ArticlePDF Available

A design of distribution network and development of efficient distribution policy

Authors:

Abstract and Figures

The distribution system design considered in this paper is derived from current operations of a major consumer product company in India who manages products for nationwide distribution. The company is charged with Central Sales Tax (CST) along with local tax VAT for the trade that happens across the state borders while distributing the products from the factories. To avoid CST, company could consider opening warehouses in each state. This will have an extra cost of opening and operating warehouse in each state. So, this needs a cost trade-off between the option of continuing with the existing system or opening a new warehouse. Accordingly, a two-stage integrated solution framework is proposed in this study. In the first stage, a mathematical model is proposed to solve the facility location and production distribution policy for the company by minimising the cost comprising production costs at factory, fixed and labour cost of new warehouses and shipment cost from factories to warehouses and warehouses to distributors. In the second stage, a breakeven analysis is proposed to compare the new proposed network with the existing network based on the distribution cost. The workability of the proposed framework is demonstrated using a pseudocase developed based on the observation of the company, particularly for nationwide existing distribution.
Content may be subject to copyright.
Int. J. Logistics Systems and Management, Vol.
Copyright © 2011 Inderscience Enterprises Ltd.
A design of distribution network and development
of efficient distribution policy
M. Mathirajan*
Operations Research Group,
Anna University Tiruchirappalli,
Tiruchirappalli 620 024, Tamil Nadu, India
E-mail: msdmathi@mgmt.iisc.ernet.in
E-mail: mathirajan@tau.edu.in
*Corresponding author
K. Manoj
Department of Management Studies,
Indian Institute of Science,
Bangalore 560 012, Karnataka, India
E-mail: Manoj.K@unilever.com
V. Ramachandran
Operations Research Group,
Anna University Tiruchirappalli,
Tiruchirappalli 620 024,
Tamil Nadu, India
E-mail: rama@annauniv.edu
Abstract: The distribution system design considered in this paper is derived
from current operations of a major consumer product company in India who
manages products for nationwide distribution. The company is charged with
Central Sales Tax (CST) along with local tax VAT for the trade that happens
across the state borders while distributing the products from the factories.
To avoid CST, company could consider opening warehouses in each state.
This will have an extra cost of opening and operating warehouse in each state.
So, this needs a cost tread-off between the option of continuing with the
existing system or opening a new warehouse. Accordingly, a two-stage
integrated solution framework is proposed in this study. In the first stage, a
mathematical model is proposed to solve the facility location and production
distribution policy for the company by minimising the cost comprising
production costs at factory, fixed and labour cost of new warehouses and
shipment cost from factories to warehouses and warehouses to distributors. In
the second stage, a breakeven analysis is proposed to compare the new
proposed network with the existing network based on the distribution cost. The
workability of the proposed framework is demonstrated using a pseudocase
developed based on the observation of the company, particularly for nationwide
existing distribution.
Keywords: design of the distribution system; CST and VAT; mathematical
model; facility location; breakeven analysis; case study.
A design of distribution network and development 109
Reference to this paper should be made as follows: Mathirajan, M., Manoj, K.
and Ramachandran, V. (2011) ‘A design of distribution network and
development of efficient distribution policy’, Int. J. Logistics Systems and
Management, Vol.
Biographical notes: M. Mathirajan is with Anna University Tiruchirappalli
(AU-T), Tiruchirappalli, and since May 2008, he is on lien for two years from
Indian Institute of Science (IISc), Bangalore. He is currently the Professor and
Head of the Department of Management Studies of AU-T. He received both
MSc (Engineering) by research and PhD in Operations Management from IISc.
He was a Postdoctoral Fellow at Singapore MIT Alliance of Nanyang
Technological University, Singapore. His research interests are in the areas of
mathematical modelling and development of heuristic methods for Operations
Management, Logistics/Supply Chain Management problems. He has been
guiding doctoral research students at IISc and AU-T.
Manoj Kumar received an MBA (2006–2008) from the Indian Institute of
Science (IISc) Bangalore. He had an interest in operations area and hence he
took a role of a Subject Matter Expert in the area of Customer Development
and Sales. He is currently working in the area of Information Management in
Unilever where he looks at the business needs of information in the space of
sales and distribution and how these information needs can be met. He obtained
a Mechanical Engineering Degree from Visveswaraya National Institute of
Technology (VNIT) in 2003.
V. Ramachandran received both PhD in Computer Applications in Power
Systems and ME in Power Systems Engineering from Anna University,
Tamil Nadu. He obtained his BE in Electrical and Electronics Engineering from
Madras University. Since 1980s, he has been with Anna University as Faculty
in various capacities, including Secretary for Tamil Nadu Engineering
Admissions. He was on sabbatical with Faculty of Information Technology,
Multimedia University, Malaysia. His current areas of research include media
streaming, service-oriented architecture, evolutionary programming in power
system applications and distributed models in power system analysis. He is
currently the Vice Chancellor of Anna University Tiruchirappalli,
Tiruchirappalli.
1 Introduction
The distribution network is an important part in the supply chain management and the
distribution network design problem is a very important strategic decision in supply
planning and management. Distribution network design problems involve both (a) the
optimisation of the flows of goods and (b) the improvement of the existing distribution
network (Ambrosino and Scutella, 2005). A good distribution network design can be used
to achieve a variety of supply chain objectives ranging from low cost to high
responsiveness (Chopra, 2003; Gunasekaran and Ngai, 2004; Liu and Hao, 2006).
In today’s rapidly changing economic and political conditions, it is a big challenge
for companies to constantly evaluate and optimally configure their distribution network.
An important strategic issue related to the design/re-design and operation of a physical
distribution network in a supply chain system is the determination of the best
sites/facilities for intermediate stocking points or warehouses (Melo et al., 2009). The use
110 M. Mathirajan et al.
of warehouse (facility) provides a company with flexibility to respond changes in the
marketplace and can result in significant cost savings owing to economies of scale in
transportation or shipping costs, inventory costs, tax structure while moving the product
from factory to warehouses located across the states within the country. Also,
the combined facility location/network design problems are useful for modelling a
number of situations in which trade-offs between facility costs, network design costs and
operating costs must be made (Melkote and Daskin, 2001).
During the configuration/reconfiguration of the distribution network design, operation
managers and planners need to address questions such as where to produce, where to
open the warehouses, and how it should be distributed through the distribution network.
The complexity of this problem increases for these operations managers when the selling
model of the company is multi-echelon, in which the products are shipped from
the factories to warehouses and from there distributed to the customers through the
distributors. In addition, the decision-makers need to consider the effects of taxes during
the design of distribution network because taxes add to their bottom-line cost of
distribution. Sometimes, it is possible for managers to avoid these taxes by a suitable
network design for their supply chain.
In this study, we are studying the distribution network of a multi-echelon supply
chain for a single product as observed in large consumer product company. The main
objective of the study is developing a mathematical model for the selection of warehouse
locations and distribution of product in the network. We have also considered the effect
of the sales tax in comparing the new proposed distribution network with the existing
network and proposed a breakeven analysis to make the decision.
2 Closely related review
The literature review is divided into two broad categories namely, (i) studies related to
facility location and distribution policy design issues, and (ii) studies related to
implementation of breakeven analysis methods in the area of supply chain management
issues.
Some researchers have looked at facility location problems as independent decision
issues (e.g., Aikens, 1985; Owen and Daskin, 1998; Bhutta, 2004; Wu et al., 2006;
Sahin and Sural, 2007; Rentizelas and Tatsiopoulos, 2010). There have been
many attempts to solve facility location problem along with other decision issues
such as allocation (e.g., Ohlemuller, 1997; Murat et al., 2010), allocation and routing
(e.g., Wu et al., 2002; Lashine et al., 2006; Nagy and Salhi, 2007), transportation and
inventory (e.g., Jayaraman, 1998; Perl and Sirisoponslip, 1988; Shen and Qi, 2007),
inventory control (e.g., Ballou, 1984; Miranda and Garrido, 2006; Wang et al., 2007;
Gebgennini et al., 2009), storage capacity calculation (e.g., Levén et al., 2004) and supply
chain management (e.g., Tsiakisa and Papageorgiou, 2008; Melo et al., 2009).
Furthermore, in the last decade, the facility location problem as an integral problem
while designing a distribution network has attracted the attention of researchers and these
types of researchers are closely related to the problem addressed in this paper. Chopra
(2003) described a framework for designing the distribution network in a supply chain by
considering various factors such as response time, product variety, product availability,
A design of distribution network and development 111
customer experience, order visibility and returnability influencing the choice of
distribution network. Klose and Drexl (2005) classified the facility location models for
distribution systems design. They also reviewed some of the contributions and
summarised the continuous locations models, network location models, mixed-integer
programming models and applications in an effective way.
Avittathur et al. (2003) studied the effect of CST rates and product variety
on Distribution Centre (DC) locations and concluded that CST has an effect. They
developed a non-linear mixed-integer programming model with an objective function of
minimising total cost to determine DC locations considering the impact of CST. They
have validated the proposed mathematical model using a real-life numerical problem
and for the simplicity of the analysis, they considered only the northern India for the
numerical problem. Their findings are based on the effects of DC fixed cost, product
variety, service levels, transportation cost and CST in deciding on number of DC
locations. Finally, their analysis concluded that CST rate is an important factor to
determine the optimal number of distribution locations.
Amiri (2006) studied the problem of designing a distribution network in a supply
chain system that involves determining simultaneously the best locations of both plants
and warehouses and best policy for distributing the product from the plants to warehouses
and from the warehouses to the customers. The researcher proposed mathematical model
by considering the traditional costs associated with facility location and distribution
problems such as variable cost due to supplying the products from plant to warehouses
as well as from warehouses to customers and fixed cost due to opening and operating
warehouse with specific capacity level. Also, proposed heuristic solution procedure based
on Lagrangian relaxation of the problem and demonstrated the quality of the proposed
heuristic algorithms by conducting suitable computational experiments.
Bidhandi et al. (2009) proposed a new approach for determining supply chain
network design considering simultaneously the strategic decisions concerning facilities
selection with the tactical decisions concerning supplier, production, warehouse and
customer allocation, the facility location and allocation problem. They proposed
Mixed-Integer Linear Programming (MILP) model with (0–1) variables. Because of the
computational difficulties, they proposed a modified version of Benders’ decomposition
to solve MILP model with (0–1) variables and following this they developed a new
algorithm based on the surrogate constraints. Finally, they discussed the quality of the
new algorithm.
While making decisions for facility locations and distribution networks
simultaneously, many factors have been considered in the literature including the tax
structure of the country as an important factor. However, the literature review indicates
that even though the effects of CST on distribution costs have been studied, particularly
in India, there are no instances observed where the CST is modelled in the solution
framework to make decisions on facility locations and distribution network design.
In addition to the literature gap mentioned earlier, to the best of our knowledge,
not many researchers are seen using breakeven analysis in making decisions in the areas
of distribution network design. However, Hanna et al. (1993) used breakeven analysis to
modern production economics while modelling economies of scale and economies of
scope. They found traditional cost benefit analysis inadequate in their context of study
and suggested a modified cost-volume-flexibility breakeven analysis for their study.
112 M. Mathirajan et al.
Norton et al. (2006) used breakeven analysis in their study to compare costs of upgrading
centralised treatment facility and an alternative distributed treatment facility.
3 Problem descriptions
The problem considered in this study is observed in a large consumer product company
in India, which is involved in manufacturing of electronic consumer products. This
company felt the need to reconfigure the existing supply chain because of the current
practice of CST. The CST is a tariff that is levied on all the inter-state trade transactions.
It is payable when there is any trading goods transaction, between any two parties
(e.g., manufacturer and distributor), does not happen in the same state. In other words,
when a manufacturer located in state ‘A’ ships goods to a distributor located in state ‘B’,
the transaction will be levied upon CST as well as local sales tax (=VAT – Value Added
Tax) of state ‘B’.
Let us consider a traditional supply chain network operating in India (comprising of
28 states), which is selling electronic consumer products through distributors. There are
one or few production houses (Factories) shipping goods to distributors. Generally, there
are few distributors in each state of the country. These distributors meet the demand of
stockiest, which are more in numbers for a particular state, generally one or two per
district. Finally, the product will reach to the customer through these stockiest and the
retailers. An example of a complete chain is shown in Figure 1.
Figure 1 Traditional supply chain in India selling through distributors
Now, suppose the Factory is in state ‘A’ and the Distributor is located in state ‘B’
(Figure 2). In such case, the CST is levied on goods traded between them, in addition to
VAT. To avoid this CST, the company can open a Warehouse in state ‘B’ and transaction
is shown from warehouse to distributor, thus avoiding interstate selling and hence CST.
But, the introduction of the warehouse incurs extra fixed cost of opening a warehouse in
each state and also the operating cost of warehouses add up to the distribution cost.
Hence, this part of the supply chain (as shown in dashed box in Figure 2) is of interest to
operations manager. They have to develop a model to select the location for new
warehouse, compare the extra cost of opening and operating new warehouse with existing
network design and reconfigure the distribution network.
A design of distribution network and development 113
Figure 2 Tax structure for interstate selling and introduction of state-owned depot
3.1 Objectives of the study
Though the problem described in this section is related to a particular consumer product
company observed in Bangalore, similar problem is applicable to other companies having
supply chain across the states of India. The following objectives are emerged based on
the observation made in the industry:
1 designing new distribution network and developing optimal distribution plan
2 finding the breakeven point at which the selected potential warehouse(s)
should be operational
3 demonstrating the workability of proposed solution framework by solving
a pseudocase study.
To obtain the objectives stated in this study, the following assumptions are made in the
problem to develop the solution methodologies:
each state should have at least one warehouse
factory locations are known
locations of potential warehouses are known
only one product is considered
warehouse located in state ‘j’ can ship goods to distributors located in state ‘j’ only
inventory costs are not considered
transportation costs are assumed to be unit transportation cost.
4 Solution framework
In this section, we propose the solution framework for the problem discussed.
The decision problem addressed in this study is viewed as the two-stage integrated
decision problem, namely (1) decision on warehouse location, and (2) decision on
finalising the warehouse location in comparison with the existing distribution design
114 M. Mathirajan et al.
network. The modelling and the algorithm details of this solution framework are
discussed in the following subsections.
4.1 Stage 1: Development of mathematical model
In the first stage of the proposed solution framework, the optimal decision on location of
the potential warehouse is to be decided out of various possible locations while the
distribution policy is formulated. To get the optimal decision on warehouse location
along with optimal distribution policy, an Integer Linear Programming (ILP) is developed
to solve Facility Location and Production–Distribution Policy (FLPDP) issues
simultaneously. The assumptions made and notations used for formulating the FLPDP
model are:
Assumptions
1 Single product is considered
2 There is no breakage considered in goods shipment
3 No inventory issues are considered
4 The warehouse in state j will ship goods to distributors in state j only
5 Unit transportation costs are considered in the problem.
Notations
Sets
F: (1 I) factories and i denotes ith factory
S: (1 J) states and j denotes jth state
W: (1 K) potential warehouses and k denotes kth potential warehouse
D: (1 M) distributors and m denotes mth distributor.
Parameters
R_WH: Required no. of warehouses in state S
P_WH: No. of potential warehouses in state S
K: Maximum number of potential warehouse in state S
DBUTOR: No. of distributors in state S
M: Maximum number of distributors in state S
C_F: Capacity of factory F
Capacity: Max capacity allowed for the potential warehouse
PC_F: Production cost at factory F
FC_SW: Fixed cost of opening a warehouse W in state S
LC_SW: Labour cost/operating cost per unit of warehouse W in state S
A design of distribution network and development 115
N_SD: Demand at distributor D in state S
TC_SWF: Unit transportation cost of shipping goods from factory F to warehouse W
in state S
TC_SWD: Unit transportation cost of shipping goods from warehouse W to distributor
D (both in state S).
Decision variables
Y_SW: 1, if warehouse W is opened in state S
0, otherwise
X_SWF: Quantity of goods shipped from factory F to warehouse W in state S
X_SWD: Quantity of goods shipped from warehouse W to distributor D (both
in state S).
The proposed ILP model
Objective function
_
11 1
_
11
(_() _ ())*_ ()
_()*_()
PWH
IJ
ij k
PWH
J
jk
PC F i TC FSW ijk X FSW ijk
FC SW jk Y SW jk
== =
==
+
++
∑∑ ∑
∑∑
_DBUTOR
11 1
(_ () _ ( ))*_ ( )
PWH
J
jk m
LC SW jk TC SWD jkm X SWD jkm
== =
+
∑∑ ∑ (1)
_
11
_()_()
PWH
J
jk
X
FSW ijk C F i i I
==
≤∀
∑∑ (2)
DBUTOR
1
_ ( ) Capacity * _ ( ) ,
m
X
SWD jkm Y SW jk j J k K
=
≤∀
(3)
DBUTOR
11
_() _() ,
I
im
X
FSW ijk X SWD jkm j J k K
==
=∀
∑∑ (4)
_
1
_()_() ,
PWH
k
X
SWD jkm N SD jm j J m M
=
≤∀
(5)
_
1
_()_()
PWH
k
YSWjk RWHj j J
=
≤∀
(6)
_()PWHj K j J≤∀ (7)
DBUTOR( )jM jJ≤∀ (8)
116 M. Mathirajan et al.
_(){0,1} ,YSWjk j Jk K∈∀ (9)
_()0 ,,
X
FSW ijk i I j J k K≥∀∈∈ (10)
_
()0 , , .
SWD jkm j J k K m M≥∀∈ ∈ (11)
The objective is to minimise total cost, which includes: Production Cost, opening and
operating cost of warehouse, and transportation costs. Constraint set (2) is the Capacity
Constraint of factories. It ensures that the total quantity of goods shipped from each
factory is less than the capacity of the factory. Constraint set (3) is the Capacity
Constraint of the potential warehouses. Although in our problem the capacity constraint
of the warehouse is not considered but it is important because apart from capacity
constraint of the warehouses it also ensures the relation between quantity shipped and
binary variable Y_SW (whether the potential warehouse is selected or not). In our
problem, the value for capacity is given a large number so that this constraint set only
acts as the relation between quantity shipped and binary variable Y_SW. Constraint set (4)
is the Demand Constraint of the distributors. It ensures that the quantity of goods shipped
to a distributor is equal to the demand of that distributor. Constraint set (5) is the Flow
Balance Constraint at each potential warehouse. It ensures that the quantity of goods
shipped to the warehouse is equal to the quantity of goods shipped out of the warehouse.
It also ensures that the warehouse in state j will ship goods to distributors in state j only.
Constraint set (6) ensures that the number of warehouses selected for a state is equal
to the required number of warehouses in that state. Constraint set (7) ensures that the
number of potential warehouse locations does not exceed the maximum number of
potential warehouse in a state. Constraint (8) ensures that the number of distributors does
not exceed the maximum number of distributor in a state. Constraint set (9) ensures that
the variable Y_SW takes the binary values. Constraint sets (10) and (11) are non-
negativity constraint for the respective decision variables.
A LINGO set code, which generates the ILP model proposed in this section for any
given data, was developed and presented in Appendix 1. Both the proposed ILP model
and the LINGO set code were verified appropriately by solving a small-scale numerical
problem and by interpreting the optimal solution. Owing to the brevity of the paper,
the verification details are not presented here. As the proposed ILP model has minimum
number of binary decision variables, the computational difficulties of the proposed model
do not arise.
The ILP model proposed in this study does not involve routing decisions for serving
the warehouses from the plants and from the warehouses to distributors. These additional
decisions can be achieved using Travelling Salesman Problem’s modelling approach
by introducing a third phase in the existing solution methodologies.
4.2 Stage 2: Development of breakeven analysis procedure to determine
optimal distribution design
Once location of potential warehouse is decided, breakeven analysis is carried out,
as second stage of the problem, to find out whether it is beneficial to go with proposed
distribution design compared with the existing design partly or totally. Accordingly, a
procedure for carrying out breakeven analysis is developed here to find out that is it
A design of distribution network and development 117
beneficial for the company to open the warehouse in the state as suggested by the FLPDP
model or should it stick to the existing distribution design in that state.
The procedure developed for breakeven analysis involves comparison of the cost of
existing distribution and new proposed distribution for each state. Total cost of existing
distribution will include: Transportation cost from the factories to the distributors and the
CST for interstate selling. Total distribution cost for new proposed distribution system
will include: Transportation cost from factories to warehouses, opening and operating
cost (labour cost) at warehouses, transportation cost from warehouses to distributors.
In the end of the comparison of the cost of existing distribution and new proposed
distribution for each state, the optimal Breakeven Period is computed to decide whether
to go for new proposed distribution system and open the warehouses. This detail of the
proposed breakeven analysis is shown in Figure 3 with the following terminology used
in the proposed procedure for the breakeven analysis:
Figure 3 Flow chart for breakeven analysis
Net Gain: For state ‘j’, if the proposed distribution system is implemented, the Net Gain
in distribution cost can be found out by:
= (Transportation Cost from the factories to the distributors+ the CST
for interstate selling) (Transportation Cost fromfactories to
warehouses + Operating Cost (Labour Cost) at warehouses+
Tr
Net Gain
ansportation Cost from warehouses to distributors).
Breakeven Period: is defined as the time period in which the cost of new proposed
distribution will be equal to the cost of existing distribution. In this study, the Breakeven
118 M. Mathirajan et al.
Period is computed in two ways, namely (a) Breakeven Period without time vale of
money (denoted as N), and (b) Breakeven Period with time value of money (denoted as
Nr). If the rate of interest is known, the time value of money can be incorporated
in the computation of Breakeven Period. With this, the computation details of N and Nr
are as follows:
_
N
etGain
F
CSW
N= (12)
where
N: Breakeven Period without time value of money
FC_SW: Fixed Cost of opening a warehouse W in state S
_*
ln 1 NetGain
ln(1 )
F
CSWR
Nr R



=
+ (13)
where
Nr: Breakeven Period with time value of money
FC_SW: Fixed cost of opening a warehouse W in state S
R: Rate of interest for the company.
5 A pseudocase study
In this section, a pseudocase based on the observation from a consumer electronics
company in South India was developed and the same was used to demonstrate the
workability of the proposed solution framework presented in the previous section.
5.1 Problem description for the case study
The case study organisation is a marketing company (hereafter called as EFL) involved
in the sales and service of water purifiers and vacuum cleaners. The products marketed
by EFL come from five production facilities, which are located across the country. These
production facilities along with the product are located at the places mentioned here.
Bangalore (BLR), Karnataka State: Water purifier
Dehradun (DUN), Uttaranchal State: Water purifier
Shillong (SLG), Meghalaya State: Water purifier
Baddi (BDD), Himachal Pradesh State: Water purifier
Bhimtal (BHM), Uttaranchal State: Water purifier and vacuum cleaner.
Some of these production facilities are located in small states because of the tax benefits
given by the state governments to promote industrial growth in the above-mentioned
states. For example, Uttaranchal gives tax benefits to any company on the products
A design of distribution network and development 119
produced in their state. It brings down the cost of production in those factories and hence
an incentive to increase the production capacity in those factories.
The products marketed by EFL reaches the ultimate customer through various
distribution channels. Figure 4 shows the business distribution of the company. Around
35% goes through direct sales, Industrial division caters to 15% of sales, 25% reaches the
customers through distributors and rest 25% is through services and other sales.
Currently, 25% business through distributors (shown in a shaded box: Figure 4) is
growing at a Compounded Annual Growth Rate (CAGR) of 40%, which makes it an
important part of their business. There are around 100 of such distributors across the
country.
Figure 4 Business distribution of EFL (see online version for colours)
Considering the business through distributor channel only, the distribution network for
EFL is shown in Figure 5. The products are manufactured by the production facility units
and passed on to warehouse. From these warehouses, the products are distributed to
various EFLs’ Distributors (EFLDs). Suppose the warehouse is located in state S1 and it
is distributing goods to EFLDs located in states S1, S2, S3, , Sn, (transit points), then
the goods traded with EFLD(S1) will not attract CST because the transaction is
happening in same state. But for the goods trades with EFLDs (S2, S3,, Sn) will attract
CST (as shown in shaded box in Figure 5) because of the interstate selling. If the
company wants to avoid this CST levied on interstate selling, it has to open their own
warehouse in each state where goods can be shipped from production facilities and then
goods can be traded from these warehouses to the EFLDs in that state, thereby avoiding
any interstate selling. But, there will be extra cost of opening and operating warehouse in
each state, so EFL has to trade off between the two options.
As some of the real data related to the research problem addressed here was not
available, it is randomly generated to demonstrate the solution framework proposed in the
study on real-life size problems and the same is discussed in the next section.
120 M. Mathirajan et al.
Figure 5 Distribution network of EFL (see online version for colours)
5.2 Data generation for the case study
Only one product: water purifier is considered because it is manufactured in all five
production facilities while vacuum cleaner was produced only in one plant. The required
data for these five production facilities such as PC_F (Production Cost), C_F
(Capacity of the production facility/factory) are assumed to be known and these are
shown in Table 1.
Table 1 Production Cost (PC_F) and Capacity (C_F) of Factories
Factory PC_F (in Rs.) C_F (in Units)
BLR 6200 22,000
DUN 6350 27,500
SLG 6500 35,000
BDD 6300 12,000
BHM 5800 15,000
State-wise, the data on required no. of warehouses (R_WH), the number of potential
warehouses (P_WH) and the number of distributors (DBUTOR) are generated randomly
considering the size (in terms of area) of the state. These data are shown in Table 2.
Unit transportation cost of shipping goods from factory F to warehouse W (TC_SWF)
and unit transportation cost of shipping goods from warehouse W to distributor D
(TC_SWD) are generated based on the physical distance between the cities where
these warehouses and distributors are located. Cost is assumed to be proportional to
the distance between two cities.
Data for other parameters such as demand at distributor D in State S (N_SD), fixed
cost of opening a warehouse W in State S (FC_SW), and labour-cost/operating-cost per
unit of warehouse W in State S (LC_SW) are generated randomly by assuming uniform
distribution functions U (50, 1050), U (30,000, 90,000) and U (25, 75), respectively.
These data is shown in Appendix 2.
A design of distribution network and development 121
Table 2 Required no. of warehouses, number of potential warehouses and number
of distributors for each state
S. No. State
State
code
Required No. of
warehouses
(R_WH)
No. of potential
warehouses
(P_WH)
No. of
distributors
(DBUTOR)
1 Andhra Pradesh APR 2 5 8
2 Arunachal Pradesh ARP 1 2 2
3 Assam ASM 1 3 3
4 Bihar BIH 2 5 8
5 Chhattisgarh CHT 1 2 3
6 Goa GOA 1 2 2
7 Gujarat GUJ 2 5 9
8 Haryana HAR 1 4 5
9 Himachal Pradesh HPR 1 2 3
10 Jammu and Kashmir JNK 1 2 2
11 Jharkhand JKD 1 2 4
12 Karnataka KAR 2 5 10
13 Kerala KER 1 2 5
14 Madhya Pradesh MPR 1 4 6
15 Maharashtra MAH 2 5 10
16 Meghalaya MEG 1 1 2
17 Mizoram MIZ 1 1 2
18 Nagaland NAG 1 2 2
19 Orissa ORI 1 4 6
20 Punjab PUN 1 3 5
21 Rajasthan RAJ 2 5 8
22 Sikkim SKM 1 1 2
23 Tamil Nadu TND 2 5 9
24 Tripura TRI 1 2 2
25 Uttar Pradesh UPR 2 5 10
26 Uttaranchal UTT 2 2 3
27 West Bengal WBN 2 5 7
28 Delhi DEL 2 5 10
5.3 Results and discussion for the pseudocase
Using the proposed solution framework, presented in Section 4, the pseudocase study
problem presented in the previous section is solved. The details of the implementation
process of the solution framework for the pseudocase study are discussed in the following
sections.
122 M. Mathirajan et al.
5.3.1 Development of ILP model for case study problem
All the relevant data on the case study is given as an input to the LINGO set code
(presented in Appendix 1) for generating the proposed ILP model for the case study data.
The generated ILP model for the case study data is solved using LINGO, an optimisation
package. Only the optimal non-zero solution is extracted from the detailed optimal
solution obtained for the case study data and the same is given to a Report Generator
Programme, developed using C++, to generate the following three managerial and
operational reports for its easy implementation of the optimal solution.
‘Report 1’ provides a strategic decision on state-wise the optimal number of
warehouses required and their optimal locations. This complete report for the case study
data is shown in Table 3.
Table 3 Optimal warehouse locations selected for each state
State No. of Warehouses selected Warehouse locations
APR 2 W03, W05
ARP 1 W01
ASM 1 W01
BIH 2 W02, W04
CHT 1 W01
GOA 1 W01
GUJ 2 W03, W04
HAR 1 W03
HPR 1 W02
JNK 1 W01
JKD 1 W02
KAR 2 W01, W02
KER 1 W02
MPR 1 W04
MAH 2 W02, W04
MEG 1 W01
MIZ 1 W01
NAG 1 W02
ORI 1 W04
PUN 1 W03
RAJ 2 W01, W02
SKM 1 W01
TND 2 W01, W05
TRI 1 W01
UPR 2 W03, W04
UTT 2 W01, W02
WBN 2 W04, W05
DEL 2 W02, W05
A design of distribution network and development 123
‘Report 2’ gives an operational decision on the factory-wise the optimal distribution
policy for each state. That is, Report 2 gives the optimal quantity of goods to be shipped
from each factory to warehouses, which are optimally decided for each state. A sample
part of ‘Report 2’ related to the factory located at Bangalore is shown in Table 4.
Table 4 Optimal distribution policy for the factory located at Bangalore
Factory State Warehouse Optimal quantity Optimal FSsum Optimal Fsum
W03 2287
APR W05 2548 4835
GOA W01 647 647
JKD W02 67 67
KAR W01 5471 5471
KER W02 1902 1902
MAH W04 5248 5248
BLR
TND W01 3830 3830
22,000
FSsum: Total quantity shipped from factory ‘F’ to state ‘S’;
Fsum: Total quantity to be produced at factory ‘F’.
‘Report 3’ indicates another operational decision on state-wise the optimal distribution
policy between each of the warehouses to a set of distributors. That is, Report 3 gives the
quantity of goods to be shipped from each of the warehouses from each of the states to a
set of distributors. A sample part of ‘Report 3’ related to the state: Andhra Pradesh is
shown in Table 5.
Table 5 Optimal distribution policy for the warehouse located at the state: Andhra Pradesh
State Warehouse Distributor Optimal quantity Optimal Ssum
D01 388
D02 370
D03 688
W03
D04 841
D05 120
D06 965
D07 525
APR
W05
D08 938
4835
Ssum: Total quantity distributed from state ‘S’.
5.3.2 Breakeven analysis
As the existing cost of distribution for each state is not available, a new LINGO set code
is developed to generate data for the existing cost of distribution. It is assumed that the
products are shipped to distributors directly from the factory, without any warehouse in
between. CST was added appropriately for the trade. The existing cost of distribution
for each state thus obtained is shown in Appendix 3.
124 M. Mathirajan et al.
The existing cost of distribution and new proposed cost of distribution as obtained
from FLPDP model is compared and breakeven period is calculated using both the
formula discussed in Section 4.2 and the result has been summarised in Table 6. The rate
of interest for formula (13) is assumed to be 7%, which is cost of borrowing for most of
the big companies.
Table 6 Breakeven Period and optimal decision (to opt for new warehouse in the state)
for accepting the new proposed warehouse
Without time value of money With time value of money
State code
Breakeven
Period N
(months)
Open
warehouse?
(Yes/No)
Breakeven
Period Nr
(months)
Open
warehouse?
(Yes/No)
APR 1.5 Yes 1.6 Yes
ARP 117.4 204.9
ASM 4.5 Yes 4.7 Yes
BIH 2.7 Yes 2.8 Yes
CHT 3.5 Yes 3.7 Yes
GOA 3.8 Yes 3.9 Yes
GUJ 2.8 Yes 2.9 Yes
HAR 2.2 Yes 2.2 Yes
HPR 2.3 Yes 2.4 Yes
JNK 5.5 Yes 5.8 Yes
JKD 23.5 26.1
KAR 0.8 Yes 0.8 Yes
KER 1.4 Yes 1.4 Yes
MPR 1.1 Yes 1.2 Yes
MAH 1.2 Yes 1.3 Yes
MEG 12.1 12.9
MIZ 7 7.4
NAG 6.8 7.2
ORI 2 Yes 2.1 Yes
PUN 1.5 Yes 1.6 Yes
RAJ 1 Yes 1.1 Yes
SKM 22.2 24.6
TND 1.2 Yes 1.2 Yes
TRI 2.8 Yes 2.9 Yes
UPR 1.4 Yes 1.4 Yes
UTT 6.1 6.4
WBN 2.6 Yes 2.7 Yes
DEL 1.7 Yes 1.8 Yes
Decision = ‘Yes’ (for N, Nr 6 months): 6 month is assumed to be the managerial
decision.
A design of distribution network and development 125
Assuming that the demand data generated for the case being yearly, the breakeven period
is calculated and then converted in months. The decision to go for new proposed
warehouse locations was taken based on this period. For a state, if the opening of new
proposed warehouses is breaking even in less than or equal to six months, then we decide
to go for the opening warehouses in that state.
6 Conclusion
The problem addressed in this study is of considerable interest to operations managers
because of rapidly changing environment in which they are operating. Moreover, changes
in the tax policies by government affect the bottom-line costs of the companies. This
encourages operations managers to re-look at their distribution network and reconfigure
its design to cut their distribution costs to achieve competitive advantage. In the literature
review, we have not observed this type of reconfiguration of supply chain distribution
network with inclusion of CST.
The problem was taken as a two-stage distribution network design problem. First,
a mathematical model is developed to solve the facility location and production
distribution policy for a company by minimising the cost comprising production costs at
factory, fixed and labour cost of new warehouses and shipment cost from factories to
warehouses and warehouses to distributors. Second, a breakeven analysis procedure is
proposed for carrying out to compare the new proposed network with the existing
network based on the distribution cost. The following methods were suggested to find
breakeven period namely, (i) without considering the time value of money, (ii)
considering time value of money.
A pseudocase is developed based on the observation made in the large consumer
electronics company located in South India to demonstrate the workability of proposed
solution framework. The case study problem is solved in two stages as proposed in the
solution framework. We also developed a LINGO set code, which generate the required
ILP model as in Section 4 for any given data, and a Report Generator Programme using
C++, which converts the LINGO output (non-zero decision variables) into three easily
interpretable reports for warehouse selection, distribution policy for factories and
distribution policy for states.
ILP model proposed for stage 1 of the decision problem, The LINGO set code
developed and presented in the appendix for the proposed ILP model, the Breakeven
Analysis procedure proposed for the second stage of the decision problem, and the
Report Generator Programme developed are very simple to implement and to solve any
large-scale problems, similar to the pseudocase study presented here.
There are few limitations experienced by us during the study. One major limitation is,
there is no complete data available for the real case problem, so we developed a
pseudodata and termed our case study as pseudocase study. Although we tried to generate
data meaningfully and close to reality, it would have given more satisfaction if the real
data was available. We also did not consider the effect of inventory in developing
the model because the complexity would have increased drastically if inventory issues are
introduced in the model design. The increase in number of warehouses will increase
126 M. Mathirajan et al.
the total inventory for the company, which will affect the overall benefit gained. This can
also be taken as the future work to include inventory issues along with other issues.
Another extension of this study can be taken as incorporating the breakeven period
analysis in the mathematical model itself and convert the solution into single stage
solution. Lastly, the model proposed in this study does not involve routing decisions for
serving the warehouses from the plants and from the warehouses to distributors and this
issue can be addressed following TSP modelling process as third phase in this study.
Acknowledgements
The authors are most grateful to the referees for their valuable comments and suggestions
that helped to improve the presentation of the paper considerably.
References
Aikens, C.H. (1985) ‘Facility location models for distribution planning’, European Journal of
Operational Research, Vol. 22, pp.263–279.
Ambrosino, D. and Scutella, M.G. (2005) ‘Distribution network design: new problems and related
models’, European Journal of Operations Research, Vol. 165, pp.610–624.
Amiri, A. (2006) ‘Designing a distribution network in a supply chain system: formulation and
efficient solution procedure’, European Journal of Operational Research, Vol. 171,
pp.567–576.
Avittathur, B., Shah, J. and Gupta, O.K. (2003) ‘Distribution centre location modeling for
differential sales tax structure’, European Journal of Operational Research, Vol. 162,
pp.191–205.
Ballou, R. (1984) ‘DISPLAN: a multiproduct plant/warehouse location model with nonlinear
inventory costs’, Journal of Operations Management, Vol. 5, No. 1, pp.75–90.
Bhutta, K.S. (2004) ‘International facility location decisions: a review of the modelling literature’,
Int. J. Integrated Supply Management, Vol. 1, No. 1, pp.33–50.
Bidhandi, H.M., Yusuff, R.M., Ahmad, M.M.H.M. and Bakar, A.R.A. (2009) ‘Development of a
new approach for deterministic supply chain network design’, European Journal of
Operations Research, Vol. 198, pp.121–128.
Chopra, S. (2003) ‘Designing the distribution network in a supply chain’, Transportation Research
Part E, Vol. 39, pp.123–140.
Gebgennini, E., Gamberini, R. and Manzini, R. (2009) ‘An integrated production-distribution
model for the dynamic location and allocation problem with safety stock optimization’,
International Journal of Production Economics, Vol. 122, No. 1, pp.286–304.
Gunasekaran, A. and Ngai, E.W.T. (2004) ‘Virtual supply-chain management’, Production
Planning and Control, Vol. 15, No. 6, pp.584–595.
Hanna, M.D., Newman, W.R. and Sridharan, S.V. (1993) ‘Adapting traditional breakeven analysis
to modern production economics: simultaneously modeling economies of scale and scope’,
International Journal of Production Economics, Vol. 29, No. 2, pp.187–201.
Jayaraman, V. (1998) ‘Transportation, facility location and inventory issues in distribution network
design’, International Journal of Operations and Production Management, Vol. 18,
pp.471–494.
A design of distribution network and development 127
Klose, A. and Drexl, A. (2005) ‘Facility location models for distribution systems design’, European
Journal of Operational Research, Vol. 162, pp.4–29.
Lashine, S.H., Fattouh, M. and Issa, A. (2006) ‘Location/allocation and routing decisions in supply
chain network design’, Journal of Modelling in Management, Vol. 1, pp.173–183.
Levén, E. and Segerstedt, A. (2004) ‘Polarica’s wild berries: an example of a required storage
capacity calculation and where to locate this inventory’, Supply Chain Management, Vol. 9,
pp.213–218.
Liu, Y. and Hao, G. (2006) ‘Design of optimization model of the distribution network
oriented to the customer’, International Journal of Business and Management, Vol. 1, No. 5,
pp.36–42.
Melkote, S. and Daskin, M.S. (2001) ‘Capacitated facility location/network design problems’,
European Journal of Operational Research, Vol. 129, pp.481–495.
Melo, M.T., Nickel, S. and Saldanha-da-Gama, F. (2009) ‘Facility location and supply chain
management – a review’, European Journal of Operational Research, Vol. 196, pp.401–412.
Miranda, P.A. and Garrido, R.A. (2006) ‘A simultaneous inventory control and facility location
model with stochastic capacity constraints’, Networks and Spatial Economics, Vol. 6,
pp.39–53.
Murat, A., Verter, V. and Laporte, G. (2010) ‘A continuous analysis framework for the solution of
location–allocation problems with dense demand’, Computers and Operations Research, Vol.
37, No. 1, pp.123–136.
Nagy, G. and Salhi, S. (2007) ‘Location-routing: issues, models and methods’, European Journal
of Operational Research, Vol. 177, No. 2, pp.649–672.
Norton Jr., J.W. and Weber Jr., W.J. (2006) ‘Breakeven costs for distributed advanced technology
water-treatment systems’, Water Research, Vol. 40, No. 19, pp.3541–3550.
Ohlemuller, M. (1997) ‘Tabu search for large location-allocation problems’, Journal of the
Operational Research Society, Vol. 48, pp.745–750.
Owen, S.H. and Daskin, M.S. (1998) ‘Strategic facility location: a review’, European Journal of
Operational Research, Vol. 111, pp.423–447.
Perl, J. and Sirisoponslip, S. (1988) ‘Distribution network: facility location, transportation and
inventory’, International Journal of Physical Distribution and Materials Management,
Vol. 18, No. 6, pp.18–26.
Rentizelas, A. and Tatsiopoulos, I.P. (2010) ‘Locating a bioenergy facility using a hybrid
optimization method’, International Journal of Production Economics, Vol. 123, No. 1,
pp.196–209.
Sahin, G. and Sural, H. (2007) ‘A review of hierarchical facility location models’, Computer and
Operations Research, Vol. 34, pp.2310–2311.
Shen, Z-J. and Qi, L. (2007) ‘Incorporating inventory and routing costs in strategic location
models’, European Journal of Operational Research, Vol. 179, pp.372–389.
Tsiakisa, P. and Papageorgiou, L.G. (2008) ‘Optimal production allocation and distribution supply
chain networks’, International Journal of Production Economics, Vol. 111, pp.468–483.
Wang, Z., Yao, D-Q. and Huang, P. (2007) ‘A new location-inventory policy with reverse logistics
applied to B2C e-markets of China’, International Journal of Production Economics,
Vol. 107, pp.350–363.
Wu, T-H., Low, S. and Bai, J-W. (2002) ‘Heuristic solutions to multi-depot location routing
problems’, Computers and Operations Research, Vol. 29, pp.1393–1415.
Wu, W-Y., Bai, C-J. and Gupta, O.K. (2006) ‘A hypermarket site selection model using the grey
multi-objective decision method’, Int. J. Logistics Systems and Management, Vol. 2, No. 1,
pp.68–77.
128 M. Mathirajan et al.
Appendix 1: A LINGO set code for FLPDP
SETS:
STATE: R_WH, P_WH, DBUTOR;
WAREHOUSE;
DISTRIBUTOR;
FACTORY: C_F, PC_F;
STATE_WH( STATE, WAREHOUSE): FC_SW, LC_SW, Y_SW;
STATE_DBUTOR( STATE, DISTRIBUTOR): N_SD;
FACTORY_WH(STATE_WH, FACTORY): X_SWF, TC_SWF;
WHOUS_DBUTOR(STATE, WAREHOUSE, DISTRIBUTOR): X_SWD,
TC_SWD;
ENDSETS
DATA:
!Data input through Excel Sheet;
STATE = @OLE('E:\PROJECT\DATA.XLS');
WAREHOUSE = @OLE('E:\PROJECT\DATA.XLS');
DISTRIBUTOR = @OLE('E:\PROJECT\DATA.XLS');
R_WH = @OLE('E:\PROJECT\DATA.XLS');
P_WH = @OLE('E:\PROJECT\DATA.XLS');
DBUTOR = @OLE('E:\PROJECT\DATA.XLS');
FACTORY = @OLE('E:\PROJECT\DATA.XLS');
N_SD = @OLE('E:\PROJECT\DATA.XLS');
FC_SW = @OLE('E:\PROJECT\DATA.XLS');
PC_F = @OLE('E:\PROJECT\DATA.XLS');
C_F = @OLE('E:\PROJECT\DATA.XLS');
TC_SWF = @OLE('E:\PROJECT\DATA.XLS');
TC_SWD = @OLE('E:\PROJECT\DATA.XLS');
LC_SW = @OLE('E:\PROJECT\DATA.XLS');
CAPACITY = 50000;
!Exporting Result to Excel Sheet;
@OLE('E:\PROJECT\RESULT.XLS') = Y_SW, X_SWF, X_SWD;
ENDDATA
A design of distribution network and development 129
!Defining Variable types;
@FOR(STATE_WH:@BIN(Y_SW));
@FOR(WHOUS_DBUTOR:@GIN(X_SWD));
@FOR(FACTORY_WH:@GIN(X_SWF));
!Objectve Function: Minimising Total cost;
MIN = @SUM(FACTORY_WH(j, k, i)| k #LE# P_WH(j): (PC_F(i) +
TC_SWF (j, k, i))* X_SWF (j, k, i)) + @SUM(STATE_WH(j, k) |
k #LE# P_WH(j): FC_SW(j, k) * Y_SW(j, k)) +
@SUM(WHOUS_DBUTOR(j, k, m) | m #LE# DBUTOR(j): (LC_SW(j, k)
+ TC_SWD (j, k, m)) * X_SWD(j, k, m));
!Constraint Set (2): Capacity constraints at Factory;
@FOR(FACTORY(i):
@SUM(FACTORY_WH(j, k, i)| k #LE# P_WH(j): X_SWF(j, k,
i)) <= C_F(i));
!Constraint Set (3): Relation between X and Y Decision
Variables;
@FOR(STATE_WH(j, k):
@SUM( WHOUS_DBUTOR(j, k, m) | m #LE# DBUTOR(j):
X_SWD(j, k, m)) <= CAPACITY * Y_SW(j, k));
!Constraint Set (4): Flow Balance at Warehouses;
@FOR(STATE_WH(j, k) | k #LE# P_WH(j):
@SUM( FACTORY_WH(j, k, i): X_SWF(j, k, i)) =
@SUM(WHOUS_DBUTOR(j, k, m) | m #LE# DBUTOR(j): X_SWD(j,
k, m)));
!Constraint Set (5): Demand Constraints at Distributors;
@FOR(STATE_DBUTOR(j, m)| m #LE# DBUTOR(j):
@SUM( WHOUS_DBUTOR (j, k, m) | k #LE# P_WH(j): X_SWD(j,
k, m)) = N_SD(j, m));
!Constraint Set (6): Required no of warehouses for each
state;
@FOR(STATE(j):
@SUM(STATE_WH(j, k)| k #LE# P_WH(j): Y_SW(j, k)) =
R_WH(j));
@FOR(STATE(j):
@SUM(STATE_WH(j, k)| k #GT# P_WH(j): Y_SW(j, k)) = 0);
!Non- negativity Constraints for Decision Variables;
@FOR(STATE_WH: Y_SW >= 0);
130 M. Mathirajan et al.
@FOR(WHOUS_DBUTOR: X_SWD >= 0);
@FOR(FACTORY_WH: X_SWF >= 0);
END
Appendix 2: Additional data for the case study
N_SD, Demand data for state distributors based on U(50, 1050)
State N_SD
APR 388 370 688 841 120 965 525 938 709 218
ARP 441 172 385 644 484 404 437 297 866 776
ASM 902 976 200 903 956 77 962 61 124 157
BIH 813 806 1003 534 593 600 229 77 837 101
CHT 1007 353 313 102 475 762 787 600 821 375
GOA 343 304 584 528 519 255 132 762 184 649
GUJ 550 337 311 144 60 884 492 255 566 247
HAR 464 426 918 616 472 1037 96 884 516 344
HPR 525 762 612 445 952 957 783 1037 752 719
JNK 766 273 101 95 900 744 81 957 504 924
JKD 108 140 105 239 57 840 711 744 549 564
KAR 255 62 283 750 800 616 197 840 981 687
KER 558 820 735 290 234 994 576 616 272 863
MPR 129 793 579 910 125 828 474 994 706 656
MAH 710 757 620 521 290 615 723 828 620 757
MEG 649 656 706 440 226 480 466 615 579 793
MIZ 210 863 272 810 173 868 389 480 735 820
NAG 357 687 981 875 246 942 327 868 283 62
ORI 207 564 549 798 617 419 934 942 105 140
PUN 847 924 504 653 381 894 889 419 101 273
RAJ 362 719 752 664 621 1005 95 894 612 762
SKM 92 344 516 198 457 453 565 1005 918 426
TND 790 247 566 496 428 451 88 453 311 337
TRI 488 649 184 1000 71 660 926 451 584 304
UPR 222 375 821 380 381 828 800 660 313 353
UTT 745 101 837 105 915 205 358 828 1003 806
WBN 679 164 124 900 960 520 915 205 200 976
DEL 510 453 866 404 484 644 385 520 101 337
A design of distribution network and development 131
Appendix 2: Additional data for the case study (continued)
FC_SW: Fixed opening cost of state warehouses based on U(30,000, 90,000);
LC_SW: Labour cost (operating cost) of state warehouses based on U(25, 75)
State FC_SW LC_SW
APR 79,987 86,170 58,057 56,551 41,664 33 52 42 52 28
ARP 67,968 72,860 69,290 76,224 37,997 56 54 70 39 50
ASM 61,101 64,768 84,534 83,408 39,616 29 52 36 57 53
BIH 62,309 79,653 52,332 52,123 73,693 44 32 45 60 43
CHT 47,650 77,668 32,522 63,223 39,870 51 64 58 60 34
GOA 35,724 55,681 37,545 75,207 60,726 27 57 47 42 73
GUJ 41,989 85,660 30,088 56,425 55,415 59 53 28 26 73
HAR 80,804 86,174 83,592 53,495 83,989 49 44 41 69 27
HPR 69,562 57,401 45,422 52,102 77,215 71 49 32 61 73
JNK 58,649 60,262 66,669 79,476 77,348 71 47 66 74 67
JKD 81,666 84,766 84,408 50,609 63,474 45 61 74 51 71
KAR 38,476 33,828 82,310 77,425 82,491 27 69 59 46 49
KER 79,422 43,236 68,782 64,278 70,316 34 31 25 39 74
MPR 78,662 65,247 52,905 33,409 40,850 53 65 56 68 63
MAH 62,240 32,136 70,738 44,678 72,369 47 61 44 56 73
MEG 67,582 70,604 53,662 55,513 45,199 63 47 48 26 64
MIZ 81,708 78,338 37,523 57,814 89,414 61 32 49 47 62
NAG 37,918 45,335 63,214 49,394 52,068 62 29 67 68 67
ORI 87,394 72,998 37,797 45,232 74,055 38 41 40 27 55
PUN 34,686 40,317 69,257 68,239 64,111 57 48 36 67 61
RAJ 36,384 30,431 69,855 62,512 43,020 25 28 70 70 45
SKM 64,969 58,488 42,717 73,930 82,660 59 42 34 60 53
TND 38,321 62,762 82,220 48,064 32,709 34 71 29 59 59
TRI 41,626 85,225 85,386 84,390 40,077 39 64 43 39 27
UPR 37,020 45,977 35,273 56,420 49,708 49 26 75 38 37
UTT 84,052 63,951 77,702 69,384 50,548 68 26 38 41 33
WBN 62,017 53,320 65,244 53,233 38,675 28 54 32 40 65
DEL 76,131 54,255 62,619 74,802 46,953 39 69 67 63 58
132 M. Mathirajan et al.
Appendix 2: Additional data for the case study (continued)
TC_SWF: Unit transportation cost of shipment from factories to state warehouses based
on inter-city distances
A design of distribution network and development 133
Appendix 2: Additional data for the case study (continued)
TC_SWF: Unit transportation cost of shipment from factories to state warehouses
based on inter-city distances (continued)
134 M. Mathirajan et al.
Appendix 2: Additional data for the case study (continued)
TC_SWD: Unit Transportation cost of shipment from state warehouse to distributors
based on intercity distances
A design of distribution network and development 135
Appendix 2: Additional data for the case study (continued)
TC_SWD: Unit Transportation cost of shipment from state warehouse to distributors
based on intercity distances (continued)
136 M. Mathirajan et al.
Appendix 2: Additional data for the case study (continued)
TC_SWD: Unit Transportation cost of shipment from state warehouse to distributors
based on intercity distances (continued)
A design of distribution network and development 137
Appendix 3: Existing cost of distribution for the case study data
S. No. State State
Total transportation cost
(from factory to distributor) CST incurred on trade
1 Andhra Pradesh APR 91011 1087875
2 Arunachal Pradesh ARP 1863 137925
3 Assam ASM 41213 467550
4 Bihar BIH 43740 1047375
5 Chattisgarh CHT 21419 376425
6 Goa GOA 13109 145575
7 Gujarat GUJ 52382 809775
8 Haryana HAR 9013 651600
9 Himachal Pradesh HPR 1412 427275
10 Jammu and Kashmir JNK 5559 233775
11 Jharkhand JKD 7663 133200
12 Karnataka KAR 114606 1230975
13 Kerala KER 73047 593325
14 Madhya Pradesh MPR 25816 756900
15 Maharashtra MAH 97908 1449225
16 Meghalaya MEG 33927 293625
17 Mizoram MIZ 2673 241425
18 Nagaland NAG 26216 234900
19 Orissa ORI 51058 709650
20 Punjab PUN 9657 744525
21 Rajasthan RAJ 43104 1150200
22 Sikkim SKM 11092 98100
23 Tamil Nadu TND 97480 861750
24 Tripura TRI 20728 255825
25 Uttar Pradesh UPR 20766 1154925
26 Uttaranchal UTT 1090 378675
27 West Bengal WBN 52759 958950
28 Delhi DEL 15178 1058400
... According to Balaji and Viswanadham (2008), tax issues management is important to decision making in network planning, because the integration of fiscal policy can result in a more cost-effective supply chain. When decision-makers take into account the effects of taxes while designing a distribution network, the tax burden can be reduced (Mathirajan et al., 2011). Furthermore, in some countries, taxes are imposed on interstate and intrastate cargo transportation with differentiated rates, even among facilities of the same company (Avittathur et al., 2005). ...
... In some countries, taxes are imposed on interstate and intrastate cargo transportation with differentiated rates. Avittathur et al. (2005) and Mathirajan et al. (2011) addressed the location and distribution problem in India, where the Central Sales Tax (CST) has an interstate character. These taxes play a role of a strategic cost component in the distribution network design because depending on the location of a facility, one can pay less taxes due to differentiated rates of origin-destination locations. ...
Article
Full-text available
This paper focuses on a medium-term distribution plan for a pharmaceutical network, considering distribution center (DC) location and transportation decisions. The problem is characterized by low DC opening costs due to outsourced warehousing services. Therefore, transportation costs have a higher impact on location decisions and should be carefully evaluated. In addition, medicines vary widely in sales price and volume, demand behavior, and physical characteristics, which affects the distribution planning. This paper proposes a multi-product, multi-period, and multi-modal mathematical model integrating network design and distribution planning decisions, such as product flow, transportation modes, type of freight shipping, fleet sizing, and escorting services for high-value cargo. The mathematical formulation takes into consideration realistic features such as value-added tax, whose rate varies among locations. A Local Branching matheuristic taking advantage of the hierarchical problem structure is introduced. The result of tests using real-life data demonstrate the model’s effectiveness to support the trade-off between the location and transportation problem, corroborating the relevance of making these decisions simultaneously. A comparison of the results with a company shows opportunities to reduce overall logistics costs plus the total amount of chargeable taxes. Regarding computational running times, the proposed matheuristic outperformed CPLEX in 75% of instances.
... Several supply chain network models have been developed by previous researchers. Yuniaristanto et al. (2010) developed a supply chain network model but did not consider tax incentives, while Mathirajan et al. (2011) have considered tax incentives in their model. Several previous research models have not included aspects of facility opening in their models (Song and Zhao, 2015;Lv and Sun, 2016). ...
Conference Paper
Full-text available
Most people are now concerned about environmental issues. As the number of motorized vehicles grows, so does the consumption of fossil fuels. As a result, CO2 levels in the atmosphere rise. Starting to use alternate energy sources, such as electric automobiles, is one approach that can be implemented. By replacing the engine parts with a conversion kit, conventional motorcycles can be converted to electric motorcycles. In this study, the model is being developed by adding parameters. The cost of ordering raw supplies, storage expenses, transportation costs, and tax benefits are all factored into data processing. By using a mixed integer linear programming (MILP) model, this research offers a mathematical model for deciding which electric motorcycle conversion kit business to open. The results demonstrate that the model can be used as a decision-making tool for starting an electric motorbike conversion kit manufacturer due to tax reductions, and then starting as a manufacturer is more effective than becoming a trader.
... Other models are more robust and can include intangible factors in the location decision. A number of authors (Hamidi, Farahmand, Sajjadi, & Nygard, 2012;Kumar, et al., 2011;Mathirajan, et al., 2011) have suggested such modeling can lead to optimality in location selection. Selecting a location with the best qualitative conditions may prove more profitable than simply choosing the lowest cost location, an approach found in many transportation models. ...
Chapter
Using a qualitative business case-study method, this chapter examines how two Pittsburgh-headquartered firms (i.e., a global manufacturing and a domestic service restaurant chain) used location strategies to take advantage of availability of skilled labor, transportation facilities, tax rates, regulatory environment, real estate tax, and other considerations to expand their operations. Supplier considerations and associated strategies also were essential to the firms' successful expansion plans.
... warehouse location, vehicle routing and legal taxes respectively (Prasad and Shankar, 2011;Nidhi and Anil, 2011;Mathirajan et al. 2011). A two-warehouse inventory model with First in, First out (FIFO) and Last in, First out (LIFO) dispatching policies has been proposed considering partial backlogging rate in order to optimise inventory costs (Jaggi et al. 2013). ...
Article
Full-text available
This paper comprehensively discusses the existing state-of-the-art warehousing literature and highlights concerned research issues as per the proposed taxonomy. All previous reviews broadly focus on warehouse design and operational issues ignoring performance measures affecting each function of warehousing and overall productivity. Therefore, this paper tries to identify the specific performance measures and methods, and explore their impact on the overall logistics system. Each individual section has been concluded with important findings and these are summarised along with converged research guidelines in the last section. The proposed gaps would provide a future road map for research in existing and other unexplored directions in warehouse operations and management.
Chapter
Offshore outsourcing basically describes the practice of contracting to outside vendors in another country, especially in cases where the client company has no direct ownership. This operational strategy of low-cost, global expansion and the increased capability of vendors in India, China, the Philippines, South Korea has opened a great number of avenues of traditional organizational functions of IT. Such functions as software development, call centers, and accounting are typically offshored. Since the late 1990s, offshore outsourcing has included more sophisticated finance and accounting functions, beyond that of simple data input and transactions. Offshore finance and accounting outsourcing is especially becoming an attractive option for many companies. The obvious benefits are gaining access to scarce and valuable skills, cutting costs, and domestic and global achieving competitiveness. However, there are risks as well. A relatively balanced approach of benefits and risks are discussed in this chapter.
Chapter
This chapter details the growth of Samsung Electronics Company in its innovative use of automatic identification and data capture technologies in managing supply chain relationships meet goals of operational efficiency and profitably. Through a combination of web and business literature sources and personal interviews, documentation of the company's successfully implemented supply chain management (SCM) applications as well as technological advancements and development are found in the chapter. SCM systems and IT applications documented at Samsung are designed to help provide assistance to other companies dealing with similar issues.
Article
Full-text available
Original Equipment Manufacturers (OEM) that assemble multiple durable products are actively engaged in adopting remanufacturing practices. This study addresses the introduction of a new business proposal, which is carrying out remanufacturing of components from acquired returns by an OEM through the set-up of required Reverse Flow Capacities (RFC), which are capacities for dismantling the returns and capacities for remanufacturing various components. Hence, by incorporating the activity on remanufacturing of components, an OEM creates a new cheaper source for obtaining components, in addition to the conventional sources of manufacturing and/or purchasing. The components obtained from any of these three sources can be used to assemble the multiple durable products for satisfying demands. The study refers to this problem as “Multi Products–Component Remanufacturing (MP-CR) problem”. Furthermore, it is identified that the incorporation of MP-CR problem has tremendous potential in the Indian automobile sector. With these, this study address two strategic issues and an operation issue for the incorporation of new MP-CR problem to an OEM. The first strategic issue is the determination of RFC that must be set-up. This strategic issue is addressed in synchronize with the operational issue of determining the optimal Inventory & Production Plans (I&PP) for assembling multiple products for a given planning horizon. In addition, once the optimal RFC is determined, the second strategic issue is the determination of optimal period (called as Breakeven Period (BEP)), after which the OEM would reclaim the capital investment for setting-up the optimal RFC. In order to address these three interrelated decisions, a systematic methodology is proposed. In the proposed methodology, first an Integer Linear Programming Model is developed for the existing system of the OEM to determine the optimal I&PP. Next, this model is extended to determine the optimal RFC and corresponding optimal I&PP. The solution from both proposed models are used for obtaining the Breakeven Period (BEP), which indicates when the capital investments breakeven. Finally, the proposed methodology is demonstrated by developing suitable data observed from an Indian Automobile OEM.
Chapter
To make products and processes sustainable, Original Equipment Manufacturers (OEM) adopt remanufacturing practices. Component Remanufacturing (CR) problems arise for any OEM that produce durable products and acquires returns. Returns are of three categories: Commercial Returns, End-of-Use Returns and End-of-Life Returns for remanufacturing components, which are equivalent in all aspects as compared to the manufactured and/or purchased components. The final product assembled using manufactured and/or purchased and/or remanufactured components by OEM are also equivalent in all aspects to satisfy the demand. The remanufacturing of components from the return will be carried out by installing required capacities for both dismantling the returns and remanufacturing the various components (called together as Reverse Flow Capacities (RFC)). Moreover, the final products’ demand can be satisfied by the same product and/or by Backordering (BO) and/or by Product Substitution (PS). With these problem characteristics of CR, this study focuses on three research problems. Accordingly, the first problem of focus is on the “Multi-Products-CR (MP-CR) system with RFC&Break-Even Period problem”, where the strategic decisions of RFC installations and the break-even period for the capital investments are determined. In the second problem, demand management by BO along with the real-life procurement restrictions are incorporated into the MP-CR System and are termed as “MP-CR System with BO problem”. Finally, the additional and important problem characteristics on PS mechanism are incorporated in the third problem, and this is the “MP-CR System with BO&PS problem”. A systematic 5-step methodology, in which the MP-CR system is modelled as an Integer Linear Programme, is proposed to address the first problem. For the second and third research problems, Integer Non-Linear Programming models are proposed. All the proposed models determine the optimal inventory and production decisions for the MP-CR system. A suitable numerical example, based on data from a real-life Indian Automobile OEM, is developed for representing each of the three research problems and for demonstrating the workability of the proposed models. Finally, managerial insights are presented from the solutions obtained.
Chapter
The nature of SCM research is constantly evolving and must address a variety of concerns like poor service, large inventory levels, and friction among suppliers and manufacturers. Analytical databases and techniques in SCM are an important part of this research. Many researchers and practitioners have depended on secondary data, but given the dynamic nature of global competition, more recent and relevant data must be gathered. These efforts need to be geared to the development of properly managed supply chain relationships and corporate sustainability initiatives that ultimately promote broad-based sustainable development objectives for the good of people, plants, and profits (i.e., triple bottom-line).
Article
Full-text available
Recently, international facilities location problems have attracted a great deal of attention in literature. Consequently, there has been an explosive growth in analytical studies in the field of facility location and, more specifically in global facility location. This paper provides a review of the existing literature with respect to factors involved in global trade and develops a taxonomy of existing analytical models. Some avenues of further research are also suggested.
Article
Full-text available
Polarica is a company in northern Europe buying, refining and selling wild berries and other specialty foods. During the last couple of years the volume of wild berries, mostly blueberries, has increased a lot. This expansion forces Polarica to consider investments in freezing-in capacity and cold-storage capacity. A simple heuristic model was constructed, from which it was concluded that additional volumes of frozen blueberries require more storage facilities and it was also recommended that the location should be based on load-distance analysis. The way this problem is tackled and solved can be copied and hopefully it presents ideas for other similar studies.
Article
Recently it has been demonstrated that the use of simulated annealing is a good alternative for solving the minisum location-allocation problem with rectilinear distances compared with other popular methods. In this study it is shown that the same solution quality and a great saving of computational time can be achieved by using tabu search. It is also possible to transfer this method to location-allocation problems with euclidean distances.
Article
As the retailing market has become more and more competitive, the selection of an attractive location plays a critical role in the performance of retail outlets. Due to their huge investment and operation costs, potential locations must go through a series of careful evaluations before deciding on the final site. Previous studies on the evaluation of site selection have adopted the multiplicative competitive interaction (MCI) model to decide the most preferred locations. The MCI model has been critiqued as relying too much on the opinions of potential customers and might need to be offset by the opinions of marketing experts. This article presents an alternative evaluation procedure to help retailers, especially hypermarketers, make a location decision by using the grey multi-objective decision method. The methodology is applied to a leading marketer in Taiwan and the results indicate that this model provides a significantly different approach from the MCI model.
Article
DISPLAN is a large-scale computer model used for the strategic planning of physical supply and distribution networks. This planning may involve the determination of the number, size, and location of plants, warehouses, inventories, terminals, and like facilities. Although a number of mathematical models exist to treat this problem, a unique feature of DISPLAN is its handling of nonlinear inventory costs.The proper economic analysis of network designs requires that all relevant costs be balanced to achieve the minimum cost configuration of facilities. These costs include facility inbound and outbound transportation costs, warehouse storage and handling costs, order processing costs, facility fixed costs, and inventory carrying costs. Among these costs, it is vital that inventory carrying costs be accurately described since they are a major factor determining the number of facilities in a network. Due to the limitations of many solution methods, inventory carrying costs are neglected or they are treated as a linear function of the number of facilities in the network. However, if a company's inventory policy is based on the concept of the economic order quantity, then the relationship between the number of facilities and the total inventory in the network is nonlinear, commonly referred to as the inventory consolidation effect. Inaccurate representation of this effect through linear approximation can lead to an incorrect number of facilities in the network.The methodology of DISPLAN is a heuristic procedure that uses the 3-dimensional transportation algorithm of linear programming in an iterative fashion to converge on the minimum cost network configuration subject to facility capacity and customer service constraints. Warehouse fixed costs and inventory costs are recomputed on a per-unit basis after each iteration, once the warehouse throughput has been established. These per-unit costs can be added to the per-unit transportation, handling, and order processing costs of the linear programming algorithm. The revised problem is resolved. The process is repeated until the best number, location, and size of warehouses are determined.Because of the nonlinear inventory cost curve, the procedure may terminate at a local optimum on the total cost curve. The stopping rule that terminates the procedure when the change in total costs between successive iterations is less than a given percentage is not absolute. Additional iterations may be computed to help assure that the global optimum is realized.Finally, a reconsideration routine is included in the computational procedure for those cases where fixed costs are a high proportion of total costs. The main heuristic may leave too many warehouses in the network and this routine explores a reduced number of warehouses. This assures that a substantial cost-saving opportunity is not overlooked.The DISPLAN methodology has been applied to many network configuration problems in a variety of industries. Typical of these are retailing, manufacturing, and spare parts distribution. The method has shown modest computer running times that are linear with problem size.
Article
This paper proposes a mixed integer linear programming model and solution algorithm for solving supply chain network design problems in deterministic, multi-commodity, single-period contexts. The strategic level of supply chain planning and tactical level planning of supply chain are aggregated to propose an integrated model. The model integrates location and capacity choices for suppliers, plants and warehouses selection, product range assignment and production flows. The open-or-close decisions for the facilities are binary decision variables and the production and transportation flow decisions are continuous decision variables. Consequently, this problem is a binary mixed integer linear programming problem. In this paper, a modified version of Benders' decomposition is proposed to solve the model. The most difficulty associated with the Benders' decomposition is the solution of master problem, as in many real-life problems the model will be NP-hard and very time consuming. In the proposed procedure, the master problem will be developed using the surrogate constraints. We show that the main constraints of the master problem can be replaced by the strongest surrogate constraint. The generated problem with the strongest surrogate constraint is a valid relaxation of the main problem. Furthermore, a near-optimal initial solution is generated for a reduction in the number of iterations.
Article
Purpose – The purpose of the paper is to present an integrated model for the location of warehouse, the allocation of retailers to warehouses, and finding the number of vehicles to deliver the demand and the required vehicle routing in order to minimize total transportation costs, fixed and operating costs, and routing costs. Design/methodology/approach – The model assumes that the number of plants has already been determined and answers the following questions: what is the number of warehouses to open? How warehouse are allocated to plants? How retailers are allocated to warehouses? Who are the retailers that will be visited and in what order? How many vehicles are required for each route? What are the total minimum costs? Findings – The model was formulated as a mixed integer linear programming model and solved using Lagrange relaxation and sub‐gradient search for the location/allocation module and a traveling salesman heuristic for the routing module. The results for the randomly selected problems show that the deviation in objective function value ranges between 0.29 and 2.05 percent from the optimum value. Also, from the CPU time point of view, the performance was very good. Originality/value – An attempt is made to integrate location, allocation, and routing decisions in the design of a supply chain network.
An integrated model for distribution network design is proposed which explicitly represents the required level of customer service.
Article
Examines the interdependence between facility location, transportation and inventory issues in a distribution network design problem. Management of inventories, determination of transportation policy, and location of plants and distribution centers are normally carried out by different groups of people in an organization. These activities interact, however, when the transportation is used to replace inventory, an increase in the number of warehouses increases total system inventory or location of warehouses would dictate the type of transportation mode choice or carrier that needs to be used. The proposed model, FLITNET, is expected to provide a more complete representation of the trade-offs that exist among the location, transportation and inventory cost components, and lead to an optimal solution.
Article
Recently it has been demonstrated that the use of simulated annealing is a good alternative for solving the minisum location–allocation problem with rectilinear distances compared with other popular methods. In this study it is shown that the same solution quality and a great saving of computational time can be achieved by using tabu search. It is also possible to transfer this method to location–allocation problems with euclidean distances.