Modelling Integration and Responsiveness for Supply Chain
Ashish Agarwal!
School of Engineering and Technology,
Indira Gandhi National Open University
Maidan Garhi
New Delhi 110 068 (India)
E-mail: ashish_ka@yahoo.com
Ravi Shankar
School of Management
Asian Institute Of Technolog
P.O.BAG NO. 4,
Klong Luang, Pathumthani,
Thailand 12120
E-mail: r.s@rediffmail.com
Purnendu Mandal
Information Systems and Analysis Department,
College of Business, Lamar University,
Beaumont, TX 77710, USA
E-mail: Purnendu.mandal @lamar.edu
Word count: 6074 and abstract with 162 words
; Corresponding Author:
Ashish Agarwal,
School of Engineering and Technology, I.G.N.O.U., Maidan Garhi, New Delhi 110 068 (India)
Email: ashish_ka@yahoo.com
Phone and Fax: 91-11-29532863
Modelling Integration and Responsiveness for the Supply Chain
Abstract
A key feature of present day business is the fact that it is the supply chains that
compete, not companies and the success or failure of supply chains is ultimately
determined in the marketplace by the end consumer. Getting the right product, at
the right time to the consumer is not only the linchpin to competitive success, but
also the key to survival. Hence, customer satisfaction and market place
understanding are critical elements for consideration when attempting to establish a
new supply chain strategy. Based on the literature review, survey results, and
discussion with experts, causal relationships among supply chain performance
variables have been developed. On the basis of these causal relationships, a
framework has been modeled using system dynamics approach to capture the
dynamic impact of performance variables on the supply chain integration and
responsiveness for a period of eighteen months. This framework is useful in
analyzing the dynamic impact of different policies towards integration and
responsiveness of a supply chain.
Keywords: Supply chain performance, system dynamics, responsiveness, integration
INTRODUCTION
A supply chain is the set of structures and processes a company uses to deliver
product and services to a customer. It consists of:
(i) stock and flow structures for the acquisition of the inputs to the process, and
(ii) management policies governing various flows of material, information, and
ownership.
Managing the flow of material from source to an ultimate customer involves many
issues related to design, planning and control of supply chains. Efficient management of
these activities offers opportunities in terms of cost and lead-time reductions and
improves quality, the latter by employing a unanimous view on quality at the source
(Person and Olhager, 2002). Many supply chains are complex systems having high-order,
multiple loops, and non-linear feedback structures. Oscillation in order and inventory,
amplification in order and expected inventory, and lag in order and material flow are
aggregate behavior in these supply chains (Forrester, 1961). These behaviors are
understood to occur due to weak supply chain integration and poor responsiveness
(Towill, 1997; Christopher, 2000). To understand the effect of integration and
responsiveness on the performance of a supply chain, a system dynamics (SD) model has
been developed and presented in this paper.
In this paper, initial levels of certain performance variables of supply chain (SC) in
fast moving consumer goods were obtained with the help of experts’ opinion. After a
period of twelve months, levels of these variables were again obtained by taking opinion
from the same group of experts. These variables have causal relationships among them.
Therefore the level of these variables, after a period of twelve months, can be obtained
using system dynamics modeling. No significant difference is observed between
perceived and simulated level of SC performance variables, hence the SD model is
further used to analyze the variables responsible for SC performance improvement under
different scenarios. The objective behind developing the SD model is to derive the
learning insights (and not the quantitative estimate) from the causal behavior of variables
and its impact on supply chain performance. As suggested by Sterman (2000), these
insights are useful as an aid for policy formulation.
IDENTIFICATION OF SUPPLY CHAIN PERFORMANCE VARIABLES
The supply chain under study is a part of a supply chain involved in fast moving
consumer goods (FMCG) business. This supply chain consists of FACG Company,
suppliers, and dealers (Figure 1).
Supplier 1
Flow of material
+ ————
Flow of ownership
Dealer 1
jource
5 Customer
Dealer 2
Flow of information
+
Flow of money
Figure 1: Supply Chain System
The Annual turnover of the company in FMCG business is approximately Rs. 400
crores. The company has single production plant in India with approximately 400
employees. Number of zone-wise suppliers is approximately 50. It has a nationwide
network of approximately 800 dealers to facilitate the customers. Its top management is
committed for integration of its supply chain. Majority of its dealers are connected to its
extranet
These dealers provide real-time information about market conditions and demand.
Based on the sales forecast and dealer orders, production plans are formulated. These
plans percolate in the supply chain through communication to the suppliers to enable
them to plan their production in advance. Latest IT tools such as extranet and e-mails are
used to communicate these plans to the suppliers.
In the proposed system dynamics model, to identify supply chain performance
variables, and to obtain their initial and expected values, brainstorming sessions were
conducted with experts from the trading partners of the supply chain.
Initial meeting was held with the management of supply chain. In this meeting, 5
experts from the company and its trading partners had been identified. These five experts
had more than ten years of experience in the area of purchasing and supply chain
management. Literature related to responsiveness and integration of supply chain had
been circulated among the experts. Within a period of fifteen days, a brainstorming
session was organized to identify the variables. In all, twenty-six variables had been
identified in this session. The number was reduced to fifteen as some variables were of
same nature. The literature related to these fifteen variables had been circulated among
the experts. After seven days, a session was organized to establish the causal relationship
among the variables. Since the causal relationships among all variables could not be
identified in this s
ssion, so another meeting was conducted to complete this task. In this
meeting, experts were also asked how these variables could be benchmarked.
After five days, the list of variables and causal diagram were circulated among the
experts for any further modification. According to experts’ opinion causal diagram was
reconstructed.
The major purposes of the brainstorming sessions were:
(i) To develop a causal diagram of the whole supply chain system: Experts were
asked about the causal relationship among variables (how change in a
particular variable affects the other variables?). The integrated causal loop
diagram of the proposed model reflecting the interactions of the supply chain
performance variables is shown in Figure 2.
(ii) To guesstimate the initial value of enabling and resulting factors: A complex
system like this does not easily yield model parameter values; hence there was
a need to take experts’ opinion on the initial value of model parameters.
(iii) To capture experts’ expectation on plausible model behavior, which could be
used as reference mode for model validation: Experts were asked to project
(estimate) the desired value of different enabling and result variables at the
end of twelve months of operation of the supply chain system.
Fifteen supply chain performance variables are considered for developing system
dynamics model. These variables are categorized as enablers, results and inhibitors. The
seven enabling variables are market sensitiveness (MS), process integration (PI), delivery
speed (DS), centralized and collaborative planning (CCP), new product introduction
(NPI), data accuracy (DA), and use of IT tools (UIT). Out of these variables MS, DS, DA
and NPI are responsiveness enablers (RE) while PI, CCP, and UIT are integration
enablers (IE).
Five variables have been identified as indicator of outcomes or results. Results are
important for the improvement of supply chain performance. It is observed that lead-time
reduction (LTR), service level improvement (SLI), cost minimization (COM), customer
satisfaction (CUS), and quality improvement (QJ) are achievable through efficiently
managing the IE and RE enablers.
Inhibitor variables reduce the effect of enablers and results on supply chain
performance. The identified inhibitor variables are uncertainty in the market (UNC), lack
of trust among trading partners (LOT), and resistance to change among partners as well
as among employees (RTC).
Improvement in the performance level of enablers helps in achieving better level of
results. Inhibitors retard the rate of improvement of enablers and results. However,
continuous improvement in the levels of enablers reduces the rate of influence of
inhibitors on them. Overall improvement in enablers and results indicates the
improvement in performance of the supply chain.
CAUSAL RELATIONSHIPS AMONG SUPPLY CHAIN PERFORMANCE
VARIABLES
In a causal loop diagram (Figure 2), the arrows indicate the direction of influence
with the plus or minus sign depending upon the type of influence. All other things being
equal, if a change in one variable generates a change in the same direction in the second
variable relative to its prior value, the relationship between the two variables is referred
to as positive. If the change in the second variable takes place in the opposite direction
the relationship is negative (Forrester, 1961; Goodman, 1983; Mohapatra et al., 1994).
For the proposed SD-model, causal loop relationships among fifteen performance
variables are presented in Figure 2.
LOT
Figure 2: Causal Relationship among Supply Chain Performance Variables
In this figure, one of the responsiveness enablers for supply chain performance
improvement is market sensitiveness. Market sensitiveness means that the supply chain is
capable of judging and responding to the real demand (Christopher, 2000). Increase in the
market sensitiveness will improve the supply chain performance. Increase in this enabler
also makes an effect on results like service level improvement, lead-time reduction and
customer satisfaction. Present level of service level improvement is indicated as 92 units
while desired level is of 99 units. Therefore, there is a gap of 7 units in the service level
improvement. Improvements in the level of enablers will minimize the gap, which results
in the improvement in Supply-chain Performance Index (SP Index). SP Index, gap in lead
time reduction (GLTR), gap in cost minimization (GCOM), gap in service level
improvement (GSLI), gap in quality improvement (GQUJ), gap in customer satisfaction
(GCUS), actual lead time reduction (ALTR), actual cost minimization (ACOM), actual
service level improvement (ASLI), actual customer satisfaction (ACUS), actual quality
improvement (AQUI), desired lead time reduction (DLTR), desired cost minimization
(DCOM), desired service level improvement (DSLI), desired customer satisfaction
(DCUS), and desired quality improvement (DQUI) are the other variables considered in
this model. SP Index is an additive function.
An increase in market sensitiveness rate in the supply chain will increase new
product introduction (Power et al., 2001; Christopher, 2000; Jayaram ef al., 1999).
Customers are the final judge of how well the organization performs, and what they say
counts. It is their perception that will determine whether they remain loyal or seek better
providers. An increase in market sensitiveness will drive firms to listen to the customers
and act quickly on what they say. Dissatisfied customers must be heeded closely, for they
often deliver the most valuable information (Christopher, 2000). An increase in new
product introduction will induce the organization to set high desired customer
satisfaction goal (Jayaram et al., 1999) which in turn will increase gap in customer
satisfaction. Gap in customer satisfaction is the result of difference between desired
customer satisfaction and actual customer satisfaction. Gap in customer satisfaction will
also have a positive effect on new product introduction.
An increase in market sensitiveness rate will have a positive impact on use of IT
tools, which in turn will have positive relation with customer satisfaction, lead-time
reduction, quality improvement, service level improvement, and cost minimization.
Increase in the level of desired customer satisfaction is possible through increase in
desired service level, assuming same level of actual service level. Increase in gap in
service level improvement will have effect on gap in lead-time reduction, which in turn
attempts to accelerate the growth in SP index. To remain competitive, it is very important
that all trading partners collaborate with each other in the supply chain.
An increase in centralized and collaborative planning will tend to increase actual
quality improvement in supply chain (Handfield and Pannesi, 1992). Quality
improvement is very important for the success of any SC. Increase in actual quality
improvement will tend to increase SP index, which will have a positive effect on market
sensitiveness. Increase in actual quality improvement will reduce gap in quality
improvement, which in turn will have a positive effect on centralized and collaborative
planning.
Increase in use of IT tools helps to improve centralized and collaborative planning
among trading partners of a supply chain. Increase in the level of these enabler variables
reduces the level of uncertainty, lack of trust, and resistance to change and innovation
(Holmberge, 2000; Svensson, 2001; Agarwal and Shankar, 2003, Swaminathan and
Tayur, 2003). Decrease in the level of inhibitors accelerates the growth in the levels of
enabler, result and SP index.
10
MODEL VALIDATION
Validation of the SD model is necessary to establish sufficient confidence in a
model on some chosen criteria. The results of the proposed model are validated for the
SC, which incorporates a network of suppliers, company and distributors. The company
was doing well in 2003 but it was not able to fully meet the customer demand and gain its
market share. Its trading partners did not have significant level of trust among
themselves. Resistance to change was another barrier for the trading partners in
implementing any new policy to improve its business performance (Agarwal and
Shankar, 2002). The management of the company was unable to predict the influence of
uncertainty on the performance. The SD model approach appears to be useful for them to
capture and analyze the impact of variables on the SC performance. The average levels of
the variables to gauge the SC performance for the year 2003 and 2004 are judged by the
experts from the trading partners of the SC in fast moving consumer goods business.
In the proposed system dynamics model, initial level of SC performance variables
was obtained during a brainstorming session. In this session, experts were asked to judge
the present levels of enabler, result, and inhibitor variables for the supply chain on a scale
of 1 to 100. Experts were also asked to give their opinion regarding the desired level of
variables. Finalization of the levels of the variables could not be completed in one
meeting; so two more meetings were conducted. After getting the levels of variables,
flow diagram was developed for the proposed model.
The score for each variable was averaged and presented in Table 2. It has been
assumed that SP Index is an additive function of enabler and result variables and has a
11
maximum score of 1200. The term SP Index has been developed to indicate the level of
performance status of the supply chain.
The level of the SC performance variables after a period of three months was
obtained through SD modeling using ithink 7.0.2 software. Results obtained from the
simulation were neither shown nor discussed with the experts to avoid any biasness. After
a period of three months, all the experts were requested to make judgment related to
current levels of the variables. This exercise was carried out to find out the variation
between the level of the variables obtained through simulation and the level perceived by
the experts. This variation was found to be significant. The values of decision fractions
used in the SD modeling were then modified in consultation with experts. The simulated
values of levels of variables were compared with judgmental values after a period of five
months. After little modification in the values of decision fractions, the variation between
the levels of the variables obtained through simulation and the levels judged by the
experts after twelfth month was not found as significant. The insignificant variation in
these two values (Table 1) validates the proposed SD model.
MODEL RESULTS
The maximum levels of variables are obtained in brainstorming session for the SC,
which reflect business performance of the supply chain for the year 2003-2004. The
causal relationship between market sensitiveness and other performance variables is
presented as a closed loop in Figure 2. On the basis of causal relationship, a system
dynamics model is developed.
12
Table 1: Validation of Results Based on Experts’ Opinion and SD Model in the Twelfth
Month
Initial | Desired Score of variables in the twelfth month
Enablers Levels | Levels Avg. score | Results (based | % Deviation from
(a (Experts on SD) SD prediction
month) opinion)
Delivery speed 89 95 91 90.25 0.82
Data accuracy 95 98 96 95.10 0.94
Centralized and Collaborative 30 70 45 44.18
Planning 1.82
Market sensitiveness 60 70 70 69.5 0.71
New product introduction 65 85 69 67.41 2.30
Process integration 60 as 65 61.24 5.78
Use of IT tools 50 85 60 59.28 1.20
Total score of enablers 449 578 496 486.86 1.82
Results
Cost minimization 73 80 77 76.36 0.83
Customer satisfaction 92 95 93 93.46 -0.49
Lead time reduction 90 98 95 95.54 -0.57
Quality improvement 95 100 97 96.17 0.86
Service level improvement 92 99 95 94.99 0.01
Total score of results 442 472 457 456.52 0.11
Inhibitors
Lack of trust 29 10 10 20.74 -9.16
Uncertainty 20 10 10 13.62 2.71
Resistance to change 10 Ei 05 6.10 12.86
Total score of inhibitors 55 25 25 40.46 71.15
SP Index 891 1050 953 943.48 1.0
Here, for example, the SD equations for market sensitiveness are represented in the
following set. Here, “L” denotes level equations and “N’” denotes initial value.
L MS.L = MS.K + (DT) (RMS. KL)
N MS = 60
MS Market sensitiveness (Units)
RMS Market sensitiveness Rate (Units/Month)
The rate of market sensitiveness (RMS) during time interval “KL” depends on level
of market sensitiveness, gap in lead time reduction result, gap in customer satisfaction,
gap in service level improvement, gap in cost minimization and gap in quality
improvement. In the above set of SD equations, “R” denotes a rate variables equation.
13
It is assumed that the market sensitiveness rate depends entirely on the level of
market sensitiveness, gap in lead-time reduction, gap in customer satisfaction, and gap in
service level improvement. Similarly other equations have been developed for other
variables.
The base year for the model has been taken as year 2003 when scores are captured
for different variables for the supply chain. The simulation time period is 20 months. Rate
of improvement in the levels of enablers and results is high during first few months.
Later, this rate of improvement decreases. Level of inhibitors also gradually decreases.
The decrease in the level of inhibitors is due to improved process integration, centralized
and collaborative planning and effective use of IT tools. As the impact of inhibitors
reduces, and the influence of enablers and results increases, SP Index gradually improves.
Increase in SP Index indicates the performance improvement of the supply chain (Figure
3).
& «501 2: ENB 3: RST
a
2
iM
2
ae
1: 930
3: 455 ee *
oe
Zo
1: 890 4H
2: 445 [1
3: 440
1.00 5.75 10.50 1.25 20.00
Page 9 Months 6:34 PM. Tue, Apr 05, 2005
NI a BF ? Variation in SP Index with enabler and result variables
Figure 3: Variation in SP Index, Enabler (ENB), and Result (RST) Variables
14
MAPPING OF SUPPLY CHAIN
Performance of a supply chain depends on the growth in integrating enablers (IE)
and responsiveness enablers (RE) (Agarwal et al., 2005;Gunasekaran et al., 2004).
According to its ability to achieve growth in the levels of integrating enabler and
responsiveness enabler, supply chains can be put into four categories namely: laggard,
complacent, disillusioned and leader.
‘Laggard’ supply ‘Complacent’
chain supply chain
Low
t High [C D
3 ‘Disillusioned’ ‘Leader’ supply
fa supply chain chain
5 Moderate
& A B
5
a
3
et
s
2
g
5
Low Moderate High
Growth in Integrating Enabler —>
Figure 4: Integrating Enabler — Responsiveness Enabler Grid
These are shown in the integrating enabler— responsiveness enabler grid-graph in
Figure 4.
Type ‘A’: Laggard Supply Chain
The group of supply chains falling under this category is not able to achieve high
growth in the levels of integrating and responsiveness enablers, resulting into low growth
in results. The SP Index of such supply chains does not improve even after a long period.
15
Under severe competition when there is no protection from customers, the performance
of such supply chains crashes to the bottom. These supply chains hardly survive.
Type ‘B’: Complacent Supply Chain
The group of supply chains that fall in this category is capable of deploying strong
integrating enablers for survival and growth. Still these SCs are not able to achieve high
responsiveness enablers. The SC is also aware of need to deploy responsiveness enablers,
but their inclination is more towards waste reduction, quality improvement, and cost
minimization. Normally, such supply chains are able to survive only under protected
environment. Despite capabilities to better serve the customers under uncertainty, these
are complacent.
Type ‘C’: Disillusioned Supply Chain
This group of supply chains is capable in deploying strong responsiveness enablers
but is not able to deploy strong integrating enablers in supply chains. This appears to be a
very temporary but unsustainable phase.
The level of results is high due to high growth in the level of responsiveness
enablers. Normally, such supply chains are able to respond to customer demand, but
internally these are weak in a volatile market.
Type ‘D’: Leader Supply Chain
Such group of supply chains is capable in deploying strong responsiveness as well
as integrating enablers. Internal integration as well as supply chain capability related to
responsiveness is very high. The SC believes in the strategy to implement strong enablers
to achieve high results. It has high conviction that there is a direct relationship among
enablers, results, inhibitors, and SP Index. Management of the SC focuses on integrating
16
enablers like use of advanced IT tools for information sharing across the supply chain,
and also believes that this is only a first step towards improving performance. There is no
quick way to achieve the high performance level. Such supply chains have high top
management commitment. There is a total involvement of each and everyone in the
supply chain to achieve the performance maturity level to sustain the survival and
growth.
The model helps the companies to take policy decisions arising out of the dynamic
nature of the system. Therefore, policy experimentation and scenario analysis are
undertaken to see the effect of different scenarios. The results of these experimentations
and scenario analysis may help the companies to take timely action in unfavorable
scenario to avoid adverse effect on the performance. Policy experimentation and scenario
analysis have been attempted for deriving better insight so that timely actions can be
taken to bring supply chain on a right path of performance journey.
POLICY FORMULATION
Experimentations have been conducted for the sub-models depicted in Figure 5.
The first sub-model deals with situation when growth in the levels of integration, as well
as responsiveness enablers is low. In second sub-model, effect of high growth in the level
of integrating and responsiveness enablers has been analyzed. Third sub-model describes
the effect of low growth in the level of integrating enablers and high growth in
responsiveness enablers on supply chain performance. Forth sub-model analyzes the
effect of high growth in integrating enablers and low growth in responsiveness enablers.
17
Low growth in
integrating enablers and
High growth in
responsiveness enablers
high growth in | > | and high growth in
responsiveness enablers integrating enablers
Base-Model
Low growth in High growth in integrating
responsiveness enablers enablers and low growth in
and low growth in responsiveness enablers
integrating enablers |
Figure 5: Framework for Policy Experimentation
The objective of these models is to get insights for making policy formulation to
improve SP Index of the supply chain under consideration (Mohapatra et al., 2000).
Results obtained from SD modeling analysis are displayed in Table 2. Second column of
the table shows the desired percentage improvement in the level of SP Index compared to
the level at base month in year 2003. Third and forth columns indicate the percentage
improvement required in JE level and the percentage improvement required in RE level
respectively. Third and forth columns show the alternative policy to achieve the desired
target range of corresponding second column. For example, if management wants to
improve the SP Index level by 3% from its initial level, the proposed SD model provides
three alternative policies to bring the SP Index to the desired level (Table 2). Similarly,
many such combinations may be evolved.
18
(a) By Improving the Levels of IE by 3.9% and RE by 5% from their Initial Levels
Incremental change in the levels of CCP, PJ, and UIT have been made to improve
the level of JE by 3.9 % while the levels of performance variables like MS, DS, DA and
NPI have been changed to improve the level of RE by 5% from its initial level.
Simulation results indicate that with 3.9% growth in IE and 5% growth in RE, desired
improvement in the level of SP Index up to 3% could be achieved.
(b) By Improving the Levels of IE and RE by 7.7% and 3.1% from their Initial Levels,
respectively
Levels of variables like CCP, PI and UIT have been incrementally changed to
improve the level of JE by 7.7%. Similarly, incremental change in the levels of MS, DS,
DA and NPI has been made to improve the level of RE by 3.1%. With these growths in IE
and RE, desired improvement in the level of SP Index (up to 3% from its initial level) can
be achieved.
Similarly under alternative policy (c), desired level of SP Index is achieved by
making incremental changes in the levels of JE and RE variables.
(c) By Improving the Levels of IE and RE by 11.7% and 1.2% Times from their Initial
Levels, respectively.
The policy formulation helps to derive insight that desired percentage improvement
in SP Index (up to 3%) can be achieved with higher growth in the level of IE (11.7%) and
comparatively lower growth in the level of RE (1.2%). The desired percentage
improvement up to 3% can also be achieved with lower growth in the level of IE (7.7%)
and higher growth in the level of RE (3.1%). Therefore, relative influence of change in
the level of IE and RE on SP Index can be analyzed for making policy for the supply
chain under consideration.
19
Table 2: Alternative Policy for Improving the Level of SP Index
S.N Range of SP Index Alternate Policy for Improving SP Index
Improvement Level (L) IE Improvement level RE Improvement level
(%) (%) (%)
1 3.0<L<0.0 3.9 5.0
Ta eh
1 1.2
2 6.0<L<3.0 11.8 10.2
15.7 8.2
19.8 6.3
3 9.0<L<6.0 28.1 10.6
32.3 8.6
36.6 6.7
4 12.0<L<9.0 32:9) 10.7
36.6 8.7
40.9 6.7
strategic policy decision for its performance improvement. Using SD modeling result,
management can analyze the impact of varying growth in responsiveness level and
integration level. The simulation results for the considered model indicate that influence
of growth in integration level is more significant on SP Index than that of responsiveness.
SENSITIVITY ANALYSIS
improvement in SP Index of the supply chain. The effect of change in the levels of RE
and IE on SP Index can be captured through the SD model by varying the growth in the
level of one type of enabler with negligible growth in other type of enabler. Simulation
results from the SD model show that the increase in the level of SP Index is less sensitive
to the changes in level of RE (Figure 6 and Table 3).
20
Results provide learning insights to the management of the supply chain to take
The proposed SD model provides an aid to management in policy formulation for
—e— SP Index
95
9.0 ae
85
8.0
75 a
7.0
65 er
6.0
% Growth in SP Index
0.3 1.4 24 3.5 45 5.6 6.6 7.7
% Grow th in RE
Figure 6: Growth in the Level of SP Index with Incremental Growth in the Level of
Responsiveness Enabler (RE) and Constant Growth in the Level of Integrating Enabler (IE)
Table 3: Growth in the SP Index with Incremental Growth in the Responsiveness Enabler
(RE) at Constant Growth in the Integrating Enabler (IE)
S.N. Incremental Constant Growth in | SP Index % Remark
Growth in the the Level of IE Increase
Level of RE in SP
RE % IE % Index
Increase Increase
in RE in JE
Base Model
1 309.00 - 140.00 - 851.00 -
With incremental
2 | 310.00 0.32 147.62 5.44 907.38 6.63 growth in the
level of RE and
3 31322 1.37 147.63 5.45 910.63 7.01
4 | 316.45 2.41 147.64 5.46 913.89 7.39 _| significant
5_| 319.70 | 3.46 | 147.65 | 5.46 o17.16 | aan | Sewnupn3
6 322.95 4.51 147.66 5.47 920.45 8.16 b e :
een observed.
7_| 326.23 | 5.58 | 147.67 | __5.48 923.15 55 | Crowh mite
level of RE can
8 | 329.51 | 6.64 | 147.69 | 5.49 927.06 8.94 | beonly achieved
up to 7.7 % at
5.4-5.5 %
9 | 332.81 TN 147.70 5.50 930.38 9.33 constant growth
in IE.
21
It has been observed from Figure 7 that increase in SP Index level is sensitive to the
changes in level of IE (Table 4).
—@— SP Index
15.0
14.0
13.0
12.0
11.0
10.0
9.0
8.0
7.0
% Growth in SP Index
6.0
% Growth in IE
Figure 7: Growth in the Level of SP Index with Incremental Growth in the Level of
Integrating Enabler (IE) and Constant Growth in the Level of Responsiveness Enabler
(RE)
Table 4: Growth in the SP Index with Incremental Growth in the Integrating Enabler (IE)
at Constant Growth in the Responsiveness Enabler (RE)
S.N. | Incremental Growth Constant Growth in SP Index % Increase Remark
in the Level of IE the Level of RE in SP Index
TE % Increase RE % Increase
in IE in RE
Base Model
1 140.00 : 309.00 : 851.00 :
With incremental
2 141.52 1.09 329.32 6.58 920.61 8.18 growth in the
level of TE and
3 | 147.69 5.49 329.51 6.64 927.06 8.94 constant growth
in the level of RE
4 | 153.97 9.98 329.71 6.70 933.63 9.71 significant
growth up to 13.8
5 160.37 14.55 329.90 6.76 940.32 10.50 % in the level of
6 | 166.90 19.21 330.10 6.83 947.14 Lizio, _| SP Indexes
been observed.
7_| 173.55 | 23.96 | 330.30 6.89 954.08 ci | Gowen the
level of TE can be
2 >: achieved up to
8 180.3: 28.80 330.50 6.96 961.15 12.94 33.7 % at 6.6-7.0
9 187.21 3372 330.70 7.02 968.35 13.79 % constant
growth in RE.
22
LIMITATIONS OF THE MODELS
SD models are developed to derive insights related to implications of different
scenarios and policies (Sterman, 2000). The numerical values are only indicators of the
situation. Proper benchmarking and scaling have not been carried out in this research.
The model is specific to a particular supply chain. The values are captured after
discussing with managers. However, there may still be gap in making them understand
the model, scale of measurement etc.
However, since the insights are quite generic and carries a lot of meaning for
practicing managers, this research has been attempted in this paper.
DISCUSSION
One of the main purposes of system dynamics is to model the ways in which its
information, action and consequences components interact to generate dynamic behavior.
The results of the system dynamics model show that the supply chain performance index
improves as the level of integrating and responsiveness enablers improves. The
advantages of using simulation are to test proposed SD model for different levels of RE
and JE under different scenarios. The causes of faulty behavior can be easily diagnosed
and feedback loops can be tuned to obtain better behavior. The management group can
derive insights from the simulation results and take policy decisions for supply chain
performance improvement.
From Figure 3, it has been observed that the levels of SP Index and enablers for the
supply chain initially improve during first and sixth month. This improvement is a result
of management efforts towards supply chain integration. In integrated supply chain, all
23
trading partners work with common goal and interact with each other. Frequent
interaction among them reduces the influence of inhibitors like lack of trust, resistance to
change and innovation, and uncertainty. Trading partners are involved at planning stage
of product development. Clear picture of customer demand is available to all trading
partner; therefore they act on the same data set. These, in turn, enhance the
responsiveness of the supply chain. Improvement in the level of enabler results into
reduction in lead-time, improvement in service level, cost minimization, quality
improvement, and customer satisfaction. During ninth and thirteenth month, rate of
improvement in enabler and result is slower than the earlier months, which may be due to
arrival of new competitors and enhanced desired value of customer satisfaction.
Table 2 displays four desired ranges of SP Index for which management of the
supply chain wants to formulate policy. From the results, it has been observed that the
desired range of SP Index could be achieved by varying the growth in responsiveness
enabler and in the integrating enabler. Figures 6 and 7 illustrate that change in the rate of
SP Index is slightly more sensitive towards growth in integration level as compared to
growth in responsiveness level of the SC.
CONCLUSIONS
A system dynamics model has been developed to understand the behavior of RE
and IE on the performance improvement of a supply chain involved in fast moving
consumer goods business. Supply chain performance Index (SP Index) is a measure of
performance level of the supply chain and is an additive function of certain variables. The
simulation has been carried out for a period of eighteen months to capture the influence
of RE and IE on SP Index. Result from SD model shows that though both responsiveness
24
and integration are important for improving SP Index of the supply chain but high growth
in integration level significantly influences the SP Index of the supply chain. High growth
in integration level can be achieved through strategic partnership, process integration,
centralized and collaborative planning and use of IT tools. System dynamics model
developed in this paper is based on causal relationship among performance variables for
the supply chain. Causal relationships among variables have been established in a
brainstorming session. These relationships are generic in nature and are well-established
statements in the literature on supply chain. According to Lalonde and Masters (1994), a
supply chain can only succeed if all the members of the supply chain have the same goal
and the same focus of serving customers. Establishing the same goal and the same focus
among supply chain members is a form of policy integration.
The primary purpose of developing a SD model is to develop an understanding that
provides insight into the system and helps solve important problem (Sterman, 2000: page
850). Insights are important in such models rather than numerical values of variables. The
model can predict the rate of improvement of different performance variables based on
their specific strengths and gaps. The SD model helps managers to get insight for
continuously monitoring the performance and taking corrective measures on problem
areas. Therefore, SD model presented in this paper helps management to understand the
implication of interdependence of variables on supply chain performance, which is the
goal of every member in the supply chain. The model recommends dedicated focus on
the improvement in the integration level of supply chain along with reinforcement in the
responsiveness capability for creating value for the ultimate customer.
25
REFERENCE
1,
10.
11.
Agarwal, A., and Shankar, R., 2002, Modeling integration and responsiveness on a
Supply Chain Performance: A System dynamics approach, International Journal
System Dynamics and Policy-Making, XIV, | and 2, 61-83.
Agarwal, A., and Shankar, R., 2003, On-line Trust building in e-enabled supply
chain, Supply Chain Management: An International Journal, 8, 4, 324-334.
Agarwal, A., Shankar, R., and Tiwari, M.K., 2006, Modeling metrics of lean, agile
and leagile supply chain: An ANP-based approach, European Journal of
Operational Research, (Elsevier Press), Vol. 173, 211-225.
Agarwal, A., Shankar, R., and Mandal P.., 2007, Effectiveness of IT in Supply
Chain Performance: A System Dynamics Study, International Journal of
Information Systems and Change Management, Vol. 1, No. 3, 241-261.
Christopher, M., 2000. The Agile Supply Chain, Competing in Volatile Markets,
Industrial Marketing Management, 29, 37-44.
Forrester, J. W., 1961, Industrial Dynamics, MIT Press, Cambridge, MA.
Goodman, M. R., 1983, Study Notes in System Dynamics, MIT Press, Cambridge,
Massachusetts and London, England.
Gunasekaran, A., Patel, C., and. McGaughey, R., 2004, A framework for supply
chain performance measurement, International Journal of Production Economics,
87, 333 -347.
Handfield R.B., and Pannesi R.T., 1992, An empirical study of delivery speed and
reliability, International Journal of Operations and Production Management; 12, 2,
58-72.
Holmberge, S., 2000, A system perspective on supply chain measurement,
International Journal of Physical Distribution and Logistics Management, 30, 10,
847-868.
Jayaram, J., Vickery, S.K., and Droge, C., 1999, An empirical study of time based
competition in the North American automotive supplier industry, Jnternational
Journal of Operation and Production Management, 19, 10, 1010-1033.
26
12. LaLonde, B.J. and Masters, J.M., 1996, The 1996, Ohio state university survey of
career patterns in logistics, Proceedings of the Annual Conference of the Council of
Logistics Management, October 20-23, 115-138.
13. Mohapatra P.K.J., Mandal P., and Bora M.C., 1994, Introduction to System
Dynamics Modeling, University Press India.
14. Mohapatra P.K.J., Mandal P., Love P.E.D. and Smith J., Li H., 2000, Modelling
and Analysis of Rework in the Design Process of Construction Projects,
International Journal of System Dynamics and Policy Planning, 12, 2, 1-17.
15. Power D.J., Sohal A.S., and Rahman S., 2001, Critical success factors in agile
supply chain management: An empirical study, /nternational Journal of Physical
Distribution and Logistics, 31, 4, 247-265.
16. Sterman J.D., 2000. Business Dynamics: System Thinking and Modeling for a
complex world, Irwin McGraw-Hill.
17. Svensson G., 2001, Perceived trust towards suppliers and customers in supply
chains of the Swedish automotive industry, Jnternational Journal of Physical
Distribution and Logistics Management, 31, 9, 647-662.
18. Swaminathan, J.M and Tayur, S.R., 2003, Models for supply chain in E-business,
Management Science, 49, 10, 1387-1406.
19. Towill D.R. 1997, The seamless supply chain-the predator’s strategic advantage,
International Journal of Technology Management, Special issue on Strategic Cost
Management, 13, 1, 37-56.
Biography
Ashish Agarwal is Reader in Mechanical Engineering at School of Engineering & Technology, Indira
Gandhi National Open University, New Delhi. He has earned his doctoral degree from IIT Delhi, India. His
areas of interest are Cellular Manufacturing System, Agile Manufacturing System, Supply Chain
Management and System Dynamics Modeling. His research papers have appeared in European Journal of
Operational Research, Supply Chain Management: An International Journal, International Journal of
System Dynamics and Policy Making, and Work-Study, Journal of Advances in Management Research,
etc.
Email: ‘ish_ka@ yahoo.com
Ravi Shankar is Associate Professor at Department of Management Studies, IIT Delhi. His are:
interest are Industrial Engineering/ Operations Management, Supply Chain Management, e-Business,
Operations Research, and Fuzzy Modeling. He is a Ph.D. from I.LT Delhi, India. He has around twenty
years of teaching and research experience. He is also the group chair of Sectoral Management in LIT.
Delhi. His research papers have appeared in International Journal of Production Research, International
Journal of Production Economics, European Journal of Operational Research, Supply Chain Management:
of
27
An International Journal, International Journal of System Dynamics and Policy Making, Productivity,
Computers & Operations Research, Computers and Industrial Engineering, Journal of Advances in
Management Research, Journal of Enterprise Information Management, etc. Currently he is faculty at
Asian Institute of Technology, Thailand.
Email: ravil @dms.iitd-ernet.in
Purnendu Mandal is a Professor and the Chair person in the Information Systems and Analysis
Department, Lamar University, Beaumont, Texas. His teaching and research interests are in the areas of
database management systems, e-commerce, strategic management information systems and management
information systems, system dynamics. His research papers have appeared in European Journal of
Operational Research, International Journal of Production Economics, Management Decision, International
Journal of Operations and Production Management, International Journal of Quality and Reliability
Management, Logistics Information Management, Intelligent Automation and Soft Computing: An
International Journal, International Journal of Technology Management, ASCE Journal of Management in
Engineering, Decision, Applied Mathematical Modelling, etc.
E-mail: Purnendu.mandal @lamar.edu
28
