Understanding supply chain resilience

Full text

34
S C M R
•
J/F 
www.scmr.com
Understanding Supply Chain Resilience
By Steven A. Melnyk, David J. Closs, Stanley E. Griffis, Christopher W. Zobel, and John R. Macdonald
Steven A. Melnyk is a professor of Operations and Supply Chain Management in the Department of Supply Chain
Management, Michigan State University. He can be reached at melnyk@msu.edu. For more information, broad.msu.
edu/facultystaff/melnyk.
David J. Closs is the John H. McConnell Chaired Professor of Business Administration in the Department of Supply
Chain Management at Michigan State. He can be reached at closs@broad.msu.edu. For more information, visit
broad.msu.edu/facultystaff/closs.
Stanley E. Griffis is an Associate Professor in the Department of Supply Chain Management at
Michigan State University. He can be reached at griffis@broad.msu.edu. For more information,
visit http://broad.msu.edu/facultystaff/griffis.
Christopher W. Zobel is the R.B. Pamplin Professor of Business Information Technology at Virginia
Tech. He can be reached at czobel@vt.edu. For more information, visit www.czobel.bit.vt.edu.
John R. Macdonald is Assistant Professor of Supply Chain Management at Michigan State
University. He can be reached at johnmac@msu.edu.
Resilience is at the heart of current supply chain
management thinking. Understanding the concept,
and where to invest in resilience, can lead to supply
chains that quickly respond to and recover from
costly disruptions.
W
hen Boeing announced plans to
assemble the 787 Dreamliner
in late 2003, it introduced a
new concept to the assembly of
a commercial aircraft. Instead
of building the plane from the
ground up, subcontractors from around the globe
would deliver completed subassemblies to Boeing’s
factory in Everett, Wash. for  nal assembly. While
the approach was intended to create a leaner manu-
facturing process, development of the new aircraft
was beset by numerous supply chain related disrup-
tions—events that interrupt the  ow of products and
information between raw materials, production, and
the end customer.
One of those disruptions occurred just last January
TALENT VISIBILITY MANUFACTURING RESILIENCE COMPLIANCE

www.scmr.com
S C M R
•
J/F 
35
Understanding Supply Chain Resilience
2013, when the Dreamliner was grounded by the FAA
due to overheating of its new lithium-ion battery. As a
result, Boeing needed to slow production of this innova-
tive aircraft until it determined the source of the over-
heating—a source that appeared to lie within in its supply
chain, according to news reports. The question for Boeing
was how quickly it could identify the source of the over-
heating and recover from the disruption.
At bottom, that was a question of how resilient the
Dreamliner supply chain was. Boeing is not alone. In
today’s increasingly dynamic and turbulent world, one
where the supply chain plays an increasingly more impor-
tant role, numerous events occur each day that threaten
to disrupt operations and jeopardize the ability to perform
effectively and ef ciently. These events include natural
and man-made disasters such as equipment failures,  res,
labor disputes, supplier defaults, political instability, and
terrorist attacks. Each can have devastating effects on a
 rm. Such disruptions reinforce the insights that not only
can supply chain disruptions affect opera-
tions; they often result in  nancial dam-
age well beyond the immediate opera-
tional impacts.
One approach to dealing with
disruptions is the development
of supply chain systems that are
resilient. However, this notion
of resilience, which is at the
heart of so much of our current
thinking about supply chain risk
and management, is often not well-
de ned and sub-
ject to a great deal
of confusion.
While many con-
sultants, researchers,
and managers agree on
the importance of supply
chain resilience, there is less
agreement on what it is, how it
operates, and how and where to
invest to mitigate risk and recover
from disruptions—to shape and
in uence resiliency. This article
draws on the expertise of the authors,
prior research, anecdotes, and recent events
to de ne and further explore this concept.
Speci cally, we propose that resilience hap-
pens by design and not by accident. The resil-
ient supply chain requires two critical capacities : the
capacity for resistance and the capacity for recovery.
The  rst, resistance, de nes the supply chain’s ability to
delay a disruption and reduce the impact once the dis-
ruption occurs. The second, recovery, de nes the supply
chain’s ability to recover from a disruption.
The remainder of the article identi es and discusses
the tradeoffs between these two resilience capacities,
how each responds to issues of supply chain uncer-
tainty and risk, and investments that  rms can make to
enhance supply chain resilience capabilities. The con-
clusion: Resilience is a capability that must  t the spe-
ci c needs of each  rm.

36
S C M R
•
J/F 
www.scmr.com
Resilience
Supply Chain Resilience Defined
The concept of resilience traces its roots back to the work
of C.S. Holling, an ecologist who rst noted the charac-
teristics of a resilient ecological system in 1973. Since
then, the notion of resilience has been applied to elds as
diverse as psychology, systems thinking, disaster manage-
ment, and more recently, supply chain management.
For some, resilience is a reactive capability that
occurs after a disruption or shock has taken place.
Others see resilience as more proactive efforts toward
helping the rm prepare for a disruption. In light of
these divergent observations, it is not surprising that
there is confusion surrounding this key concept.
To the authors, supply chain resilience is “the ability
of a supply chain to both resist disruptions and recover
operational capability after disruptions occur.“ As men-
tioned above, viewed from this perspective, resilience
consists of two critical but complementary system com-
ponents: the capacity for resistance and the capacity for
recovery. Let’s look more closely at those elements:
• Resistance capacity is the ability of a system to min-
imize the impact of a disruption by evading it entirely
(avoidance) or by minimizing the time between disrup-
tion onset and the start of recovery from that disruption
(containment).
• Recovery capacity is the ability of a system to return
to functionality once a disruption has occurred. The pro-
cess of system recovery is characterized by a (hopefully
brief) stabilization phase after which a return to a steady
state of performance can be pursued. The nal achieved
steady-state performance may or may not reacquire orig-
inal performance levels, and is dependent on many dis-
ruption and competitor factors.
Exhibit 1 portrays the impact of a disruption over
time, from the moment that the disruption originates
somewhere in the system (at time TD) until the system
has returned to some form of steady-state (TR).
In this illustration, we can identify the four stages of
resilience, which are avoidance, containment, stabiliza-
tion, and return. Exhibit 1 also denes the sequence of
events, or time series signature, in a disruption as well
as the typical system response for a typical disruption.
Those would include inventory levels, cash ow, and
asset availability to name just a few.
Two variables are central to understanding this illus-
tration, T and R. T denotes the time at which a spe-
cic event occurs while R denotes the relative impact
of the event as measured in terms of dollars, units lost,
change in ll rate, or some other metric that is important
to a rm’s performance. Taken together, time (T) and
response (R) are important because they dene inec-
tion points in the time series signature where a change
in state can be observed.
The differences between the variance events listed in
Exhibit 1 identify traits of interest to management. For
example, TO-TD, or the gap between the moment at
which the disruption took place (TD) and the moment
that that disruption began to affect the rm (TO), tells
management how long it will take for the rm’s perfor-
mance to be impacted; this time interval also identies
the maximum amount of early warning that the rm can
count on to begin taking action to minimize the negative
effects of the disruption.
When supply chain disruptions and their traits are
observed, it is interesting to compare how the policies
and strategies used by the rm can affect the various
events identied in Exhibit 1 and Exhibit 2 in terms of
both time and impact.
Once recovery is complete, rms often reect upon
their experience to document appropriate lessons and
identify system renements to reduce future risks. This
completes a supply chain resilience cycle of: Avoidance
➝ Containment ➝ Stabilization ➝ Return ➝ Review ➝
Avoidance.
Resistance and Recovery
To illustrate the concepts of resistance and recovery,
consider the 2011 Japanese earthquake and subsequent
tsunami. In the wake of these twin disasters, it quick-
ly became apparent that suppliers for both Nissan and
Toyota facilities lacked adequate resistance capabilities
when faced with an event of this magnitude. Nissan,
however, exhibited signicant capacity for recovery. It
resumed operations and regained lost market share more
quickly than Toyota. Nissan was able to achieve this by
EXHIBIT 1
Time Series Display of Supply Chain Resilience Factors
Source: Michigan State University
S
yste
m
R
esponse
R(
t
)
Time
(
t
)
RR
R
RT
T
RC
C
TD
Avoidance
n
TO
TPT
C
TR
R
et
u
rn
Cont
tainment
t
m
n
S
Stabilization
S
b
a

www.scmr.com
S C M R
•
J/F 
37
accessing alternative suppliers, while Toyota stayed with
existing suppliers. Nissan’s supply chain thus provided a
differential advantage over that of Toyota, despite their
highly similar supply chain networks and locations rela-
tive to the earthquake/tsunami. Although full avoidance
of a supply chain disruption is an admirable goal, acci-
dents and disruptions will still occur. Instead, rms need
to develop the ability to deal directly with events that are
unavoidable.
Exhibit 3 below illustrates an alternate view of sup-
ply chain resilience, which characterizes resilience into
the capacities for resistance and recovery along with the
respective phases: avoidance, containment, stabilization,
and return.
While rms would clearly prefer to possess a high
capacity for both resistance and recovery, it is more likely
that rms will have a mix of these qualities. In particu-
lar, given resource constraints and competitive factors,
rms may need to choose where it is best for them to
invest limited resources. That is, the rm may not be
able to afford to invest in both improving resistance and
recovery. With this in mind, the resistance and recovery
matrix (Exhibit 4) characterizes possible positions that
a rm might nd itself in
with regard to varied levels
of these attributes.
Supply chains exhibit-
ing low capacities for both
resistance and recovery
would have low resistance:
They would experience
nearly every disruption
while also having slow and
weak recoveries as a result
of a lack of ability to recover
effectively. These supply
chains are “fragile.” Their
long-term prognosis is very
poor since they likely will
not last and won’t grow, unless protected by unique mar-
ket or regulatory conditions. For example, some indus-
tries in Sri Lanka over the past decade have suffered
multiple disruptions due to civil war, theft, power out-
ages, monsoon rains, and ooding. Firms in these indus-
tries have survived, however, because effective competi-
tion does not exist or because competition chooses not
to compete in such market or regulatory environments.
As a result, fragile supply chains that provide poor qual-
ity customer service persist because the customer base is
conditioned to accept low customer service.
In contrast to fragile supply chains with low resis-
tance, those that exhibit high levels of resistance are able
to alleviate potential risks more easily. When they also
possess the capacity for effective recovery, they quickly
rebound from those events that are unavoidable. Such
supply chains are classied as “hardy.”
General Motors (GM) is an example of a hardy sup-
ply chain. According to reports, GM constantly moni-
tors its supply chain to minimize disruptions and, when
necessary, to facilitate recovery. That was the case dur-
ing the Thailand oods of 2011. Despite having plants
C
ontainment
EXHIBIT 3
Tree of Supply Chain Resilience
Source: Michigan State University
R
esilience
R
esistance
Avoidance
R
eturn
R
ecovery
Stabilization
EXHIBIT 4
Resistance and Recovery Matrix
Source: Michigan State University
R
esistance
C
apacity
R
ecovery Capacity
H
i
g
h
L
o
w
Low
H
i
g
h
F
rag
i
le
Resistant
but Sluggish
Vulnerable
but Responsive
Hardy
Event Type Full Name Operational Denition
EXHIBIT 2
Description of Time Series Inection Points
Source: Michigan State University
E
vent
F
u
ll
N
ame
Specic time period in which the triggering event is initiated.
The time period in which the system being studied feels the
impact of the triggering event.
Time period in which the system reaches its climax.
The system response at the climax.
The time period in which the system begins to recover from
the disturbance.
The system response at the turning point; the response at which
the system transitions from being impacted by the disturbance
to recovering from the disturbance.
The time period in which the system returns to steady-state.
The system response level at the recovery period (may differ
from the pre-disturbance response level).
Time of Disturbance
Time of Onset
Time of Climax
Response at Climax
Turning Point
Response at Turning Point
Time of Recovery
Response at Recovery
TD
TO
TC
RC
TP
RT
TR
RR
T-Time
T
T
R-Response
T
R
T
R

38
S C M R
•
J/F 
www.scmr.com
Resilience
and suppliers in the area, GM experienced limited dis-
ruptions to the ow of materials because it was able to
resist the onset of problems better than its competitors.
When disruptions became unavoidable, GM was robust
enough to quickly work through them and recover.
Somewhere between fragile and hardy there exists
two middle positions. Supply chains that are character-
ized by an ability to adequately minimize disruptions, but
an insufcient ability to quickly recover, are ”resistant
but sluggish.” These supply chains exhibit high levels of
resistance, but if the system is ultimately disrupted, the
supply chain impacts are negative. These supply chains
are like a heavyweight boxer who is able to take signi-
cant attacks, but who is knocked down for a signicant
amount of time if pushed too far.
The use of the term “sluggish” in this case does not
imply ineptitude or lack of desire to restore operation,
but rather insufcient capability to do so. This may arise,
for example, from lack of recovery training as resources
are focused toward resistance instead. The chemical
industry is a case in point. Although these rms devel-
op relatively strong defenses
against a disruption, if a spill
or other event occurs it may
lead to serious consequences
that built-in recovery capabil-
ities might not be sufcient
to address quickly due to the nature of such spills.
The other middle position is characterized by sup-
ply chains that exhibit low resistance to disruptive
events, but quickly overcome their impact. These supply
chains are termed “vulnerable but responsive.” Similar to
an electrical fuse in a building, these are easily knocked
ofine, but they have the capacity to quickly recover.
An example of such a supply chain might be that of the
clothing manufacturer/retailer Zara. The fashion indus-
try is routinely beset by both supply and demand disrup-
tions, but Zara (as well as some other manufacturers) has
invested heavily in exible manufacturing so that it can
respond quickly to such changes. Recognizing that these
demand disruptions are the nature of its market, Zara
has invested in responsive systems to facilitate recovery.
While the “fragile” position is clearly undesirable and
the “hardy” position the brass ring, the existence of the
“middle” positions requires acknowledgement that rms
may reside there for two reasons.
• First, there may be limited resources with which to
invest in both capabilities.
• Second, there may be limited control over the envi-
ronment in which a supply chain operates.
The different manifestations of this lack of control in
a supply chain require rms to consider the notions of
supply chain resilience, risk, and uncertainty.
Supply Chain Resilience, Risk,
and Uncertainty
The distinctions between supply chain resilience, risk, and
uncertainty are often blurred and unclear. Unfortunately
this issue is exacerbated by the fact that some use risk and
uncertainty interchangeably, implying that these two con-
cepts are the same. Yet, this is not the case. While linked,
they are separate and distinct concepts.
Risk exists so rms have to deal with the possibili-
ties of encountering situations
that can adversely affect them.
However, not all future events are
equally unknown. Past experience
offers some insight regarding what
events could occur, the probabil-
ity of occurrence, and the impact.
Firms can predict the likelihood of
these events over a set time period
to help them determine how to potentially react when
they occur. Events with a greater likelihood and signi-
cant potential impact require greater preparation.
In contrast, uncertainty considers unpredictable
events. Typically, these are events that have not been
previously encountered. Alternatively, they are events
where the type of event falls outside of past experience.
To understand the differences, consider what happened
at the Fukushima Daiichi nuclear plant following the
Tohoku earthquake and tsunami.
This represented the largest nuclear disaster since
the meltdown of the reactor in Chernobyl in 1986. It
caused the evacuation of 100,000 people from their
homes. When 11 of Japan’s 50 nuclear reactors closed
immediately following the earthquake, the capacity to
produce electricity was reduced by some 40 percent. In
addition, key air and seaports shut down, affecting the
global supply of semiconductor equipment and materi-
als for consumer electronics, as well as parts sourced in
Japan for the wings, landing gear, and other major sys-
tems for Boeing’s 787 Dreamliner.
The full avoidance of a supply chain
disruption is an admirable goal. However,
accidents and disruptions will still occur. For that
reason, firms need to develop the ability to deal
directly with events that are unavoidable.

40
S C M R
•
J/F 
www.scmr.com
Resilience
Yet, in studying the events that took place at
Fukushima, one can see the interplay of risk and uncer-
tainty. When the plant was rst built in the 1960s, the
expected maximum height of a tsunami was 5 meters
(16.4 feet); the seawall built at the plant to resist this
potential risk event was 5.7 meters (19 feet). The tsu-
nami-generated wave that hit the plant was 13 to 15
meters (43 to 49 feet) in height. This event reects the
uncertainty that is always present. While plans were
made to resist a tsunami wave, the planners did not
foresee such a large tsunami
hitting the plant.
You could argue that what
Fukushima Daiichi needed
was a system that was ideally
hardy but at a minimum was
vulnerable but responsive.
Instead, what they had was
a system that was resistant
but sluggish. The notion of a
resistant but sluggish supply
chain and a vulnerable but
responsive supply chain may
also be considered in this
context of supply chain risk
and supply chain uncertainty.
Under conditions of uncer-
tainty, such as in the fashion
industry, the best approach
to building resilience may be
to invest in the capacity to
recover from an unpredictable
disruption. On the other hand,
faced with the known risk of
a chemical spill, the chemi-
cal industry’s policy of avoid-
ing such disruptions is more
appropriate, especially given
the extent of the damage that
would otherwise result.
By differentiating between
risk and uncertainty, we can
uncover an important rule of
thumb for resilience: When
faced primarily by risk, it makes
sense to invest in improving
resistance; when dealing with
uncertainty, it is more appropri-
ate to invest in improving recov-
ery capabilities.
Investing in Supply Chain Resilience
Resilience can be more properly regarded as a derived
system property. That is, it is the result of the invest-
ments a rm makes over time, not a ‘free’ benet of
existence. Moreover, it can be generated through many
different types of investments. These are summarized in
Exhibit 5.
Some of these investments, such as inventory and
capacity buffers, are direct investments. Investments in
safety stock or increased lead-times buffer the system
EXHIBIT 5
Eight Categories of Resilience Oriented Investments
Source: Michigan State University
Investing in the ability of the firm to identify
potential problems in the supply chain as
close to the event occurrence as possible.
Investments in improving the quantity,
speed, and quality of information flowing
within the supply chain.
Designing and implementing supply chains
that can be configured and reconfigured
quickly in response to environmental changes.
Protecting the system from supply chain
shocks in the form of theft, damage, or
counterfeiting.
Applying investments in other areas that can
be drawn on by the firm when a shock occurs.
Typically, these investments create goodwill
or a willingness to let the firm address its
supply chain problems.
• Improved Information Technology
or Information Sharing
• Early Warning by Supply Chain Partners
• Forecasting
• Demand Sensing
• Monitoring of Performance in the
Supply Chain
Examples of Investments
• Improved Information Te chnology
• Effective Communication
• Information Visibility
• Supply Base Management (Strategies
for Better Managing Suppliers at the
Major, Minor, and Scouting Levels)
• Supply Base Configuration
• Choosing Flexible Supply Chain Partners
• Human Resources – Capacity
• Human Resources – Capability/Experience
• Inventory
• Operating Flexibility
• Excess Operating Capacity
• Redundancy
• Excess/Safety Lead Time
• Firewalls
• Quarantine
• Strengthened Physical Systems
• Transportation Alternatives
• Variable Bills of Material
• Planning for Contingencies
• Training/Rehearsing
• Risk Assessment
• Insurance
• Marketing Position/Brand Equity
• Supply Chain Capital
• Relationships with Suppliers
• Relationships with Customers
• Supplier Loyalty
• Customer Loyalty
• Support for Innovation
• Support for Dynamic Partnering
• Revenue Management
Investment Strategies
Designing contingency plans for possible
supply chain shocks and testing of plans so
that the various groups know what they must
do and what their specific responsibilities are.
Changing either flows or product specifi-
cations in response to supply chain problems.
Creation of excess cushions in the form
of inventory, capacity, or lead times.
8
Indirect Investments
7
Preparedness
1
Discovery
2
Information
3
Supply Chain Design
4
Buffers
5
Operating Flexibility
6
Security
I
nvestment
S
trategies
S
ummarize
d

www.scmr.com
S C M R
•
J/F 
41
much like the shock absorbers on an automobile smooth
a bumpy road.
Indirect investments in areas such as brand equity and
customer loyalty can also have an impact on resilience.
While these investments are not focused directly on enhanc-
ing the resilience of supply chain systems, they offer capa-
bilities that the rm can draw on to deal with unexpected
breakdowns in its systems. As detailed in the sidebar, indirect
investments in brand equity and relationships with custom-
ers enabled Proctor & Gamble to recover from production
problems when it introduced its Tide Pods product.
Furthermore, these investments can be mapped to
specic stages within the four phases of resilience. The
challenge for the rm is that of determining the choices
between concern for supply chain risk or uncertainty and
determining which quadrant (as illustrated in Exhibit 4)
is both most appropriate and as representing the best
value for the rm’s investment investments.
Exhibit 6 illustrates that many of these investments
affect multiple stages of resilience. Note that these
investment values are qualitative approximations of
value; other values may be realized in various types of
supply chain situations. In reviewing this exhibit, also
note that the strength of the impact is indicated by the
greenness of the shading—a moderate impact is denoted
by + and yellow shading; a strong impact is denoted by
++ and green shading.
This exhibit provides example categories that can be
mapped to the four phases of resilience. It also suggests
how resilience investments affect the four phases of resil-
ience in differing ways. What it does not convey is the
nature of the impact—whether it is through main effects
(where an investment such as supply chain design affects
directly resilience) or through interactions (the interaction
between two or more factors found in the table).
No More Happy Accidents
While there is a great deal of con-
fusion about supply chain resil-
ience, it really comes down to
two separate but interrelated ele-
ments: resistance and recovery.
Further, where your rm chooses
to invest in building resilience is
really a function of whether you
are faced by uncertainty (in which
case you invest in recovery) or risk
(which justies the investment in
resistance).
Managers can make those
investments in supply chain resil-
ience through multiple venues in
ways that are both appropriate to
the risks a rm wants to mitigate
and that make sense to the parties
involved. The result is that resil-
ience is now becoming a supply
chain property that supply chain
managers can shape and inuence.
That happens by design and is no
longer a happy accident.
jjj
In August 2011, Proctor & Gamble announced the introduction of the Tide Pod. This
was an innovative detergent delivery system combining a detergent, stain remov-
er, and brightener into one easy-to-use pod. The product was intended to increase
demand in what had become a mature market. Unfortunately, P&G had to delay the
actual market entry date until early 2012 due to production challenges that limited
how much product would be available at retail outlets to support a broad product
launch. The breakdown gave P&G’s competitors in the home laundry market seg-
ment an advanced warning of P&G’s intent and a chance to seize market share in the
more profitable one-dose, convenience market segment.
Still, P&G was able to correct the original supply chain problems and recover
from these disruptions. By December 2012, P&G was projecting first year retail sales
totaling $500 million for the pods. Given that most new products are considered a
success if they achieve $50 million in sales, this turnaround is significant. Moreover,
because of production constrained product scarcity, P&G has offered no promo-
tions or discounts on the sales of this premium-priced product.
One reason for P&G’s ultimate success, despite their supply chain challenges,
can be found in the indirect investments that P&G has made in brand equity and
customer loyalty.
In other words, Tide brand loyal customers trusted the Tide brand, and despite
market entry by competitors’ all-in-one detergent products, P&G’s customers were
willing to wait until P&G resolved their supply chain problems and brought their
product to market.
P&G’s investment in resiliency
EXHIBIT 6
Alternative Investments in Resilience and Their Impact
on Avoidance, Containment, Stabilization, and Return
Strategies
Indirect Investment
Discovery
Information
Supply Chain Design
Buffers
Operating Flexibility
Security
Preparedness
Source: Michigan State University
Containment
Avoidance Stabilization Return
0
++
++
+
+
0
++
++
++
0
+
++
++
++
++
++
0
0
+
++
++
++
0
++
0
0
+
+
0
++
0
++

CopyrightofSupplyChainManagementReviewisthepropertyofPeerlessMediaandits
contentmaynotbecopiedoremailedtomultiplesitesorpostedtoalistservwithoutthe
copyrightholder'sexpresswrittenpermission.However,usersmayprint,download,oremail
articlesforindividualuse.