ISSN 2683-345X

journal homepage: http://ijiemjournal.uns.ac.rs/

International Journal of Industrial

Engineering and Management

Volume 12 / No 2 / June 2021 / 102 - 114

Original research article

Simulation Optimization of Manufacturing

Takt Time for a Leagile Supply Chain
with a De-coupling Point
L. N. Pattanaik
Department of Production Engineering, Birla Institute of Technology, Mesra, Ranchi, India

ABSTRACT ARTICLE INFO

Achieving agility with leanness in supply chains is considered to be a challenge for industry Article history:
and academia. In order to cope with dynamic demands at extreme downstream, buffer stocks
Received October 24, 2020
at various points on the supply chain can be seen as a solution. Although it may improve the
agility feature but extra inventories at warehouses affects the leanness of the supply chain Revised April 1, 2021
adversely. The aim of the present paper is to address this issue by finding an optimum rate Accepted April 5, 2021
of production at the factory which is directly related to Takt time concept of lean manu- Published online May 13, 2021
facturing in order to fulfill the dynamic demand patterns at the downstream retailers end
while minimizing intermediate stock inventories. The model is conceived as a leagile supply Keywords:
chain with a de-coupling point at the warehouse or distribution center between retailers and Simulation Optimization;
plant. A discrete event simulation model for the supply chain is developed in WITNESS® WITNESS®;
to experiment with various rates of production before finding the optimum value. The two Lean Manufacturing;
performance measures representing fulfillment of product demands and inventory carrying Leagile Supply Chain;
costs are expressed in equivalent cost units for optimization. Two demand scenarios for Takt Time;
a two product supply chain are simulated to identify the optimal rate of production while Decoupling point
illustrating the solution methodology. The simulation optimization approach to address this
problem of leagile supply chain is found to be effective and practical.
*Corresponding author:
Laxmi Narayan Pattanaik
[email protected]

1. Introduction the market. In order to retain the core competence

of manufacturing enterprise, its systems and supply
Supply chains often pose challenges in decision chains should be altered in response to the changing
making as it involves several entities and parameters requirement [1]. In the present dynamic world, in or-
like vendors, warehouses, logistics supports, variable der to be competitive, it is integral that the customer
demands and costs, etc. Uncertainties in events and demand be met unceasingly. The market demand is
forecasted demands add more complexities to this volatile because of lot of factors such as increasing
problem. In recent times, the demand for products demands for customization, advances in technology,
is highly unpredictable owing to the market dynam- seasonal variations, catastrophes, etc. It is, therefore,
ics and uncertainties. Further, the need of providing imperative that advancements in modeling of supply
a wide range of product variety to cater the custom- chain management systems rise up to meet the chal-
ized demands also aggravate the problem. In a global lenges [2-4].
manufacturing environment, more and more prod- In this context, the concepts of lean and agile sup-
ucts with shorter life cycles have been introduced to ply chain models are worth a mention. Supply chain

Published by the University of Novi Sad, Faculty of Technical Sciences, Novi Sad, Serbia. DOI: http://doi.org/10.24867/IJIEM-2021-2-280
This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions
Pattanaik 103

agility is a key to adapting to market variations more experiment with varying production rates under dy-
efficiently, inventory reduction, enabling firms to re- namic demand scenarios.
spond to demand more quickly and integrating with In section 2, literature pertaining to leagile supply
suppliers more effectively. chains and simulation based optimization of supply
Lean supply chains were primarily designed chains are compiled. The leagile supply chain mod-
for the removal of waste from all the related func- el along with various assumptions, parameters and
tions. Leanness means developing a value stream to constraints, objectives for optimization are described
eliminate all waste and to achieve a balanced pro- in section 3. In section 4, experimentation with the
duction schedule. However, lean supply chains are model developed in WITNESS® is illustrated using
considered to be attainable for relatively stable and hypothetical data sets for two stochastic demand pe-
predictable demand with low variety. On the other riods, various costs, capacity of warehouse and pro-
hand, agile supply chains provide with the solutions duction rates. The output from simulation model is
to problems where there is fluctuating demand with analyzed to arrive at optimal takt or production rate
high variety. These two paradigms of lean and agile for two different product types in each demand sce-
supply chains are merged to develop the conceptual nario. The concluding remarks and future scope of
model of leagile supply chain. The tradeoff between the work are presented in the last section.
leanness and agility was balanced in a leagile supply
chain system by identifying an appropriate de-cou-
pling point [5]. 2. Literature review
De-coupling point is defined as the point where
This literature review encompassed the relevant
the model changes from push to pull based system.
areas for the present work like agile supply chains,
The push model works on the principle of anticipa-
leanness in supply chain, leagile supply chains,
tion of customer orders while pull model is executed
de-coupling point concept, application of simulation
when customer demand is known with certainty. The
optimization and some case studies. Literature from
supply chain exhibit lean features before the decou-
these areas are presented here in the same order.
pling point from the upstream and agility after it to-
Over the years it has become apparent that mar-
wards downstream. The de-coupling point is located
kets are now increasingly volatile and less predictable.
such that it favors the need for responding to a fluc-
So the need for a more agile response has grown.
tuating demand in downstream while allowing a stat-
Agility is the company-wide capability that includes
ic level of manufacturing schedule in the upstream.
organizational structures, information systems, logis-
The decoupling point can be considered as the point
tics, procurement and production to respond to vol-
where order-driven and the forecast driven functions
atility. In such market conditions of increasing levels
merge [5-6].
of product variety and customization, the ability to re-
One of the key characteristic of the lean model
spond to customer orders in time can provide a crit-
presented in this paper is the takt. It is a preset pro-
ical competitive advantage. Xiaomei and co-authors
duction rhythm associated with lean philosophy and
[7] emphasized that the traditional supply chains fail
defined as the time interval between two consecutive
to cope up with the uncertainty in the market owing
finished products to ensure the continual flow of fin-
to development of economy, information technology
ished products needed to meet customer demand.
and shortened product life cycle. Preference of cus-
The reciprocal of production rate is mathematically
tomized and diversified products by end customer,
equivalent to the takt time. The benefits of adopting
uncertainties and disruptions necessitates an inherent
takt during production include balanced utilization of
flexibility within the supply chain network to ensure
resources, minimization of waste in finished invento-
the reconfigurability; a primary requirement when
ry, fulfillment of demand in schedule, etc.
developing an agile supply chain system [8-11].
The aim of the present work is to find the opti-
Lean manufacturing concepts have been in prac-
mum production rate or takt time of a manufactur-
tice over few decades and well accepted as an effec-
ing/assembly system associated with a leagile supply
tive tool. The roots of lean philosophy can be traced
chain having stochastic and dynamic demands at
to Toyota Production Systems (TPS) of Japan. The
multiple retailers. The objectives are to avoid excess
lean approach is applicable where there is relatively
stocks of inventory at warehouse (considered as the
stable and predictable demand with low variety [12].
de-coupling point) while meeting the demands of the
Leanness and agility, even being very different as
downstream retailers. A discrete-event simulation
concepts have been successfully merged within total
model is developed for the leagile supply chain to
supply chains by [5]. The combination of agility and
International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)
104 Pattanaik

leanness into supply chain with the strategic place- namic system as compared to other analysis tools,
ment of a de-coupling point is termed as leagility [13]. simulation has an edge due to the dynamicity and
The drawback of lean supply chain is the inability to randomness it can provide to the user. Simulation
respond to end customer customized demands, thus, has been used for years in the areas of supply chain,
leagile supply chain has been proposed in the indus- manufacturing and business has led to a wide range
try to combine the advantages of both agile and lean of successful applications in different areas such as
paradigms. Compared with traditional supply chains, design, planning and control, strategy making, re-
leagile supply chain has the advantages of informa- source allocation, training, etc. [23].
tion sharing, shorten length of chain, order guidance Simulation is the best practice to evaluate the sys-
and close cooperation among stake holders. tem performance closely to real situation. Simulation
Christopher and Towil [14] put forward the idea Optimization (SO) appears as popular technique and
to bring together the lean and agile philosophies to has received considerable attention from both sim-
highlight the difference in the two approaches and ulation researchers and practitioners which can be
suggested that these can be combined for better re- achieved using software packages [24]. Ran et al. [25]
sults and advantages. The leagile supply chain focus is presented a review on applications of SO to design
to effectively handle uncertain demands by deferring and operation of manufacturing systems to address
the products as far as possible towards the customer the inherent stochastic properties. By dividing the
end. Hoek [15] highlighted benefits of the postpone- problems into local and global optimization category,
ment strategy, like reduced inventory, increase in flex- they further classified on the basis of discrete or con-
ibility and multiplicity of production, easy forecasting tinuous nature. Maedeh et al. [26] solved a multi-ob-
and better personalization according to the customer jective problem using a hybrid of SO with regression
demand. The importance and advantages of leagile analysis for unreliable and unbalanced production
supply chain has been discussed by a number of re- lines. In a recent review, classified applications of
searchers [9] [16-17]. Ambe and Badenhorst-Weiss simulation optimization to supply chain problems in
[18] proposed a framework for leagile supply chain general with focus on resilience was found. Hybrid-
appropriate for the auto industry and the implemen- ization of SO with meta-heuristics was suggested as a
tation of which would result in cost reduction and prospective future research direction [27].
the supply chain being more responsive. Shukla and Demand uncertainty, in particular is an important
Wan [19] in their work presented an optimization factor to be considered in the supply chain design and
approach for a leagile inventory model. They first operations. Due to advance of global manufacturing,
formulated a non-linear integer programming model the decentralized optimization of multi-tier supply
which was solved in real-time using three variants of chains for multiple retailers and manufacturers be-
genetic algorithm. Komoto and other authors [20] in comes more and more important. Nishi and Yoshida
their paper on multi-objective reconfiguration meth- [28] have addressed the optimization of multi-period
od of supply chains through discrete event simulation bi-level supply chains under demand uncertainty. The
worked on a case study to show how the multi-objec- optimization algorithm to derive Stackelberg equilib-
tive optimization has been implemented in discrete rium for multi-period bi-level supply chain planning
event simulation. Peirleitner et al. [21] compared two problem is developed and dealt using simulation.
different solution methods for determining optimal Matheus et al. [29] proposed a simulation based op-
parameter settings for lot size Q and reorder points. timization approach to cope with supply chain plan-
The first method is an analytical optimization mod- ning and control of high uncertainty scenarios i.e.
el assuming a single-stage, single-product inventory stochastic behavior and dynamic events, addressing
system which is applied independently for all supply areas of material inventory, production and trans-
chain partners. Optimal parameters are identified portation. In their work, they discussed a simulation
for all partners and then re-evaluated in the dynamic based optimization approach to simultaneously deal
and stochastic simulation model. Results show that if with the planning and control of the material inven-
analytical optimal parameters are evaluated with sim- tory, production and transportation areas combining
ulation, which includes the dynamics and interdepen- the capabilities from metaheuristics and simulation
dencies between the supply chain members, lower models. Their proposed approach was implemented
service levels than initially predefined were achieved. in a test case and claimed a convergence to a solu-
A synchronized logistic model to address various is- tion within a short span of time. Liotta et al. [30] also
sues of a dynamic supply chain was developed [22]. opine that simulation-based optimization is a strategy
Considering supply chain as a complex and dy- for dealing with uncertainty in the supply chain. In

International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)

Pattanaik 105

addition, Truong and Azadivar [31] suggested that is achieved by satisfying the customers by adjusting
managing a supply chain is much more complex than the takt time and reorder points according to the de-
dealing with one facility because of existing conflict- mands.
ing objectives and dynamic properties of the system. Takt is a pre-determined production rhythm asso-
Hence, they proposed a simulation-based approach ciated with lean philosophy and defined as the time
to deal with supply chain configuration design. interval between two consecutive finished products
A case study of a mobile phone manufactur- needed to meet customer demand. It is expressed as
ing industry was undertaken. They studied mobile the ratio between total available time for production
phone firm’s operational configuration and pro- and total customer demands during that duration.
posed real-time decision support mechanism based The reciprocal of production rate is mathematically
on agent-based discrete-event simulation to estimate equivalent to the takt time. By adopting takt during
performance of average inventory levels over the production, balanced utilization of resources, mini-
system-wide supply chain [1]. Using WITNESS®, a mization of waste in finished inventory, fulfillment of
supply chain model can be built and analyzed with- demand in schedule, etc. can be achieved.
out the need to physically carry out tests in real life A leagile supply chain model with one manufac-
[32]. turer, one warehouse and six retailers (Figure 1) is
In this paper, a leagile supply chain model is sim- presented here to illustrate the proposed simulation
ulated using WITNESS® to give optimum takt times optimization approach. Some assumptions for the
for different demand periods with conflicting objec- model are summarized here.
tives of minimizing stock-outs and inventory costs.
• Required inventories for production are
supplied just-in-time and no shortages occur
3. Model of the leagile supply chain • The manufacturer and the supply chain
deals with two product varieties, product
The supply chain adopted in the present paper A and B
has leagile characteristics to meet volatile market de- • The retailers use an inventory replenishment
mand as well as to ensure optimum inventory level model with re-order point and re-order
to minimize cost. To achieve this, the decoupling quantity as parameters
point is set at the warehouse which is between the
• The six retailers are divided into three
production facility and the retailer(s). Products are
clusters based on their geographical
continuously manufactured at a predetermined rate
locations and distances from the warehouse
based on takt time and pushed to the warehouse after
which the goods are pulled by the retailer(s) accord- The leagile supply chain model has the following
ing to the customer demand. Leanness is achieved parameters:
by producing the optimum amount and avoiding ex-
pi Reorder point for retailer cluster i,
cess inventory and cost associated with it while agility
in units

Figure 1. Model of the leagile supply chain with six retailers in three clusters

International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)

106 Pattanaik

q Reorder quantity, in units the customer.

Ri Retailer cluster i In this paper a supply chain model is optimized
having two conflicting objectives to maximize the ser-
DA Demand for product type A, units/day
vice level and minimize the inventory level.
DB Demand for product type B, units/day The two objectives were combined to calculate
MPA Market price of product A per unit the total loss incurred due to high inventory at the
MPB Market price of product B per unit warehouse and losses due to poor service level at re-
CPA Cost price of product A per unit tailers. Losses due to high inventory are considered
only after the inventory level at the warehouse cross-
CPB Cost price of product B per unit
es a threshold level IT.
ti Minimum transportation time from
warehouse to retailer cluster i, days
αi Exponential variation in transportation
time from warehouse to retailer cluster i,
days
Ti Transportation times from warehouse to
retailer cluster i, Ti = (ti + αi), days
ηA The optimum production rate of product
type A at the plant, units/day
ηB The optimum production rate of product
type B at the plant, units/day
λA Production capacity of the plant for
product A, units/day The takt based production rate with the lowest
value of total loss TL gives the optimum production
λB Production capacity of the plant for product
rate for the respective demand cycle
B, units/day
Rc Retailer holding capacity, units
Wc Warehouse storage capacity, units 4. Simulation optimization using
ɣi Delivery capacity from warehouse to WITNESS®
retailer cluster i, units
Simulation optimization is considered as an effec-
TL Total loss incurred for a product, cost units
tive analytical tool to arrive at the optimal solution
TI Total loss due to excess inventory at
without implementing any classical, conventional or
warehouse for a product, cost units
meta-heuristic based computation. Problems in sup-
TS Total loss due to unavailability of a product
ply chains are predominantly of combinatorial opti-
at retailers, cost units
mization types, which can be solved using simulation
IW Inventory at warehouse
optimization with lesser computational complexity.
IT Threshold level for inventory
The software used in the present work, WIT-
NESS® is industry-standard simulation software with
The above described model is simulated to find the ability to model a wide range of process and op-
the optimum production rate for two product types eration tasks. It is a software platform for dynamic
with the objective to maximize the service level system modeling and simulation, which is developed
(availability of products at retailers) and at the same by the British Lanner, to cater the needs of industrial
time minimize the inventory level at the warehouse. and business systems and processes (Men and Zhou,
A higher inventory level improves the service level 2011). It has a wide range of application areas, a large
means the customer demands are fulfilled with less number of model elements, a powerful simulation
stock-out situations but higher inventory can be costly engine, a convenient graphical interface operation
for the firm as excess inventory will result in greater function and a hierarchical modeling function.
storage costs, risk of obsolescence, pilferage, insur- The leagile supply chain model is optimized using
ance premium, etc. On the other hand lower inven- the approach of simulation optimization. Simulation
tory level can results in stock-outs which will impair optimization deals with the situation in which the ana-
the service levels causing loss of sales and goodwill of lyst would like to find which of the many sets of mod-
el specifications (input parameters and/or structural

International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)

Pattanaik 107

assumptions) would lead to optimal performance. The simulation model was run for two hypotheti-
Initially, a trial model was developed in WITNESS® cal demand cycles to understand the behavior of the
with single supplier, manufacturer, warehouse and a supply chain under volatile demand conditions. Re-
single retailer which was later expanded to the pro- ferring to the notations for various parameters and
posed model as presented in Figure 2. variables as expressed in section 3, the numerical in-
WITNESS® is Lanner Group’s simulation soft- put data used for the model are as follows:
ware package which provides a visual, interactive and
λB 30-60 units/day α1 Exp (0.125) days
interpretative approach to simulation without the
need for compilation. The software has been adopt- Rc 1000 units α2 Exp (0.25) days
ed in discrete-event problems from various areas like Wc 6000 units α3 Exp (1.0) days
automotive, pharmaceutical, aerospace, electronics,
defence, services, etc. The WITNESS® Manufactur- ɣ1 30 units t3 4.0 days
ing Performance Edition has been specially designed ɣ2 55 units CPA 100 cost units
for manufacturing applications. It is ideally suited to
ɣ3 75 units MPA 160 cost units
a variety of production and storage layout, logistical
modelling and supply chain modelling scenarios. t1 0.5 days CPB 150 cost units
t2 1.0 days MPB 250 cost units
4.1 Experimentation with simulation model IT 500 units
For experimentation on the constructed model of Case (i) First demand cycle
the leagile supply chain, two product types A and B
DA : Uniform distribution [5, 10] units/day
are considered where A is cheaper and of higher de-
mand than B. When the stock with the retailer drops DB : Uniform distribution [5, 7] units/day
below the reorder point p units, the retailer places an p1 = 5 units
order of reorder quantity q to the warehouse. The
warehouse receives the order and forwards to ‘Pick- p2 = 10 units
ing’ where the orders are dispatched to the retailer p3 = 50 units
via delivery.
The demand is taken to be uniformly distributed
The model was simulated for 1000 days with 250
integer values between the upper and lower bound
days as warm-up period and the statistics of the in-
values. The model was run for different production
ventory accumulated at the warehouse and the unful-
rates within the plant capacity ranges λA and λB. The
filled customer demand percentage at each retailer
results for average unfulfilled customer demand per-
were found. The unfulfilled customer demand per-
centage of all the retailers and the inventory accumu-
centage is a direct indication of service level.
lated in the warehouse were found for products A
Service level %=(1-unfulfilled customer
and product B as produced in Table 1.
demand %)

Figure 2. Screen print of the leagile supply chain in WITNESS®

International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)

108 Pattanaik

The losses due to poor service level at retailer Referring to (1) and (2),
end and high inventory at warehouse are calculated
by using Eq. (1) and (2). An example is given here to
explain the calculation of total loss.
Production rate (Product A) = 40 units/day
Unfulfilled customer demand = 7.17%
Inventory level at warehouse = 40
Simulation run time = 1000 days Similarly, the total loss TL is calculated for all oth-
er production rates for product A and product B as
Average demand per day = 7.5 units/day given in Table 2.
Profit for product A = 60/unit

Table 1. Simulation results of the first demand cycle

Product A Product B

Production Ave. % unfulfilled Inventory at Ave. % unfulfilled Inventory at

Production rate
rate customer demand warehouse customer demand warehouse

40 7.17 40 30 10.86 0

41 5.32 41 32 8.14 0

42 3.78 42 34 3.95 34

43 2.47 43 35 1.86 35

44 2.12 550 36 0.86 450

45 1.89 1580 37 0.99 1920

48 1.33 5000 38 0.73 2950

49 1.55 6000 39 0.86 4200

50 1.45 5970 40 0.85 5450

Table 2. Total losses for products at various production rates in first demand cycle

Product A Product B

Production
TS TI Total Loss Production rate TS TI Total Loss
rate

40 32257.5 0 32257.5 30 65180 0 65180

41 23932.5 0 23932.5 32 48810 0 48810

42 17010 0 17010 34 23680 0 23680

43 11122.5 0 11122.5 35 11180 0 11180

44 9517.5 5000 14517.5 36 5170 0 5170

45 8490 108000 116490 37 5940 213000 218940

48 5992.5 450000 455992.5 38 4390 367500 371890

49 6952.5 550000 556952.5 39 5170 555000 560170

50 6540 547000 553540 40 5100 742500 747600

International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)

Pattanaik 109

In Figure 3, the trade-off between average of carrying extra inventory.

percentage customer unfulfilled demand and inven- The data for product A from Table 2 are plotted
tory level at warehouses can be clearly observed. In in Figure 4 to find the optimum production rate of
order to reduce the unfulfilled customer demand, 43 units/day (highlighted in a circle) that gives the de-
higher inventory stocks at warehouses and retailers sired objective of minimum total loss.
are required which again leads to the disadvantages

Unfulfilled demand
Inventory level

Figure 3. Variation in unfulfilled demand and inventory level with production rate of A

Figure 4. Variation in total loss with production rate for product A

Figure 5. Variation in unfulfilled demand and inventory level with production rate of B

International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)

110 Pattanaik

Similar to the Figure 3 (for Product A), Figure p2 = 10 units

5 indicates the same variation and trade-off relation
p3 = 50 units
between unfulfilled demand and inventory level of
product B. The data for average unfulfilled customer demand
Figure 6 shows the optimum production rate for percentage of all the retailers and the inventory accu-
product B as 36 units/day where the total loss in- mulated at the warehouse was collected for product
curred is minimum. A and product B separately as presented in Table 3.
Here, it can be seen that the average unfulfilled
Case (ii) Second demand cycle
customer demand percentage for product A is very
DA : Uniform distribution [10, 15] units/day high and is not satisfying a service level of at least
95%. The reason is that the re-order point, pi being
DB : Uniform distribution [7, 10] units/day
too low. The re-order point of the retailer needs to
p1 = 5 units be heightened as the demand increases. After recog-

Figure 6. Variation in total loss for product B

Table 3. Simulation results of the second demand cycle

Product A Product B

Production Ave. unfulfilled Inventory at Ave. unfulfilled Inventory at

Production rate
rate customer demand % warehouse customer demand % warehouse

60 11.12 90 45 7.15 45

65 7.63 70 48 3.76 270

67 6.81 1910 49 3.94 1740

68 6.51 2720 50 3.91 2660

69 7.56 4740 51 3.98 4240

70 7.68 5870 52 4.61 5600

72 6.94 6000 55 4.42 5995

75 6.83 5835 60 4.48 5930

International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)

Pattanaik 111

nizing this, the model was simulated by enhancing at all the levels of production rates as produced in
the re-order points as below. Table 5. The corresponding data is shown below in
Table 10.
p1 = 10 units
The data from Table 5 are plotted to mark the
p2 = 20 units optimum production rate for products A and B in
p3 = 80 units Figures 7 and 8 respectively. The optimum produc-
tion rates for the products A and B are 69 and 49
The simulation results for the revised model are units/day as highlighted in circles.
given in Table 4.
Total loss TL is calculated for both the products

Table 4. Simulation results of the second demand cycle with revised re-order points

Product A Product B

Production Ave. unfulfilled Inventory at Ave. unfulfilled Inventory at

Production rate
rate customer demand % warehouse customer demand % warehouse

60 10.78 60 45 7.83 45

65 6.89 65 48 2.91 48

67 5.36 150 49 1.86 200

68 4.47 68 50 1.85 790

69 3.73 340 51 2.36 2480

70 4.31 1040 52 2.48 3840

72 3.63 3430 55 2.78 5775

75 4.35 5705 60 2.35 5940

Table 5. Total losses for products at various production rates in second demand cycle

Product A Product B

Production
TS TI Total Loss Production rate TS TI Total Loss
rate

60 80825 0 80825 45 66526.6 0 66526.6

65 51662.5 0 51662.5 48 24692.5 0 24692.5

67 40187.5 0 40187.5 49 15795.8 0 15795.8

68 33550 0 33550 50 15753.3 43500 59253.3

69 27987.5 0 27987.5 51 20074.1 297000 317074.1

70 32287.5 54000 86287.5 52 21108.3 501000 522108.3

72 27212.5 293000 320212.5 55 23601.6 791250 814851.6

75 32625 520500 553125 60 20003.3 816000 836003.3

International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)

112 Pattanaik

5. Conclusion lean portion of the supply chain. Similar to the takt

time concept of lean manufacturing, the optimal pro-
In the present paper, a trade-off problem iden- duction rate is capable of meeting customer demands
tified between inventory cost and shortage cost in a as pulled from the downstream retailers as well as
leagile supply chain. The two conflicting objectives minimize the inventory carrying cost at the ware-
were maximization of service level at the retailer end house or distribution center. To implement leagile
and minimization of inventory at the warehouse, supply chain, different demand cycles were consid-
the decoupling point. A simulation optimization ap- ered to understand how the model can respond to
proach implemented using WITNESS® simulation changes in the market demand. The data for two hy-
software to find the optimal rate of production for pothetical demand cycles were generated using ran-

Figure 7. Variation in total loss for product A in second demand cycle

Figure 8. Variation in total loss for product B in second demand cycle

International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)

Pattanaik 113

dom distributions to run the simulation model for acteristics: Creation of virtual groups by enhanced
production flow analysis. Int. J. Prod. Econ. 85(3), 305-318.
two product types. The bi-objective function outputs
[7] Xiaomei, Li, Zhaofang, Mao, Guohong, Xia, Fu Jia, 2008.
were transformed into the same cost units for ease of Study on manufacturing supply chain leagile strategy driven
comparison and analysis. factors based on customer value, 4th International
An adequate mean service level of 95% and above Conference on Wireless Communications,
Networking and Mobile Computing, Dalian, China
was reached while minimizing inventory costs for the DOI: 10.1109/WiCom.2008.1642
two demand scenarios. The optimal rate of produc- [8] Dev, N. K., Ravi Shankar, Prasanta Kumar Dey, 2014. Re
tion for the two products in two different demand cy- configuration of supply chain network: an ISM-based
roadmap to performance. Benchmarking: An International
cles was found. It was also established a fact that the Journal. 21(3), 386-411.
re-order point plays a key role as the demand fluctu- [9] Purvis, L., Gosling, J., Naim, M. M., 2014. The
ates. The retailer must increase its re-order point to development of a lean, agile and leagile supply network
meet the increasing customer demand. taxonomy based on differing types of flexibility. Int. J. Prod.
Econ. 151, 100-111.
Application of simulation optimization approach [10] Shashi, Piera Centobelli, Roberto Cerchione and Myriam
to a leagile supply chain to find optimum production Ertz, 2020. Agile supply chain management: where did it
rate is a novel exploration which this paper report- come from and where will it go in the era of digital
transformation?, Industrial Marketing Management, 90,
ed. As a future scope of the present work, the model 324-345.
could be expanded by adding multiple manufactur- [11] Mansoor Shekarian, Seyed Vahid, Reza Nooraie and
ing plants and warehouses at different geographical Mahour Mellat Parast, 2020. An examination of the impact
of flexibility and agility on mitigating supply chain
locations to represent typical automobile firms. The
disruptions, International Journal of Production
effects of several other parameters like location of Economics, 220, 107438.
de-coupling point, truckload capacity, re-order quan- [12] Yuik, C. Jia and P. Puvanasvaran. 2020. Development of
tity, etc. can also be studied on the unfulfilled de- Lean Manufacturing Implementation Framework in
Machinery and Equipment SMEs. Int. J. Ind. Eng. Manag.,
mands and inventory levels. 11(3), 157-169.
[13] Naylor, J. B., Mohamed M Naim, Danny Berry, 1999.
Leagility: Integrating the lean and agile manufacturing
Acknowledgements paradigms in the total supply chain. Int. J. Prod. Econ. 62,
107-118.
The author wish to thank the two anonymous [14] Christopher, M., Towill, D., 2001. An integrated model for
reviewers for their valued review and suggestions to the design of agile supply chains. Int. J. Phys. Distr. Log.
31(4), 235-246.
improve the content and presentation of the paper. [15] Hoek, R. V., 1998. Reconfiguring the supply chain to
implement postponed manufacturing. Int. J. Logist. Manag.
9(1), 95-110.
Funding [16] Rahiminezhad Galankashi, M., Helmi, S.A. 2016.
Assessment of hybrid Lean-Agile (Leagile) supply chain
This research did not receive any specific grant strategies, J. of Manuf. Tech. Manage. 27(4), 470-482.
from funding agencies in the public, commercial, or [17] Huang, Y. Y., Li, S. J., 2010. How to achieve leagility: a case
study of a personal computer original equipment
not-for-profit sectors. manufacturer in Taiwan. J. Manuf. Syst. 29(2-3), 63-70.
[18] Ambe, I. M., J. A. Badenhorst-Weiss, 2010. Strategic supply
chain framework for the automotive Industry. Afr. J. Bus.
References Manage. 4(10), 2110-2120.
[19] Shukla, S. K., Wan, H. D., 2010. A leagile inventory-
[1] Dev, N. K., Ravi Shankar, Angappa Gunasekaran, location model: formulation and its optimisation. Int. J.
Lakshman S. Thakur, 2016. A hybrid adaptive decision Oper. Res. 8 (2), 150-173.
system for supply chain Reconfiguration. Int. J. Prod. Res. [20] Komoto, H., T. Tomiyama, M. Nagel, S. Silvester, H.
54(23), 7100–7114. Brezet, 2005. A multi-objective reconfiguration method of
[2] Naim, M. M., Gosling, J., 2011. On leanness, agility and supply chains through discrete event simulation, 4th Inter
leagile supply chains. Int. J. Prod. Econ. 131(1), 342-354. national Symposium on Environmentally Conscious
[3] Chatterjee, P. and Stević, Ž. 2019. A two-phase fuzzy Design and Inverse Manufacturing, Tokyo, Japan.
AHP-fuzzy TOPSIS model for supplier evaluation in DOI: 10.1109/ECODIM.2005.1619238
manufacturing environment. Operational Research in [21] Peirleitner, A. J., Thomas F., Klaus, A., 2016. A
Engineering Sciences: Theory and Applications, 2(1), 72-90. simulation approach for multi-stage supply chain
[4] Klochkov, Y., Gazizulina, A., and Muralidharan, K. (2019). optimization to analyze real world transportation effects.
Lean six sigma for sustainable business practices: A case Proceedings of the Winter Simulation Conference (WSC),
study and standardization. International Journal for Quality Washington, DC, USA. DOI: 10.1109/WSC.2016.7822268
Research, 13(1), 47-74. [22] Diamantino Torres, Ana Raquel Xambre and Leonor
[5] Rachel Mason-Jones, Ben Naylor, Denis, R. Towill, 2000. Teixeira, 2016. Development of Synchronized Logistics
Lean, agile or leagile? Matching your supply chain to the Scenarios, Int. J. Ind. Eng. Manag., 7(2), 85-94.
marketplace. Int. J. Prod. Res. 38(17), 4061-4070. [23] Mohsen, J., Tillal, E., Aisha, N., Lampros, K. S., Terry, Y.,
[6] Prince, J., Kay, J. M., 2003. Combining lean and agile char 2010. Simulation in manufacturing and business: A review.
Eur. J. Oper. Res. 203(1), 1–13.

International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)

114 Pattanaik

[24] Othman, S., Noorfa, H., 2012. Supply chain simulation

and optimization methods: an overview, 3rd International
Conference on Intelligent Systems Modelling and
Simulation, Kota Kinabalu, Malaysia, DOI:
10.1109/ISMS.2012.122
[25] Ran Liu, Xiaolei Xie, Kaiye Yu, Qiaoyu Hu, 2018. A survey
on simulation optimization for the manufacturing system
operation, Int. J. of Mod. Sim. 38 (2), 116-127.
[26]
Maedeh Mosayeb Motlagh, Parham Azimi, Maghsoud
Amiri, Golshan Madraki, 2019. An efficient simulation
optimization methodology to solve a multi-objective
problem in unreliable unbalanced production lines, Expert
Systems With Applications, 138, 112836.
[27] Rafael D. Tordecilla, Angel A. Juan, Jairo R. Montoya-
Torres, Carlos L. Quintero-Araujo, Javier Panadero, 2021.
Simulation-optimization methods for designing and
assessing resilient supply chain networks under uncertainty
scenarios: A review, Simulation Modelling Practice and
Theory, 106, 102166.
[28]
Nishi, T., Yoshida, O., 2016. Optimization of multi-
period bi-level supply chains under demand uncertainty.
Procedia CIRP, 41, 508–513.
[29] Matheus, C. P., Enzo, M. F., Apolo, M. C. D., Mirko, K.,
Michael, F., 2018. Towards a simulation-based optimization
approach to integrate supply chain planning and control,
51st CIRP Conference on Manufacturing Systems, Procedia
CIRP. 72, 520–525.
[30] Liotta, G., Kaihara, T., and Stecca, G., 2016. Optimization
and simulation of collaborative networks for sustainable
production and transportation. IEEE T. Ind. Inform. 12(1),
417–424.
[31]
Truong, T. H., Azadivar, F., 2003. Simulation based
optimization for supply chain configuration design.
Proceedings of the Winter Simulation Conference, New
Orleans, LA, USA, DOI: 10.1109/WSC.2003.1261560
[32] Ong, J. Q., Latif, M., Kundu, S., Tyagi, G. K., Sehgal, P.,
2014. Exploiting WITNESS simulation for SCM. Int. J. of
Res. Manage. Sc. Tech. 2(2), 103-109.

International Journal of Industrial Engineering and Management Vol 12 No 2 (2021)