Upload
vandien
View
214
Download
2
Embed Size (px)
Citation preview
A Supply Chain Optimization Model:
Minimizing Transportation and
Inventory Cost
Nairi Nazarian
A thesis submitted in partial fulfillment
of the requirements for the degree of
BCHELOR OF APPLIED SCIENCE
Supervisor: R.H. Kwon
Department of Mechanical and Industrial Engineering
University of Toronto
March, 2007
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
1
Abstract
In this thesis I have studied the three in-stock strategies – flow-through,
regional and single DC central stock - currently used by Hudson’s Bay Company
(Hbc) and thus developed a simple transportation-inventory model in order to
compare their total costs. I have also described a distribution model proposed by
(Berman et al [2006]) in which the model is formulated as a non-linear integer
optimization problem. Due to the non-linearity of the inventory cost in the
objective function, two heuristics and an exact algorithm is proposed in order to
solve the problem.
The results obtained from the transportation-inventory models show that the
single DC and regional central stock strategies are more cost-efficient
respectively compared to the flow-through approach. It is recommended to take
the single DC and the regional central stock strategies for slow moving and
demanding products respectively.
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
2
Acknowledgements
I would like to thank my supervisor, Professor Roy Kwon, for providing me with help
and support especially for his guidance in determining my thesis topic.
I would also like to thank D.A. Power at Hbc’s Supply Chain Optimization for helping
and supporting me throughout the course of the thesis project.
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
3
TABLE OF CONTENTS
Abstract ………………………………………………………………….……………….i Acknowledgements……………………………………………………………………………….….ii Table of Contents………………………………………………………………………………….….iii List of Symbols…………………………………………………………………….……iv
List of Figures……………………………………………………………………….…..x
List of Tables……………………………………………………………………….……xi
1. Motivation…………………………………………………………………....….1
2. Introduction………………………………………………………………..…….2
3. Literature Review…………………………………………………………..……3
Supply Chain Network Design…………………………………………..…..3
Hudson’s Bay Company (Hbc) Inventory Structure and Product Flow….…4
4. Models………….…………………………………………………………..……7
Preliminary Distribution Model…………………..…………………….…...7
Cross-dock and Direct Shipment Models (Berman et al[2006])……….…...12
Distribution Strategies………………………………………….……13
Model Assumptions…………………………………………………15
Formulation………………………………………………………….16
Hbc Transportation and Inventory Model……………………..….………...18
Transportation Model Formulation………….…………………..…..20
Inventory and Transportation Model Formulation…………….…….20
Programming Using OPL……………………………………………21
Data Acquisition……………………………………………………..21
5. Comparison of Models – Results……………………..…………………….……22
6. Future Research..………………………………………………………….……..28
7. Conclusion……………………………………………………………………….28
8. References………………………………………………………………….……29
9. Collections of Figures and Tables………………………………………………..30
Appendix A…………………………………………………………………………..37
Appendix B………………………………………………..…………………………39
Appendix C…………………………………………………………………………..44
Appendix D……………………………………………………………………….….46
Appendix E…………………………………………………………………….…….48
Appendix F…………………………………………………………………………...50
Appendix F-1………………….…………………………..………………….….50
Appendix F-2…………………………………………………………………….53
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
4
List of Symbols
Preliminary model
m = Number of stores or demand points
n = Number of potential DCs
l = Number of suppliers
t = Number of potential cross-dock locations
=jD Weekly/Annual demand from store j
=iK Potential capacity of DC at site i
=hS Supply capacity at supplier h; truckload (TL) and less-than-truckload (LTL)
=eW Potential cross-dock capacity; Dummy Variable
=eF Fixed cost of processing shipment from supplier at cross-dock site e
=eV Variable cost of processing shipment from supplier at cross-dock site e
=iF Fixed cost of investment at DC i
=iV Variable cost of processing shipment from supplier at DC i
=hic Cost of shipping one unit from supplier h to DC i
=hjc Cost of shipping one unit from supplier h to store j
=hec Cost of shipping one unit from supplier h to cross-dock site e
=ejc Cost of shipping one unit from cross-dock site e to store j
=ijc Cost of shipping one unit from DC i to store j
=ic Cost of holding one unit at DC i
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
5
=jc Cost of holding one unit at store j
=iy 1 if DC is located at site i, 0 otherwise
=ey 1 if cross-dock is done at site e, 0 otherwise
=ejx Quantity shipped from cross-dock e to store j
=ijx Quantity shipped from DC at site i to store j
=hix Quantity shipped from supplier h to DC at site i
=hex Quantity shipped from supplier h to cross-dock at site e
=hjx Quantity shipped from supplier h to store j
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
6
Cross-Dock and Direct Shipment Models Suggested by (Berman et al
[2006])
I = Set of Suppliers
J = Set of Plants
P = Set of products
K = Set of cross-docks
C= Truck capacity
=d
ijc Direct transportation cost of shipping one truckload of products from
supplier i to plant j
=i
ikc Inbound transportation cost of shipping one truckload of products from
supplier I to cross-dock k
=o
kjc Outbound transportation cost of shipping of one truckload of products from
cross-dock k to plant j
=d
ijt Direct transportation time (periods) from supplier i to plant j
=k
ikt Inbound transportation time (periods) from supplier i to cross-dock k
=o
kjt Outbound transportation time (periods) from cross-dock k to plant j
=kT Time spent transferring from inbound to outbound at cross-dock k
=pb Truck capacity occupied by one unit of product p
=ph Inventory-carrying cost of one unit of product p per period
F = },0,,,:),,{( >∈∈∈ ijpdPpJjIipji set of flows
},0,,:),,{( >∈∈= ijpdPpIipjiF set of flows to plant j
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
7
},0,:),,{( >∈= ijpij dPppjiF set of flows from supplier i to plant j
}),,(:{ ijij FpjipP ∈=
}),,(:),{( jj FpjipiIP ∈=
=ijpkx {1 if flow (i,j,p) is shipped through cross-dock k; 0 otherwise
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
8
Transportation Model
m = Number of stores or demand points
n = Number of potential DCs (Distribution Centres)
l = Number of suppliers
t = Number of products
=pjD Demand for product p at store j
=iK Potential capacity of DC at site i
=phiS Supply capacity at supplier h to DC i for product p
=pijT Cost of transporting product p from supplier at DC site i to store j
=iC DC Cost of processing shipment from supplier at DC site i;
DC cost includes: Receive – Put away – Pick - Ship
=pijX Number of product p shipped from DC i to store j
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
9
Inventory and Transportation Model
m = Number of stores or demand points
n = Number of potential DCs (Distribution Centres)
l = Number of periods
k = Number of products
=pjtD Demand for product p at store j in period t
=iK Potential capacity of DC at site i
=phiS Supply capacity at supplier h to DC i for product p
=ijT Cost of transporting of one unit of product from supplier at DC site i to
store j
=iC DC Cost of processing shipment from supplier at DC site ij;
DC cost includes: Receive – Put away – Pick - Ship
pI = Inventory cost for product p
pjII =Initial inventory product p at store j
=pijtX Number of product p shipped from DC i to store j at period t
pjtY = Number of product p stored at store j in period t
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
10
List of Figures
Figure 1 Schematic representation of the process that takes
products to get onto stores’ shelves
Figure 2 Distribution strategies (Shapiro [2005])
Figure 3 Distribution strategies (Berman et al [2006])
Figure 4 Product delivery routes
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
11
List of Tables
Table 1 Distribution Stratégies and Inventory Costs for a full
truck (Berman et al [2006])
Table 2 Products’ shipment through Toronto cross-dock
Table 3 Number of Shower Cleaner Kits shipped from Toronto
DC – Single DC central-stock strategy
Table 4 Number of ZipLoc Sandwich Bags shipped from Toronto
DC – Single DC central-stock strategy
Table 5 Number of Shower Cleaner Kits stored in the stores
(Inventory)
Table 6 Number of ZipLoc Sandwich Bags stored in the stores
(Inventory)
Table 7 Cost Comparison
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
12
1. Motivation
In summer 2006 I had an opportunity to work as a replenishment intern
at the Hudson’s Bay Co. (Hbc) and gain valuable hands-on experience in the
domain of Supply Chain (SC). Prior to my internship at the Hbc I had also had
summer research experience in the SC area with Professor Kwon. My research
topic was to basically develop a Mathematical Optimization model for the Safety
Stock in a Multi-Echelon Inventory system subject to uncertain demand with a
quoted Guaranteed Service Time at each echelon. My combination of the
research and practical SC experience and interest towards this area as well as
the recommendation of Professor Kwon made me to prepare my thesis on the
Optimization of the Hbc’s Supply Chain network. The project that I am assigned
to is part of the Hbc Strategic Inventory Optimization and Logistics Network
Design project that determines the inventory deployment decisions regarding the
selection of locations and facilities where products get stored or cross-docked.
The project was initially introduced to me by D. A. Power at the Hbc’s
Optimization division.
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
13
2. Introduction
Achievement of significant cost savings and improvements in profitability
requires a typical retail company to make long-term decisions regarding the
structure of its supply chain network and bringing its facilities, suppliers and
customers closer together under the strategic supply chain planning (Shapiro
[2005]). SC network optimization models allow us to model multiple inventory
deployment decisions under multiple scenarios and finally analyze the results of
each scenario. That being said my purpose here is to model the Hbc’s logistics
network, and to study and analyze the trade-off between taking flow-through
and central stock approaches- which will be explained in “Supply Chain Network
Design” section- in terms of total costs as well as in-stock position. In the next
step the model will be coded in OPL (Optimization Programming Language). In
this thesis I have studied two major distribution strategies of Hbc’s supply chain
network namely cross-dock and central stock approaches for one of its suppliers-
SC Johnson. As part of my study I have developed a transportation-inventory
model for a single source to multiple destinations scenario. I have also studied
the distribution model proposed by (Berman et al [2006]) that can be used for
Hbc’s distribution system.
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
14
3. Literature Review
3.1 Supply Chain Network Design
Before discussing the importance and effect of the inventory deployment
decisions in the supply chain network optimization, it is beneficial to go over the
major reasons for going through the process of the supply chain redesign and
optimization. According to Shapiro et al [2004] the triggering events for the
design of the supply chain can be mergers, acquisitions, inconsistent service
levels, rapid growth, poor asset utilization, rising costs, etc. These are all key
drivers that can force the companies to reconsider their supply chain in order to
better understand and meet their customers’ needs by efficiently utilizing their
manufacturing and distribution resources and potentials. According to (Shapiro et
al [2004]), in order to study the primary objectives of supply chain design
projects, SLIM Technologies-a leading provider of supply chain management
solutions- has chosen a number of companies that had conducted the project.
Almost all companies had agreed that cost reduction, maximization of asset
utilization and service improvement were the top three motives for the redesign
of the supply chain.
According to (Chopra [2003]) there are four decisions to be made
regarding the design of the supply chain network. The first decision is related to
the facility role-decisions that define the role and processes that are performed
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
15
at each facility. The second decision is regarding the facility location. Since
facility location decisions affect subsequent SC decisions; therefore, all aspects of
these decisions have to be carefully studied. Two major considerations would be
the cost associated with locating the facility and its proximity to the market.
Capacity allocation decisions also determine the amount of capacity to be
allocated to each facility and finally decisions regarding market and supply
allocations decide what markets and which supply sources should each facility
respectfully serve and be fed from.
Fisher [1997] in his paper “What is the right Supply Chain for your
product?” emphasizes the consideration of the nature of the demand for one’s
products before devising a supply chain. In other words what he identified as the
main cause of problems in supply chain was a mismatch between the type of
product and the type of supply chain. According to Fisher [1997] on the basis of
products demand patterns, they either fall into functional or innovative
categories of products.
3.2 Hudson’s Bay Company (Hbc) Inventory Structure and Product
Flow
Before getting into the modeling stage of the Hbc’s inventory across its
supply chain I would like to briefly explain the three approaches to Hbc’s in-stock
strategy namely flow-through, central stock (rapid replenishment and slow
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
16
moving). Furthermore I will explain what happens in an Hbc DC from the very
first stage where the product is received from the supplier until it reached onto
the stores’ shelves.
As mentioned earlier the Hbc in-stock strategy consists of 3 approaches:
The first approach is flow-through which locates the stock close to suppliers for
the most cost efficient movement of product from suppliers to stores. The
second approach which is regional central stock (rapid replenishment) locates
central stock inventory close to stores for fast allocation when it is needed. And
finally the last approach which locates slow-moving central stock in one facility to
accommodate fluctuations in demand and special buys is called single DC central
stock (slow-moving).
That being said, the industry best practice is to stock fast-moving SKUs
regionally to ensure a reliable in-stock position whereas the Hbc’s current
network is well-designed for flow-through; therefore, Vision 2524 (industry best
practices) is adopted to determine the best stocking strategies for commodities.
The other issue that is worth knowing is the process that a product goes
through in order to reach onto the stores’ shelves. This process contains 4 stages
(Figure 1). The first stage is receipt which is responsible for booking receiving
appointments from the suppliers and also managing the receiving of shipment
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
17
and printing of labels of products. The above mentioned roles at the first stage
fall under Inbound Management System (IMS) and National Purchase Order
Management System (NPOMS) respectively. The second stage is also responsible
for receiving shipments, allocating pallets to storage slots and printing labels. At
this stage the shipment is either delivered based on flow-through or central stock
approaches. Warehouse Management System (WMS) and NPOMS control this
stage. At the third stage (Splitcase) pickers handle the received batches by store.
This stage falls under pick and pack (P&P) system. As the batches got handled
by store, they move to the next stage where they get sorted and get ready for
shipment. This stage falls under Warehouse Control System (WCS). Lastly the
batches get shipped to the stores by tractors and trailer. The Transportation
Management System (TMS) is responsible for managing the fleet of tractors and
trailers.
Figure 1 - Schematic representation of the process that
takes products to get onto stores’ shelves (Courtesy of Hbc)
CENTRAL
STOCK
FLOW
THROUGH
SPLITCASE SHIP TRANSPO
RT RECEIVE
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
18
4. Models
4.1 Preliminary Distribution Model
Before getting into the modeling portion of my work, I would like to briefly
explain inventory requirements for each of the product groups and how they
vary according to different distribution channels (Shapiro [2005]). As shown in
Figure 2, products can flow in three different paths. In the first path, product is
shipped through a cross dock to a store, meaning that no inventory is held in
that place. Inventory is only held at a store. Costs associated with this path are
transportation as well as fixed and variable processing costs at cross dock site.
Cost of transportation is also related to the shipment volume (either truckload
(TL) or less-than-truckload (LTL). In the second path, product is directly shipped
by the supplier to stores. The only costs associated with this path are the costs
of transportation and inventory at stores. In the third path inventory is only held
both at DCs and stores. Again transportation, inventory holding and fixed as well
as variable processing costs are considered for this path.
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
19
Figure 2 – Distribution strategies (Shapiro [2005])
In the modeling of the network, I have considered a supply chain in which
suppliers ship product either directly to stores, or cross dock site or DCs as
explained earlier. Location and capacity allocation decisions have to be made for
distribution centre (DC). Multiple DCs may be used to satisfy demand at a
market.
The goal is to identify distribution locations as well as quantities shipped
between various points that minimize the total fixed and variable costs. Define
the following decision variables:
The preliminary version of the problem is formulated as the following integer
program:
Suppliers
Distribution
Centre (DC)
Cross-Dock
Stores
(1) (2) (3)
vf CC ,
tC
tC
tC
hvf CCC ,,
tC
hC
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
20
Objective Function
Min
∑ ∑∑∑∑∑= = =
+
=
+
=
++
=
++
l
h
l
h
t
e
xc
n
i
xc
t
e
VFy
n
i
VFy hehehihieeeiii
1 1 1
*
1
*
1
)(*
1
)(*
∑ ∑ ∑∑ ∑∑= = == =
+
=
+
l
h
t
e
m
j
xc
n
i
m
j
xc
m
j
xc ejejijijhjhj
1 1 1
*
1 1
*
1
*
Subject to
1. ∑∑∑= = =
≤++
n
i
t
e
h
m
j
hjhehi Sxxx1 1 1
)( for h=1,…, l
Total amount shipped from a supplier cannot exceed the supplier’s capacity
2. ii
m
j
ij yKx *1
≤∑=
for i=1,…,n
Amount stocked in the DC cannot exceed its capacity
3. ∑ ∑= =
≡−
l
h
m
j
ejhe xx1 1
0 for e=1,…,t
The amount shipped out of a cross-dock site is exactly equal the amount
received from the supplier
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
21
4. ∑ ∑= =
≥−
l
h
m
j
ijhi xx1 1
0 for i=1,…,n
The amount shipped out of a DC site cannot exceed the amount received from
the supplier
5. ∑∑∑= = =
=+
l
h
n
i
t
e
jijej Dxx1 1 1
for j=1,…,m
The amount shipped to a customer must cover the demand
The objective function minimizes the total fixed and variable costs of the
supply chain network. The constraint in equation 1 specifies that the total
amount shipped from a supplier cannot exceed the supplier’s capacity. The
constraint equation 2 enforces that amount stocked in the DC cannot exceed its
capacity. The constraint in equation 3 states that the amount shipped out of a
cross-dock site is exactly equal the amount received from the supplier. The
constraint in equation 4 specifies that the amount shipped out of a DC site
cannot exceed the amount received from the supplier. The constraint in equation
5 specifies that the amount shipped to a customer must cover the demand.
4.2 Cross-Dock and Direct Shipment Models (Berman et al [2006])
Before getting into the details of the final model that I developed and
used for 1 supplier, 1 cross-dock and 2 distributions centers case as well as
assumptions that I made in order to maintain the linearity of the objective
function I would like to briefly describe an optimization model suggested by
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
22
(Berman et al [2006]) that is the solution to a distribution strategy selection
problem where cost functions of both direct delivery and shipment through a
cross-dock are modeled and compared.
According to (Berman et al [2006]) there are two important issues in the
supply chain area that contribute to the total cost of the supply chain network
namely transportation and inventory costs. That being said retail companies can
achieve significant savings by considering these two costs at the same time
rather than trying to minimize each separately. As mentioned above in this paper
the two distribution strategies mainly direct delivery and shipment through cross-
dock are considered where a group of products are shipped from a set of
suppliers to a set of plants. The cost function consists of the total transportation,
pipeline inventory, and plant inventory costs. The presence of the plant inventory
cost has made the model to be formulated as a non-linear integer programming
problem. According to (Berman et al [2006]) the objective function is highly non-
linear and neither convex nor concave; therefore, a greedy heuristic is suggested
to find an initial solution and an upper bound. And then a branch-and-bound
algorithm is developed based on the Lagrangian relaxation of the non-linear
program.
Before getting into the formulation portion of the model, I am going to
provide a brief background of the two distribution strategies discussed in the
paper and then briefly state the assumptions made by (Berman et al [2006]) in
order to have a solvable problem. According to Jonathan Patrick a post-doctorial
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
23
fellow at Center for Health Care Management, Sauder School of Business, UBC
“As far as your model accurately reflects reality, it is not solvable; and as far as it
is solvable, it will not reflect reality”.
4.2.1 Distribution Strategies
For many retail companies products are shipped by suppliers through one
of the following shipment strategies. The first one is direct shipment where
products get shipped directly from the supplier to the DC/plant without stop. The
second method of shipment is milk-run (peddling) where trucks pick up products
from different suppliers on their ways and finally drop them at one or several
DCs. The last but not least is cross-dock where products get shipped to DCs
through cross-dock by suppliers. Below is a graphical representation of the three
distribution strategies.
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
24
Figure 3 – Distribution stratégies (Berman et al [2006])
Each of these three distribution strategies has its own advantages and
disadvantages over the other strategies based on the delivery time and the
transportation cost criteria. Products delivered directly from the supplier to the
plant cost less due to its short distance between the supplier and the plant and
also take less time to get to the plant. As for cross-dock the transportation cost
is high due to the long distance between suppliers and plants and therefore
takes longer to deliver the products. An advantage of the cross-dock strategy
over the direct delivery is that products from several suppliers can be shipped
from a single cross-dock site to the plant which means high delivery frequency
and low plant inventory whereas in the case of direct delivery only products from
the same supplier are shipped to the same plant which means low delivery
frequency and high plant inventory. The Table 1 represents the relationship
between each distribution strategy and its associated cost.
Diretc
Milk-run
Cross-dock
Supplier
Plant
Cross-dock
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
25
Distribution
Strategy
Delivery Time Delivery
Frequency
Inventory
Cost: Pipeline
Inventory
Cost: Plant
Direct Short Low Low High
Milk-run Medium Medium Medium Medium
Cross-dock Long High High Low
Table 1 – Distribution Stratégies and Inventory Costs for a full truck (Berman et al [2006])
4.2.2 Model Assumptions
As mentioned earlier to have a solvable problem, a couple of assumptions
have been made in this paper.
1. It is assumed that the product quantities are infinitely splittable, in other
words a product can be shipped in any quantity within a vehicle shipment.
2. Delivery frequency can be any positive number and is not limited to a set
of potential nembers.
3. Products are always available for shipping at suppliers, no matter which
distribution strategy is chosen.
4. Inbound-outbound coordination at the cross-dock is ignored.
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
26
5. All units of the same flow (a flow is a combination of supplier, plant and,
product) are assigned to the same transportation option, i.e., direct or
through the same cross-dock.
6. Each truck is fully loaded. Only the volume of products is concerned when
calculating truck capacity usage. The transportation costs are only
determined by the source and destination, regardless of the weight.
4.2.3 Formulation
Below is the nonlinear integer mathematical formulation of (Berman et al
[2006])’s model.
Objective Function
Min ∑∈
0
)(Kk
k Xg
Subject to
(1) ∑∈
=0
1kk
ijpkx ,),,( Fpji ∈∀
(2) }1,0{∈ijpkx 0,),,( KkFpji ∈∈∀
In this model if k is 0-in other words, if 10 =ijpx means that flow (i,j,p) is shipped
directly. X in the objective function represents the vector of all decision variables
ijpkx and )(Xg k is the total of transportation, pipeline inventory, and plant
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
27
inventory costs to ship flows through cross-dock k. Constraint 1 ensures that
every flow is delivered and constraint 2 makes sure that the same flow follows
the same route (Berman et al [2006]).
The following two equations are the expanded forms of )(Xg k first for direct
shipment and second for shipment of flows that travel through cross-dock. The
detailed objective function can be found in (Berman et al [2006]).
Total cost of direct delivery-transportation costs, pipeline inventory costs and
inventory costs:
∑∑∑∈∈ ∈
++=
ijPpd
ij
ijpijpp
ijpijpp
d
ij
d
ij
Ii Jj
d
ijf
xdhxdhtcfXg )]
*2
*****(*[)(
0
00
In the above formula the frequency of shipment between each supplier-plant
pair (i,j) is formulated as follows:
Cxdbf
ijPp
ijpijpp
d
ij /** 0∑∈
=
The total cost of shipping flows that travel through cross-dock k consists of two
parts: the inbound transportation cost and outbound transportation cost.
According to (Berman et al [2006]) the shipping time through cross-dock k is
formulated as )( 0
kjk
i
ik tTt ++ that is used in the calculation of pipeline
inventory.
∑ ∑ ∑∑∈ ∈ ∈ ∈
+++++=
Ii Jj Jj IPpi
kjijpkijppijpkijppkjk
i
ikkjkj
i
ik
i
ikk
j
fxdhxdhtTtcfcfXg),(
0000 )]2/(****).[(**)(
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
28
Where
∑ ∑∈ ∈
=
Jj Pp
ijpkjp
i
ik Cxdbf /** and
∑∑∈ ∈
=
Ii Pp
ijpkijppkj Cxdbf /**0
For each Ii∈ and Jj∈ are the frequency of inbound and outbound shipment.
4.3 Hbc Transportation and Inventory Model
As discussed earlier a distribution model that integrates both
transportation cost and inventory cost in its objective function can be modeled as
a non-linear integer optimization program and solved by branch-and-bound
algorithms and other heuristic methods due to the inherent difficulty of such
problems.
That being said for this project in order to have a linear objective function
and be able to get answers using optimization programming languages (OPL) as
well as due to the time limitation, I have prepared my models for single source
and single destination cases and at the end compared their total costs. My
model consists of one supplier in Ontario, two DCs-one in Toronto and the other
one in Calgary and finally nine retail stores in Calgary. The DC in Toronto also
can act as a cross-dock site. Figure 4 is a graphical representation of the
possible shipment routes.
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
29
Figure 4 - Product delivery routes
Three scenarios are considered:
Scenario 1: Products are delivered from the supplier through cross-dock at
Toronto to stores in Calgary- Flow-through approach
Scenario 2: Products are shipped from the supplier to the Toronto DC-products
are stored in Toronto DC - and shipped to the stores in Calgary – Single DC
Central Stock (slow moving products) approach
Scenario 3: Products are shipped from the supplier to the Calgary DC – products
are stored in Calgary DC – and shipped to the stores in Calgary – Regional
Central Stock (Rapid Replenishment) approach
Due to the simplified nature of the model, I have formulated and coded two
simple transportation and inventory models. Below are the mathematical
formulations of the models.
Supplier- SC Johnson
Calgary DC Toronto
DC/Cross-
dock
Calgary Stores (9)
1 2 9
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
30
4.3.1 Transportation Model Formulation
Objective Function
Min
∑ ∑ ∑= = =
+∗
t
p
n
i
m
j
ipijpij CTX1 1 1
)(
Subject to
6. ∑=
=
m
j
phipij SX1
for h=1,…, l ; p=1,…,t; i=1,…,n
7. ∑=
=
n
i
pjpij DX1
for p=1,…,t; j=1,…,m
8. i
t
p
pij KX ≤∑= 1 for i=1,…,n; j=1,…,m
4.3.2 Inventory and Transportation Model Formulation
Objective Function
Min
∑ ∑ ∑ ∑= = = =
++
k
p
n
i
m
j
l
t
pjtppijtijr YIXTC1 1 1 1
**)(
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
31
Subject to
1. ∑∑= =
≤
k
p
m
j
ipijt KX1 1
for i=1,…,n; t=1,…,l
2. ∑=
−+=+
n
i
pjtpjtpijttpj YDXY1
)1( for p=1,…,k; j=1,…,m;t=1,…,l
3. pjpjt IIY = for p=1,…,k; j=1,…,m
4.3.3 Programming Using OPL
At the next stage of my research I coded the mathematical formulations
using ILOG’s OPL (Optimization Programming Language) software. In OPL
the model and the data are saved in two separate files and once it is run
successfully solutions to the objective function as well as the decision
variables defined in the model appear in a separate window at the bottom of
the coding window. Details of the codes can be found in Appendices A and D.
4.3.4 Data Acquisition
The data required to populate the models with were obtained through BIS
system at Hbc. Through BIS it is possible to run weekly, monthly, seasonal
and annual sales reports for any set of SKU/LOC (i.e. product-location)
combination as well as for any set of vendors. As mentioned earlier the data
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
32
are saved in a separate file. Details of the data can be found in Appendices B,
B and E.
5. Comparison of Models – Results
After populating the models with data and running them, the following results
were obtained from each scenario. Table 2 provides the quantity shipped
from the supplier through Toronto cross-dock to Calgary stores (scenario 1).
Since scenario 1 is a single source to multiple destination case and there is no
inventory held at the cross-dock site; therefore, the amount shipped to each
store is equal to its demand. The total delivery cost-transportation and DC
costs (receive, put away, pick and shipment) - for one period is $ 513.34
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
33
Shower Cleaner Kit ZipLoc Sandwich Bag
Calgary Store 108 1 11
Calgary Store 109 1 11
Medicine Store 243 1 4
Lethbridge Store 286 1 13
Calgary Store 433 1 17
Calgary Store 435 0 9
Calgary Store 462 1 8
Calgary Store 480 1 9
Calgary Store 498 1 15
Table 2-Products’ shipment through Toronto cross-dock
In scenarios 2 and 3 products are delivered from the supplier to the DCs
in Toronto and Calgary respectively. Products are stored and then shipped to
the stores in Calgary based on the demand at each store. The model for
these two scenarios incorporates a new decision variable namely the amount
of inventory and a cost parameter associated with it into the transportation
model from the first scenario. Tables 3 and 4 contain the number of shower
cleaner kits and ZipLoc sandwich bags that are shipped from Toronto DC to
all 9 Calgary stores in each of the 5 periods respectively. Tables 5 and 6 show
the amount of inventory held at each store after satisfying the demand in
that period. As we see in table 5 there are one or two units of shower cleaner
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
34
kit left at certain periods whereas for ZipLoc sandwich bag (table 6) the
inventory is zero. That being said the total delivery cost – transportation, DC
costs (receive, put away, pick and shipment) and inventory cost – for
scenarios 2 and 3 over all 5 periods are $ 1785.9 and $ 2201.9 respectively in
other words $ 357.18 and $ 440.38 for a single period.
The cost comparison of the three scenarios (table 7) shows that the
scenario 2 – shipment from the supplier to Toronto DC (central stock) – is
less costly compared to the scenarios 1 and 3 ($ 513.34 and $ 440.38
respectively).
Details of the solutions can be found in Appendices C and F-1 and F-2.
Period 1 Period 2 Period 3 Period 4 Period 5
Calgary_108 2 0 0 5 0
Calgary_109 1 1 1 0 3
Medicine_243 1 0 0 1 3
Lettbridge_286 1 0 0 1 3
Calgary_433 2 0 0 0 1
Calgary_435 1 2 1 0 1
Calgary_462 2 1 1 0 0
Calgary_480 1 0 0 0 0
Calgary_498 3 1 2 0 3
Table 3 - Number of Shower Cleaner Kits shipped from Toronto DC – Single DC central-stock strategy
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
35
Period 1 Period 2 Period 3 Period 4 Period 5
Calgary_108 10 14 10 1 5
Calgary_109 6 13 7 16 5
Medicine_243 4 1 7 4 6
Lettbridge_286 14 13 4 7 7
Calgary_433 6 16 8 12 16
Calgary_435 7 2 4 0 6
Calgary_462 7 4 5 3 4
Calgary_480 12 7 5 3 4
Calgary_498 17 6 7 11 8
Table 4 - Number of ZipLoc Sandwich Bags shipped from Toronto DC – Single DC central-stock strategy
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
36
Period 1 Period 2 Period 3 Period 4 Period 5
Calgary_108 0 0 0 2 2
Calgary_109 0 0 0 0 1
Medicine_243 0 0 0 0 0
Lettbridge_286 0 0 0 0 0
Calgary_433 0 0 0 0 0
Calgary_435 0 0 1 0 0
Calgary_462 0 0 1 1 0
Calgary_480 0 0 0 0 0
Calgary_498 0 0 1 0 0
Table 5 - Number of Shower Cleaner Kits stored in the stores (Inventory)
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
37
Period 1 Period 2 Period 3 Period 4 Period 5
Calgary_108 0 0 0 0 0
Calgary_109 0 0 0 0 0
Medicine_243 0 0 0 0 0
Lettbridge_286 0 0 0 0 0
Calgary_433 0 0 0 0 0
Calgary_435 0 0 0 0 0
Calgary_462 0 0 0 0 0
Calgary_480 0 0 0 0 0
Calgary_498 0 0 0 0 0
Table 6-Number of ZipLoc Sandwich Bags stored in the stores
Scenario 1 Scenario 2 Scenario 3
Total cost for one
period(CAD$)
513.34
357.18
440.38
Table 7-Cost Comparison
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
38
6. Future Research
Although it has been concluded that delivery through cross-dock has the
longest distance, and therefore the highest transportation cost and the longest
delivery time but on the other hand cross-dock can combine products from
different suppliers which means high delivery frequency and low plant inventory.
Such trade-offs between transportation and inventory costs should be studied
more in order to obtain a near-optimal or local-optimal solution for a typical
distribution problem. Given that non-linear integer optimization models are hard
to solve due to the non-linearity of their objective functions; therefore,
developing heuristic methods and branch-and-bound algorithms so far have
proven to be very successful (Berman [2006]) for that reason the search for new
heuristic methods and algorithms should be continued.
7. Conclusion
In this thesis, I have considered the problem of selecting the most cost
efficient in-stock strategy – flow-through, regional and single DC central stock
approaches– for delivering two products from a single supplier to the Hbc’s
Toronto and Calgary DCs respectively and then shipping them to the stores in
Calgary so that the total transportation, distribution center and inventory costs
are minimized. A distribution model, proposed by (Berman [2006]) for the flow-
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
39
through (cross-dock) strategy is recommended and a simple transportation-
inventory model is developed to compare the cost associated with each strategy.
The results have demonstrated that taking the Single DC central stock strategy
with the DC in Toronto is the most cost-efficient. This strategy is suitable for
slow-moving and less demanding items. The second cost efficient strategy after
the Single DC central stock is the regional central stock. Even though this
approach is costlier than the first one but it is best suited for rapid moving items
for fast allocation when it is needed.
It is recommended to adopt Single DC and Regional central stock strategies for
slow and fast moving products respectively.
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
40
8. References Shapiro, J. F., Wagner S. [2005], “Strategic Inventory Optimization” SLIM
Technologies, October, 2005
Shapiro, J. F., Wagner S., Jacobs K. [2004], “A Practical Guide to Supply Chain
Network Design” SLIM Technologies, September, 2004
Fisher M. L. [1997], “What is the Right Supply Chain for Your Product?” Harvard
Business Review, 1997
Chopra S., Meindl P. [2003], Supply Chain Management: Strategy, Planning, and
Operations, 2nd ed. Prentice Hall
Berman O., Whang Q. [2006], “Inbound Logistic Planning: Minimizing
Transportation and Inventory Cost” transportation Science, 2006
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
41
9. Collection of Figures & Tables
Figure 1 (Courtesy of the Hbc)
Figure 2 (Shapiro [2005])
Suppliers
Distribution
Centre (DC)
Cross-Dock
Stores
(1) (2) (3)
vf CC ,
tC
tC
tC
hvf CCC ,,
tC
hC
CENTRAL
STOCK
FLOW
THROUGH
SPLITCASE SHIP TRANSPO
RT RECEIVE
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
42
Figure 3 (Berman et al [2006])
Distribution
Strategy
Delivery Time Delivery
Frequency
Inventory
Cost: Pipeline
Inventory
Cost: Plant
Direct Short Low Low High
Milk-run Medium Medium Medium Medium
Cross-dock Long High High Low
Table 1 (Berman et al [2006])
Diretc
Milk-run
Cross-dock
Supplier
Plant
Cross-dock
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
43
Figure 4 (Product delivery routes)
Shower Cleaner Kit ZipLoc Sandwich Bag
Calgary Store 108 1 11
Calgary Store 109 1 11
Medicine Store 243 1 4
Lethbridge Store 286 1 13
Calgary Store 433 1 17
Calgary Store 435 0 9
Calgary Store 462 1 8
Calgary Store 480 1 9
Calgary Store 498 1 15
Table 2-Products’ shipment through Toronto cross-dock
Supplier- SC Johnson
Calgary DC Toronto
DC/Cross-
dock
Calgary Stores (9)
1 2 9
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
44
Period 1 Period 2 Period 3 Period 4 Period 5
Calgary_108 2 0 0 5 0
Calgary_109 1 1 1 0 3
Medicine_243 1 0 0 1 3
Lettbridge_286 1 0 0 1 3
Calgary_433 2 0 0 0 1
Calgary_435 1 2 1 0 1
Calgary_462 2 1 1 0 0
Calgary_480 1 0 0 0 0
Calgary_498 3 1 2 0 3
Table 3-Number of Shower Cleaner Kits shipped from Toronto DC using central-stock strategy
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
45
Period 1 Period 2 Period 3 Period 4 Period 5
Calgary_108 10 14 10 1 5
Calgary_109 6 13 7 16 5
Medicine_243 4 1 7 4 6
Lettbridge_286 14 13 4 7 7
Calgary_433 6 16 8 12 16
Calgary_435 7 2 4 0 6
Calgary_462 7 4 5 3 4
Calgary_480 12 7 5 3 4
Calgary_498 17 6 7 11 8
Table 4-Number of ZipLoc Sandwich Bags shipped from Toronto DC using central-stock strategy
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
46
Period 1 Period 2 Period 3 Period 4 Period 5
Calgary_108 0 0 0 2 2
Calgary_109 0 0 0 0 1
Medicine_243 0 0 0 0 0
Lettbridge_286 0 0 0 0 0
Calgary_433 0 0 0 0 0
Calgary_435 0 0 1 0 0
Calgary_462 0 0 1 1 0
Calgary_480 0 0 0 0 0
Calgary_498 0 0 1 0 0
Table 5-Number of Shower Cleaner Kits stored in the stores
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
47
Period 1 Period 2 Period 3 Period 4 Period 5
Calgary_108 0 0 0 0 0
Calgary_109 0 0 0 0 0
Medicine_243 0 0 0 0 0
Lettbridge_286 0 0 0 0 0
Calgary_433 0 0 0 0 0
Calgary_435 0 0 0 0 0
Calgary_462 0 0 0 0 0
Calgary_480 0 0 0 0 0
Calgary_498 0 0 0 0 0
Table 6 - Number of ZipLoc Sandwich Bags stored in the stores (Inventory)
Scenario 1 Scenario 2 Scenario 3
Total cost for one
period(CAD$)
513.34
357.18
440.38
Table 7 - Cost Comparison
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
48
Appendix A
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
49
Transportation Model in OPL Code
Model:
{string} DCs =...; {string} Stores =...; {string} Products = ...; float Capacity[DCs] = ...; float Supply[Products][DCs] = ...; float Demand[Products][Stores] = ...; assert forall(p in Products) sum(o in DCs) Supply[p][o] == sum(d in Stores) Demand[p][d]; float Trans_Cost[Products][DCs][Stores] = ...; float DC_Cost[Products][DCs][Stores] = ...; dvar float+ Trans[Products][DCs][Stores]; minimize sum( p in Products , o in DCs, d in Stores ) Trans[p][o][d] * (Trans_Cost[p][o][d] + Trans_Cost[p][o][d]); subject to { forall( p in Products , o in DCs ) ctSupply: sum( d in Stores ) Trans[p][o][d] == Supply[p][o]; forall( p in Products , d in Stores ) ctDemand: sum( o in DCs ) Trans[p][o][d] == Demand[p][d]; forall( o in DCs, d in Stores ) ctCapacity: sum( p in Products ) Trans[p][o][d] <= Capacity[o]; } execute DISPLAY { writeln("trans = ",Trans); }
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
50
Appendix B
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
51
Transportation Model’s Data
Data: DCs = { TORONTO CALGARY }; Stores = {CALGARY_108 CALGARY_109 MEDICINE LETHBRIDGE CALGARY_433 CALGARY_435 CALGARY_462 CALGARY_480 CALGARY_498}; Products = { SHOWER_KIT SANDWICH_BAG }; Capacity = [100 100]; Supply = #[ SHOWER_KIT: #[ TORONTO: 8 CALGARY: 0 ]# SANDWICH_BAG: #[ TORONTO: 97 CALGARY: 0 ]# ]#; Demand = #[ SHOWER_KIT: #[ CALGARY_108: 1 CALGARY_109: 1 MEDICINE: 1 LETHBRIDGE: 1 CALGARY_433: 1 CALGARY_435: 0 CALGARY_462: 1 CALGARY_480: 1 CALGARY_498: 1 ]# SANDWICH_BAG: #[ CALGARY_108: 11 CALGARY_109: 11 MEDICINE: 4 LETHBRIDGE: 13 CALGARY_433: 17 CALGARY_435: 9
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
52
CALGARY_462: 8 CALGARY_480: 9 CALGARY_498: 15 ]# ]#; Trans_Cost = #[ SHOWER_KIT: #[ TORONTO: #[ CALGARY_108: 2.3002 CALGARY_109: 2.3002 MEDICINE: 2.3214 LETHBRIDGE: 2.332 CALGARY_433: 2.332 CALGARY_435: 2.4592 CALGARY_462: 2.5122 CALGARY_480: 2.6394 CALGARY_498: 2.756 ]# CALGARY: #[ CALGARY_108: 0.636 CALGARY_109: 0.636 MEDICINE: 1.1342 LETHBRIDGE: 0.954 CALGARY_433: 0.6572 CALGARY_435: 0.6678 CALGARY_462: 0.6678 CALGARY_480: 0.689 CALGARY_498: 0.6996 ]# ]# SANDWICH_BAG: #[ TORONTO: #[ CALGARY_108: 2.3002 CALGARY_109: 2.3002 MEDICINE: 2.3214 LETHBRIDGE: 2.332 CALGARY_433: 2.332 CALGARY_435: 2.4592
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
53
CALGARY_462: 2.5122 CALGARY_480: 2.6394 CALGARY_498: 2.756 ]# CALGARY: #[ CALGARY_108: 0.636 CALGARY_109: 0.636 MEDICINE: 1.1342 LETHBRIDGE: 0.954 CALGARY_433: 0.6572 CALGARY_435: 0.6678 CALGARY_462: 0.6678 CALGARY_480: 0.689 CALGARY_498: 0.6996 ]# ]# ]#; DC_Cost = #[ SHOWER_KIT: #[ TORONTO: #[ CALGARY_108: 2.26807141 CALGARY_109: 2.26807141 MEDICINE: 2.26807141 LETHBRIDGE: 2.26807141 CALGARY_433: 2.26807141 CALGARY_435: 2.26807141 CALGARY_462: 2.26807141 CALGARY_480: 2.26807141 CALGARY_498: 2.26807141 ]# CALGARY: #[ CALGARY_108: 2.26807141 CALGARY_109: 2.26807141 MEDICINE: 2.26807141 LETHBRIDGE: 2.26807141 CALGARY_433: 2.26807141 CALGARY_435: 2.26807141 CALGARY_462: 2.26807141
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
54
CALGARY_480: 2.26807141 CALGARY_498: 2.26807141 ]# ]# SANDWICH_BAG: #[ TORONTO: #[ CALGARY_108: 2.26807141 CALGARY_109: 2.26807141 MEDICINE: 2.26807141 LETHBRIDGE: 2.26807141 CALGARY_433: 2.26807141 CALGARY_435: 2.26807141 CALGARY_462: 2.26807141 CALGARY_480: 2.26807141 CALGARY_498: 2.26807141 ]# CALGARY: #[ CALGARY_108: 2.26807141 CALGARY_109: 2.26807141 MEDICINE: 2.26807141 LETHBRIDGE: 2.26807141 CALGARY_433: 2.26807141 CALGARY_435: 2.26807141 CALGARY_462: 2.26807141 CALGARY_480: 2.26807141 CALGARY_498: 2.26807141 ]# ]# ]#;
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
55
Appendix C
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
56
Transportation Model’s Solution Report
Cost calculation of products cross-docked at Toronto DC: Final solution with objective = 513.34: Trans = [[[1 1 1 1 1 0 1 1 1] [0 0 0 0 0 0 0 0 0]] [[11 11 4 13 17 9 8 9 15] [0 0 0 0 0 0 0 0 0]]];
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
57
Appendix D
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
58
Transportation & Inventory Model in OPL Code
Model:
{string} Products = ...; {string} DCs = ...; int NbPeriods = ...; range Periods = 1..NbPeriods; {string} Stores = ...; float Capacity[DCs] = ...; float Demand[Products][Stores][Periods] = ...; float TransCost[DCs][Stores] = ...; float Inventory[Products][Stores] = ...; float InvCost[Products] = ...; float DC_Cost[DCs] = ...; dvar float+ Trans[Products][DCs][Stores][Periods]; dvar float+ Inv[Products][Stores][0..NbPeriods]; minimize sum( p in Products, r in DCs, s in Stores, t in Periods ) ((DC_Cost[r] + TransCost[r][s])*Trans[p][r][s][t] + InvCost[p]*Inv[p][s][t]); subject to { forall( r in DCs, t in Periods ) ctCapacity: sum( p in Products, s in Stores ) Trans[p][r][s][t] <= Capacity[r]; forall( p in Products , s in Stores, t in Periods ) ctDemand: Inv[p][s][t-1] + (sum( r in DCs) Trans[p][r][s][t]) == Demand[p][s][t] + Inv[p][s][t]; forall( p in Products, s in Stores ) ctInventory: Inv[p][s][0] == Inventory[p][s]; };
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
59
Appendix E
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
60
Transportation & Inventory Model’s Data
Data:
Products = { SHOWER_KIT SANDWICH_BAG }; DCs = { TORONTO CALGARY }; Stores = {CALGARY_108 CALGARY_109 MEDICINE LETHBRIDGE CALGARY_433 CALGARY_435 CALGARY_462 CALGARY_480 CALGARY_498}; NbPeriods = 5; Capacity = [ 400, 400 ]; Demand = [ [[2 0 0 3 0] [1 1 1 0 2] [1 0 0 1 3] [1 0 0 1 3] [2 0 0 0 1] [1 2 0 1 1] [2 1 0 0 1] [1 0 0 0 0] [3 1 1 1 3]] [[10 14 10 1 5 ] [6 13 7 16 5 ] [4 1 7 4 6] [14 13 4 7 7] [6 16 8 12 16 ] [7 2 4 0 6 ] [7 4 5 3 4] [12 7 5 3 4] [17 6 7 11 8]] ]; Inventory = [ [0 0 0 0 0 0 0 0 0] [0 0 0 0 0 0 0 0 0 ] ]; InvCost = [ 0.2 0.2]; TransCost = [ [2.3002 2.3002 2.3214 2.332 2.332 2.4592 2.5122 2.6394 2.756] [0.636 0.636 1.1342 0.954 0.6572 0.6678 0.6678 0.689 0.6996] ]; DC_Cost = [2.26807141 5.0648]; Supply = [ [ [14 5 5 7 14] [0 0 0 0 0] ] [ [83 76 57 57 61] [0 0 0 0 0] ] ];
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
61
Appendix F-1
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
62
Transportation & Inventory Solution Reports
Solution Report 1: Final solution with objective = 1785.9: Trans = [[[[2 0 0 5 0] [1 1 1 0 3] [1 0 0 1 3] [1 0 0 1 3] [2 0 0 0 1] [1 2 1 0 1] [2 1 1 0 0] [1 0 0 0 0] [3 1 2 0 3]] [[0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0]]] [[[10 14 10 1 5] [6 13 7 16 5] [4 1 7 4 6] [14 13 4 7 7] [6 16 8 12 16] [7 2 4 0 6] [7 4 5 3 4] [12 7 5 3 4] [17 6 7 11 8]] [[0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0]]]]; Inv = [[[0 0 0 0 2 2] [0 0 0 0 0 1] [0 0 0 0 0 0]
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
63
[0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 1 0 0] [0 0 0 1 1 0] [0 0 0 0 0 0] [0 0 0 1 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0]]];
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
64
Appendix F-2
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
65
Transportation & Inventory Solution Reports
Solution Report 2:
Final solution with objective = 2201.9: Trans = [[[[0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0]] [[2 0 0 5 0] [1 1 2 0 2] [1 0 1 0 3] [1 0 0 1 3] [2 0 0 0 1] [1 2 1 0 1] [2 1 0 0 1] [1 0 0 0 0] [3 1 1 1 3]]] [[[0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0] [0 0 0 0 0]] [[10 14 10 1 5] [6 13 7 16 5] [4 1 7 4 6] [14 13 4 7 7] [6 16 8 12 16] [7 2 4 0 6] [7 4 5 3 4] [12 7 5 3 4] [17 6 7 11 8]]]]; Inv = [[[0 0 0 0 2 2] [0 0 0 1 1 1] [0 0 0 1 0 0]
A Supply Chain Optimization Model: Minimizing Transportation and Inventory Costs
66
[0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 1 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0]] [[0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0] [0 0 0 0 0 0]]];