage handbook of mixed methods in social & behavioral research. Sage.

Tewari, M. 2006. Is Price and Cost Competitiveness Enough for Apparel Firms to Gain Market Share in the World after Quotas? A Review. Global Economy Journal 6(4): 1-46.

Tomlin, B. 2006. On the Value of Mitigation and Contingency Strategies for Managing Supply Chain Disruption Risks. Management Science 52 (5): 639-657.

Vugrin, E. D., D. E. Warren, and M. A. Ehlen. 2011. A resilience assessment framework for infrastructure and economic systems: Quantitative and qualitative resilience analysis of petrochemical supply chains to a hurricane. Process Safety Progress 30 (3): 280-290.

Wagner, S. M., & Neshat, N. (2010). Assessing the vulnerability of supply chains using graph theory. International Journal of Production Economics, 126(1), 121-129.

Wagner, S. M., and N. Neshat. 2012. A comparison of supply chain vulnerability indices for different categories of firms. International Journal of Production Research 50 (11): 2877-2891.

Wang, H.F., and W.K. Hong. 2007. An integrated service strategy by QFD approach: a case of a telecom company in Taiwan. International Journal of Management and Decision Making 8 (2-4/2007): 251-267.

Wang, H., M. Xie, and T. N. Goh. 1998. A comparative study of the prioritization matrix method and the analytic hierarchy process technique in quality function deployment. Total Quality Management 9 (6): 421-430.

Wasserman, G. S. 1993. On how to prioritize design requirements during the QFD planning process. IIE Transactions 25(3): 59-65.

Wu, D. and Olson, D.L. (2009), On the special issue of risk issues in operations: methods and tools, Production Planning and Control, 20:4 293-294.

Wu, D D and Olson, D (2010a), Enterprise risk management: a DEA VaR approach in vendor selection, International Journal of Production research, Vol 48, No 16, pp. 4919-4932.

Wu, D D and Olson, D (2010b), Enterprise risk management: coping with model risk in alarge bank, Journal of the Operational Research Society, 61(2): 179-190.

Wu, D., Olson, D.L.and Birge, J. R (2011), Introduction to special issue on "Enterprise risk management in operations", Int. J. Production Economics, 134: 1-2.

Wu, D., Chen, S. H. and Olson, D.L.(2014), Business intelligence in risk management:Some recent progress, Information Sciences, 256: 1-7.

Wu, T., J. Blackhurst, and V. Chidambaram. 2006. A model for inbound supply risk analysis. Computers in Industry 57 (4): 350-365.

Xie, M., T. N. Goh, and H. Wang. 1998. A study of the sensitivity of ‘‘customer voice’’ in QFD analysis. International Journal of Industrial Engineering 5 (4): 301-307.

Zhou, M. 1998. Fuzzy logic and optimization models for implementing QFD.Computers & Industrial Engineering 35(1): 237-240.

Zsidisin, G. A., S. A. Melnyk, and G. L. Ragatz. 2005. An institutional theory perspective of business continuity planning for purchasing and supply management.International Journal of Production Research 43 (16): 3401 - 3420.

46

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

Bangladesh.

Khondker , Razzak, and Ahmed Deals with important issues and challenges facing the RMG(2005) industry during the post-MFA regime. It also discusses the

competitiveness issue in relation to productivity, workingenvironment and stakeholders.

Choudhury and Hossain (2005) The authors have taken step to define the challenges and opportunitiesin the post-MFA period in the RMG sector and indicated thegovernment response to combat post-MFA challenges.

Karim (2003) Discusses how the Bangladeshi RMG sector coped startegically withthe post-MFA challenges.

Tewari (2006) Emphasized on timely supply, short lead times, low inventories,innovation and the ability to contribute to design and full packagesupply.

Ferdousi and Ahmed (2009) Investigated the improvement of manufacturing performance throughlean practice that helps to reduce lead time and cost and to improveproductivity.

Hossan, Sarker, and Afroze (2012) Discuss the recent unrest in the RMG industry which indicates politicalaction in this industry.

Islam, Begum, and Rashed (2012) Operational disturbances, manufacturers are facing competition withrespect to quality, cost and time to market.

Ahmed (2009) Analysed the main drivers of growth, challenges faced andperformance of RMG manufacturing in Bangladesh in the post-MFAperiod.

Nuruzzaman, Haque, and Rafiq Described how to create competitive advantage through SCM.(2010)Chowdhury, Dewan and Addressed supply chain disruptions and mitigation processesQuaddus (2012)Chowdhury, Dewan, Dealt with upstream supply chain barriers and mitigationNuruzzaman and Quaddus processes(2012)Haider (2007) Presents the challenges and the surface-level and deep-level

competitiveness of the Bangladeshi RMG industry.

47

Table 2: Supply chain Vulnerability factors.

Vulnerability Specifc Vulnerabilities Referencesfactors

Natural disaster (flood, cyclone) Christopher and Peck (2004); Sheffi (2005);Kleindorfer and Saad (2005); Wu et al. (2006);Blackhurst et al. (2008).

Hazard Political instability Kleindorfer and Saad (2005); Wu et al. (2006);vulnerability Blackhurst et al. (2008); Blos et al. (2009).

Fire and other accidental damage Blos et al. (2009).Labour unrest Sheffi (2005); Kleindorfer and Saad (2005); Wu et

al. (2006); Blackhurst et al. (2008).Increased competition Haider (2007); Blos et al. (2009).Non-compliance of social and Islam and Deegan (2008).

environmental factors

StrategicProblem of relation with buyer Blos et al. (2009).

(switching of buyer)vulnerability

Problem of integration and real-time Gaudenzi and Borghesi (2006).information

Problem of relation with supplier Blos et al. (2009).Plant location problem Field studyCurrency fluctuation Blos et al. (2009); Blackhurstet et al. (2008).Economic recession Blos et al. (2009)

Financial Raw material price fluctuation Blos et al. (2009).vulnerability High bank interest & fund shortage Blos et al. (2009).

Bankruptcy or credit default of any Blos et al. (2009); Blackhurstet et al. (2008).supply chain member

Shortage of skilled worker Haider (2007).Switching and absenteeism of Chowdhury et al. (2012).

workersFault in production planning and Chowdhury et al. (2012); Wu et al (2006).

Operationalinventory management

Failure of IT system and Blos et al. (2009).vulnerability

machineriesDisruption in utility supply Blos et al. (2009).Product quality defection (poor Blos et al. (2009).

quality)Illiteracy of workers and supervisors Chowdhury et al. (2012)Delay in custom clearance Colicchia (2010).Delay for Congestion and Colicchia (2010); Blackhurst et al. (2008).

Infrastructure inefficiency in portvulnerability Strike by port workers Colicchia (2010); Blos et al. (2009).

Delay in transportation for Poor Blackhurst et al. (2008).infrastructure

Suppliers’ delay Blackhurst et al. (2008).Dependence on imported material Haider (2007); Nuruzzan (2009), Craighead et al.

and lack of backward linkage (2007).Lack of alternative for some critical Craighead et al. (2007).

Demand & itemssupply Defection or nonconformity of Blackhurst et al. (2008).

vulnerability materialOpportunism of buyers (expect Ponomarov and Hollcomb (2009).

discount)Demand fluctuation/uncertainaty Wu et al. (2006).Suppliers opportunism Ponomarov and Hollcomb (2009).

48

Table 3: Supply chain resilience capabilities.

CapabilitySpecific capabilities References

Factors

Flexibility in production (different volume of Duclos, Vokurka, and Lummus (2005);order, flexible production schedule) Braunscheidel and Suresh (2009).Ability to modify a wide varity of product as Braunscheidel and Suresh (2009).per buyer requirement (mix flexibility)Flexibility in contract with SC partners Duclos et al. (2005).(Partial order and payment, partial shipment)

Flexibility Efficient and effective logistics and supply Duclos et al. (2005).chain functions (e.g. sourcing, producing,distribution)Ability to respond to additional order or Juttner and Maklan (2011).sudden demandAbility to supply new and different products Braunscheidel and Suresh (2009).to different customer groups (mix flexibility)Alternative and reserve capacity (logistical Pettit et al. (2013).

Reserve/backup options)capacity Buffer stock Pettit et al. (2013).

Backup energy source Pettit et al. (2013).Integration Sharing information with supply chain Braunscheidel and Suresh (2009);

partners Blackhurst et al. (2005)Communication and information flow with Braunscheidel and Suresh (2009).different departments (e.g. supply chain andother departments)Joint or collaborative planning (e.g product Braunscheidel and Suresh (2009).development)Communication with supply chain partners Braunscheidel and Suresh (2009).ICT supported planning and integration Narasimhan and Kim (2001).

Efficiency Waste elimination (efficient use of resource) Pettit et al. (2013); (Fiksel 2003); Sheffi(2005).

Efficient and hardworking employeesQuality control and less defection Pettit et al. (2013); Kleindorfer and Saad

(2005).Buyer satisfaction (product quality and Pettit et al. (2013).

Customer service)satisfaction & Customer loyalty and Preference Pettit et al. (2013).market position Good relation with buyers and suppliers Zsidisin, Melnyk, and Ragatz (2005); Pettit

et al. (2013).

FinancialFund availability Pettit et al. (2013); Tang (2006).

Profitability Pettit et al. (2013).strength

Insurance Pettit et al. (2013).

49

Table 4: Profile of five companies and Participants

Participant Position Company type Product(s) Company Age of Productionsize (number company and salesof employees)

D1 Manager RMG Bottom item More than 10 years 60-80,000Merchandisi manufacturing (shorts, 5,000 dozens/yearng unit trousers)

D2 Deputy Accessory Cartoon and Less than 5 years 14-general supplying unit poly bags which 1000 15,00,000manager are used for cartoons

packing each yearproducts duringshipment.

D3 Manager RMG Main product is 2000-3000 10 years 25-30,000Merchandisi manufacturing shirt. dozens/yearng unit

D4 Supply RMG Main product is More than 25 years 40-50,000chain manufacturing shirt. 5,000 dozen/yearmanager unit

D5 Deputy Accessory Process raw Less than 15 years 40-50,00000General supplying unit cotton to yarn 1000 yard/ yearmanager and then spin

the yarn forknitting.

50

Table 5: Vulnerabilities and corresponding resilience capabilities

VulnerabilitySpecific vulnerabilities Enterprises AHP Adjusted

Weights weights

factors (WHATs)

1 2 3 4 5

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

Political instability and labour unrest (HV2) y y y y y .534 .071

Hazard (.132) Fire and other accidental damage (HV3) y y y .297. .039

Sabotage y y

Piracy and theft y

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

Failure to comply with social-environmental y y y y .383 .085

Strategic (.223)

factors (SV2)

Problem of relation with buyer & suppliers y y .172 .04(SV3)

Problem of integration y y

Currency & raw material price fluctuation y y y y .253 .039(FV1)

Financial (.154) 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

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

Disruption in utility supply (OV2) y y y y y .543 .106

Operational Product quality defection (poor quality) y y y .175 .034(.195) (OV3)

Machinery breakdown and failure y y

IT system failure

Production planning problem y y

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

Infrastructural Delay in transportation for Poor y y .422 .049(.115) infrastructure and port facilities (IV2)

Strike by port workers y

Suppliers’ disruptions (DSV1) y y y .242 .043

Lack of material (Dependence on imported y y y y y .597 .108

Demand &material and lack of backward linkage)

(DSV2)

supply (.181)Buyers’ disruptions (expect discount) y y y .161 .029(DSV3)

Unpredictability of demand y

Product differentiation and customization y y y(St1)

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 y y y y y

Resience

traing and counselling (St6)Product and Process Improvement for y y y y

Strategies

efficiency and waste reduction (St7)

(HOWs)

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 y y y y yissues (St11)

ICT adoption and information intergartion y y y y(St12)

Cooperation, communication and building y y y y yrelation with buyers and suppliers (St13)

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

51

Table 6: Aspiration levels of the objectives for study 1

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

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

110f2 .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= procurement strategy (f1); x5,x6,x7,x11,x12=processing strategy (f2);x1,x3,x8 and x10=distribution strategy (f3).

Table 7: Efficient Resilient Portfolios of study 1

Objective Function Procurement Processing strategies Distribution strategies Budstrategies get

f1 f2 f3 X2 X4 X9X

13 X5 X6 X7X

11X

12 X1 X3 X8X

10

Weighted(1,1,1)

.327 .329 .231 0 1 0 1 1 1 0 0 1 0 0 1 1

110Weighted(0.2, 0.5,0.3)

.327 .411 .149 0 1 0 1 1 1 1 0 1 0 0 1 0Note: 0.2 weight for f1 ; 0.5 weight for f2 ; 0.3 weight for f3 .

52

Table 8: Vulnerabilities and Strategies for Study 2

Vulenarabilities StrategiesHV1 = Sabotage St1= Security system improvementHV2 = Political instability St2= Product customizationHV3 = Factory fire St3=Alternative sources of supplyHV4 = Natural disaster (flood, cyclone) St4= Alternative transportation routingHV5 = Foreign government policy St5= Maintaining reserve capacity

change St6= Quality controlSV1= Competitive pressure St7= Skill development trainingSV2= Reputation loss St8= Improving process technologySV3= Problem of integration with St9= Demand forecasting

supply chain members St10= Backward linkage developmentSV4 = Failure to comply with socio- St11= Quick response to customers’

environmental issues requirementsFV1= Increasing raw material price St12= Improving social andFV2= Impact of Economic recession environmental performanceFV3= Cost of Finance St13= Information integrationFV4 = Exchnage rate fluctuation St14= Cooperation and collaborationOV1= Shortage of skilled worker with suppliers.OV2= Impact of Power crisisOV3= Concentrated production

locationOV4 = Machine breakdown & failureOV5 = IT system failureOV6 = Production planning problemIV1= Delay in portIV2= Poor transportation infrastructureIV3= Delay due to export import

document processingDSV1= Suppliers’ disruptionDSV2= High dependence on imported

materialsDSV3= Buyers’ disruption

53

Table 9: Aspiration levels of the objectives for study 2

Objectiv Aspiratio Procurement Processing strategy Distribution Budgee n level strategy (f1) (f2) strategy (f3) t

Function X XX

1X

1 X X X XX

1X

1 X X XX

1

3 5 0 4 1 6 7 8 2 3 2 4 9 1

f1 .163 0 1 0 1 0 0 0 0 0 0 0 0 0

130f2 .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 1Note: x3, x5, x10, x14 = procurement strategy (f1); x1, x6, x7, x8, x12, x13 =processing strategy (f2); x2, x4, x9, and x11 =distribution strategy (f3).

Table 10: Efficient Resilient Portfolios of study 2

Objective Function Procurement Processing strategies Distribution Budstrategies strategies get

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

130Weighted(0.25,0.45, 0.3)

.092 .151 .109 0 1 0 01

1 1 0 0 0 0 0 1 0Note: 0.25 weight for f1 ; 0.45 weight for f2 ; 0.3 weight for f3 .

54

Table 11: Vulnerabilities and Strategies for Study 3

VulnerabilitiesStudy 1 Study 2

StrategiesStudy 1 Study 2

HV1-Political unrest √ √ St1-Back up capacity √ √

√ √ St2-Risk managementHV2-Fire and other accident team

√ √ St3-Focusing on √ √HV3-Natural disaster sustainability practise

√ St4-Using updated √SV1-Reputation risk Technology

√ √ St5-Customer √ √relationship

SV2-Increased competition development

SV3-Selecting wrong production St6-Developing newsite buyers & markets

SV4-Problem of relation with √ √ St7-Flexibility inbuyers & suppliers production

St8-Strict quality √SV5-Lack of sustainability planning control at different& standard stage

St9-Training & √FV1-Increased production cost development

St10-Multiple √ √FV2-Lack of order suppliers

FV3-Loss due to rejection of St11-Suppliershipment selection & evaluation

√ √ St12-Developing √ √relationship with

OV1-Disruption of utility suppliers

St13-Product √ √OV2-Switching of workers differentiation

St14-Alternative √transportation (e.g. air

OV3-lead time failure 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 ofsuppliers (late delivery, quality )

DSV4-Supply shortage

DSV5-Buyers' opportunism

55

Table 12: Aspiration levels of the objectives for study 3

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

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

80f2 .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= procurement strategy (f1); x2,x3,x4,x7,x8, x9=processing strategy (f2); x5,x6,x13 and x14=distribution strategy (f3).

Table 13: Efficient Resilient Portfolios of study 3

Objective Function Procurement strategies Processing strategies Distribution Budstrategies get

f1 f2 f3 X1X

10X

11X

12 X2 X3 X4 X7 X8 X9 X5 X6X

13X

14

Weighted .30 .696 .5970 1 1 1 1 0 0 1 1 1 1 1 0 0(1,1,1)

Weighted .30 .696 .597 80(0.25,0.40, .35) 0 1 1 1 1 0 0 1 1 1 1 1 0 0

Note: 0.25 weight for f1 ; 0.40 weight for f2 ; 0.35 weight for f3 .

56

Figure 1: QFD Framework

Note: = Custome r requirements; = Degree of importance of CRi’s; = Design Requireme nts; = Relationship Matrix (i.e. degree to which CRi is met by DRj ) A.I.= Absolute importance of DRj’s ; R.I.= Relative im portance of DRj’s.

57

#,

#,#

* #,)

&,

*,#

',)

(,!

&,#

),!

(,( !,'

(,

&

),' &,*

SCVs Weights St1 St! St# St& St( St St) St* St' St10 St11 St1! St1# A,I

HV1 0,0!! 0 0,0 0,0 0,'* 0 0 0 0,0#) 0,0!! 0 0 0,0(1 0 1,!!!

HV! 0,0)1 0 0,1( 0,&') 0 0 0,0)1 0,0)1 0 0 0,&') 0,1( 0 1,&

HV# 0,0#' 0 0,0#' 0,11) 0,#(1 0 0 0 0 0 0 0,#(1 0,11) 0 0,')(

SV1 0,0'' 0,*'1 0,1( 0 0,0'' 0,'# 0,'# 0,*'1 0,0'' 0,!#1 0,'# 0,&'( 0,!#1 0,*'1 ,0)!

SV! 0,0*( 0 0 0 0,!(( 0 0,1'* 0 0 0,0*( 0,('( 0 0 1,1##

SV# 0,0& 0,0) 0,1! 0 0,1! 0,0& 0 0,! 0 0,1! 0,!* 0,0& 0,0& 0,# 1,#*)

FV1 0,0#' 0 0,11) 0 0,1'( 0 0 0 0,#(1 0,0#' 0 0 0 0,1'( 0,*')

FV! 0,0'# 0,!)' 0,!)' 0 0 0,!! 0,(1 0,*#) 0,*#) 0,1(( 0,!)' 0,0'# 0,!! 0,*#) &,*)

FV# 0,0!# 0 0,11 0 0,0!# 0 0 0 0,0' 0,0#* 0,0!# 0 0,#* 0,(& 1,!#&

OV1 0,0(( 0 0 0 0 0 0,&'( 0,0(( 0 0 0 0,1( 0 0 0,)1(

OV! 0,10 0 0 0 0,!&) 0,10 0,10 0,10 0,10 0 0,10 0 0,10 0,!&) 1,1#

OV# 0,0#& 0,0#& 0,0* 0 0 0,#0 0,!#* 0,#0 0 0,0) 0,10! 0,0() 0,0#& 0,10! 1,#!

IV1 0,0) 0 0,( 0,1* 0,11! 0 0 0 0,1* 0,1* 0,0) 0 0,0) 0,11! 1,&(*

IV! 0,0&' 0,11& 0 0,1&) 0 0 0 0 0 0 0 0 0 0,!1

DSV1 0,0&# 0 0,#*) 0,)! 0,1!' 0,0&# 0 0 0,1 0,!1( 0,1!' 0 0,1!' 0,#*) !,!#'

DSV! 0,10* 0,#!& 0,')! 0,!(! 0,)( 0 0,10* 0,1* 0,!(! 0,)( 0 0 0,10* 0,1* #,***

DSV# 0,0!' 0,0!' 0 0,0!' 0,0*) 0,!0# 0,!1 0,!0# 0,0!' 0,0*) 0,0* 0,1&( 0,0*) 0,!1 1,&*'

A,I 1,!& #,!!( 1,#& #,)&# 1,* !,((! #,0&) !,1#1 1,'1 1,*#! !,&#* 1,)#( &,11!

Cost (1, #,& !(,! !),# !!,) 1*,* #),! 1&,# )1,* !!,& (,) 1(, !1,

RE 0,0#1&)# 0,0**('' 0,0(&1!) 0,1#)10 0,0*!!0# 0,1#()&( 0,0*1'0' 0,1&'0!1 0,0!0! 0,0*1)* 0,0&!''* 0,111!1* 0,1'0#)

Figure 2: Supply Chain Resilience model: Study 1

Note: A.I= Absolute importance; Stj = Resilience strategy j; HV, SV, FV, OV, IV, DSV

= Various vulnerabilities, RE= Resilience efficiency.

58

SCVs Weights St1 St! St# St& St( St St) St* St' St10 St11 St1! St1# St1& AIHV1 0,0! 0,!) 0 0 0,1 0 0 0 0 0 0 0 0 0 0,#)HV! 0,0# 0 0 0,1&) 0 0,() 0 0 0 0,11' 0 0 0 0,11' 0,0*& 1,0#

HV# 0,0#& 0,!#* 0 0,0#' 0 0,!)# 0 0 0 0 0 0 0,#(1 0,0( 0,10! 1,0*SV1 0,0*! 0 0,*'1 0,1( 0 0 0,'# 0,*'1 0,'# 0,!') 0,!#1 0,*'1 0,!#1 0,!#1 0,!') (,(11SV! 0,0*' 0 0 0 0 0 0,!# 0 0 0 0 0,*01 0 0,&1 1,*#&

SV# 0,0) 0 0,0) 1,! 0,0'# 0,0& 0 0,1! 0 0,1! 0,!* 0 0,0& 0,# !,#!FV1 0,0&1 0 0 0,0'1 0 0,11) 0 0 0 0,#(1 0,0#' 0 0 0 0,11) 0,)1(FV! 0,0*1 0 0,&( 0,!)' 0 0 0,0'# 0,*#) 0,(1 0,*#) 0,0'# 0,!1) 0,0'# 0,0'# 0,*#) &,&'(

FV# 0,0!# 0 0 0,!0) 0 0,0(& 0 0 0 0,0' 0,0!# 0,0(& 0 0,0(& 0,0' 0,(#OV1 0,0&* 0 0 0 0 0 0 0,&'( 0,(( 0 0 0 0,0' 0 0 1,1#(OV! 0,10& 0 0 0 0 0,'(& 0 0 0 0,1)) 0 0 0 0 0,10 1,!#)

OV# 0,0## 0 0 0 0 0,!#1 0 0 0 0 0 0 0 0 0 0,!#1IV1 0,01 0 0 0,##( 0,&' 0,1* 0 0 0 0,1* 0,1( 0 0 0 0,1* 1,(IV! 0,0&( 0 0,0* 0 0,!&( 0 0 0 0 0 0 0 0 0 0,#!(IV# 0,0# 0 0 0 0 0,1* 0 0 0 0,0* 0,10* 0 0 0 0 0,#*

DSV1 0,0& 0 0 0,#*) 0 0,!1( 0 0 0 0,!1( 0,!&( 0,1!' 0 0,0)! 0,1!' 1,#'!DSV! 0,10 0 0 0,')! 0 0,(& 0 0,!(! 0 0,#!& 0,')! 0 0 0 0,(& #,DSV# 0,0!( 0 0,0)( 0 0,1)( 0,0(* 0,1)( 0,0(* 0,1!( 0,0*) 0,0* 0,!!( 0,1!( 0,0(* 0 1,!!'

AI 1,00) 0,(0* 1,&'* !,)0! 1,*&& #,*0) 1,!& !,(## !,1#' !,)# !,0(( 1,)' 1,'1 0,)#! #,!#1Cost 11,( *!,! &(,) #),& &1, &#, #,# )(,) !(,! 1!0,& &,* 10!,# ##, &(,&

RE 0,0&& 0,01*!!& 0,0('1!( 0,0&'#0( 0,0'1(1& 0,0#)!&* 0,0')* 0,0!*!( 0,10*()1 0,01)0* 0,0#*#) 0,01(# 0,0!1)* 0,0)11)

Figure 3: Supply Chain Resilience model: Study 2

59

1.5

22

52 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

Figure 4: Supply Chain Resilience model: Study 3

60

Highlights @4 th revision

• QFD methodology is used effectively to explicate supply chain vulnerability and resiliency.• Relevance of the paper with Enterprise Risk Management (ERM) is presented.• A new multi-objective optimization method is proposed based on QFD.• The method offers new ways to find portfolio of efficient resilient strategies.• The proposed method is applied to three large garment companies of Bangladesh• The results unearth some significant findings across the three companies.• Six new references have been added.

61