Beruflich Dokumente
Kultur Dokumente
Chain Networks
1. Key design factors
2. Distribution options
3. Design of optimal distribution networks under
various configurations and constraints
1–2
BA5056 TRANSPORTATION AND
DISTRIBUTION MANAGEMENT
UNIT - I DISTRIBUTION 9
Role of Distribution in Supply chain, Distribution
channels – Functions, resources, Operations in
Distribution, Designing Distribution network
models - its features - advantages and
disadvantages.
The Role of Distribution
in the Supply Chain
Distribution – the steps taken to move and store a
product from the supplier stage to the customer
stage in a supply chain
Drives profitability by directly affecting supply
chain cost and the customer experience
Choice of distribution network can achieve supply
chain objectives from low cost to high responsiveness
2 phases: a) structure, b) detailed specifications
a) directly or through intermediaries?
b) convert the structure into locations, capabilities,
capacities, demand allocation
Network Design
Matching Supply with Demand
Warehouses /
Suppliers DCs Cross Docks Stores Customers
Product Multiple
Normal demand curve Categorization Carriers Demand shifts quickly
Product Flow
Supplier Rates
Capacity Customer
Locations
Products
Demand
Consolidation Cross Docks Pattern
Vendors
Centers & DCs Zone Base
Warehousing Transport
Restrictions Rates for
Sourcing Product Set
Policies Warehouse
Capacity Warehouse
Capability on Utilization &
Handling Products Limitations
Transportation
Facilities
and handling
Information
Desired Response Time and Number of
Facilities
McMaster-Carr (6
facilities 1-2 day lead
time most US cities)
Inventory Costs and Number
of Facilities
B&N
Amazon
B&N
PPET = 7
Amazon
PPET = 19
The retailer plays same role as before, but carrier merges in-transit the order
if from multiple manufacturers. Example: order a Dell PC and a Sony
monitor; carrier merges at a hub before delivering to customer
2. In-Transit Merge
Facility role
What is the role, what processes are performed?
Facility location
Where should facilities be located?
Capacity allocation
How much capacity at each facility?
Market and supply allocation
What markets each facility serve? Which supply
sources should feed each facility?
Factors Influencing
Network Design Decisions
Strategic factors
Technological factors
Macroeconomic factors
Tariffsand tax incentives
Exchange-rate and demand risk
Freight and fuel costs
Political
Infrastructure factors
Customer response time and local presence
Logistics and facility costs
Framework for Network Design
Decisions
Mathematical Models & Examples
1. Capacitated Plant Location Models
2. Practical considerations
3. Advanced models with multiple attributes
Models for Facility Location and
Capacity Allocation
Network design models are used to decide on
locations and capacities, and to assign current
demand to facilities
Objective:
Maximize the overall profitability of the supply chain
network, while providing customers with the
appropriate responsiveness
Many trade-offs during network design
Network Optimization Models and
Numerical Examples
Capacitated Plant Location Models (also known as
Facility Location and Capacity Allocation Models)
Model 1: Optimal Production Allocation or
Transshipment Problem
Model 2: Consider location specific costs and decide
which plants should remain open
Model 3: Enforce single sourcing for each market
K i = capacity of factory i
cij = cost of producing and shipping one unit from factory i to market j
subject to
n
n m
x D j for j 1,..., m
Min cij xij
ij
i 1
m
i 1 j 1
x
j 1
ij K i for i 1,..., n
Network Model 1
Numerical Example
Demand City
Monthly
Production and Transportation Cost
Capacity
per Thousand Units (Thousand $)
(Thousand
Supply City Atlanta Boston Chicago Denver Omaha Portland Units) K
Baltimore 1,675 400 985 1,630 1,160 2,800 18
Cheyenne 1,460 1,940 970 100 495 1,200 24
Salt Lake City 1,925 2,400 1,450 500 950 800 27
Baltimore 0 8 2 0 0 0
Cheyenne 0 0 0 6 0 0
Salt Lake 0 0 0 0 0 11
Memphis 10 0 12 0 0 0
Wichita 0 0 0 0 7 0
Z = $26,463
Network Model 2: Selecting plants
Given that you can meet supply with less plants,
decide which plants should remain open
There are now location-specific costs to consider!
yi = 1 if factory i is open, 0 otherwise
xij = quantity shipped from factory i to market j
Network Model 2: Selecting Plants
n = number of potential plant locations/capacity
m = number of markets or demand points yi = 1 if plant i is open, 0 otherwise
Dj = annual demand from market j xij = quantity shipped from plant i
to market j
Ki = potential capacity of plant i
fi = annualized fixed cost of keeping plant i open
cij = cost of producing and shipping one unit from plant i to market j (cost
includes production, inventory, transportation, and tariffs)
n n m
Min f i yi c x ij ij
subject to i 1 i 1 j 1
n
x ij D j for j 1,..., m
i 1 Note!
m
x
j 1
ij K i yi for i 1,..., n
Numerical Example
Demand City
Monthly Monthly
Production and Transportation Cost
Capacity Fixed Cost
per Thousand Units (Thousand $)
(Thousand (Thousand
Supply City Atlanta Boston Chicago Denver Omaha Portland Units) K $) f
Baltimore 1,675 400 985 1,630 1,160 2,800 18 7,650
Cheyenne 1,460 1,940 970 100 495 1,200 24 3,500
Salt Lake 1,925 2,400 1,450 500 950 800 27 5,000
City
Memphis 380 1,355 543 1,045 665 2,321 22 4,100
Wichita 922 1,646 700 508 311 1,797 31 2,200
Monthly 10 8 14 6 7 11
demand
(thousand
units) Dj
Network Model 2
Formulation of the model
Open/
Closed Atlanta Boston Chicago Denver Omaha Portland
Baltimore Open 0 8 2 0 0 0
Cheyenne Open
0 0 0 6 7 11
Salt Lake Closed
0 0 0 0 0 0
Memphis Open
10 0 12 0 0 0
Wichita Closed
0 0 0 0 0 0
Z= $ 47,401
Network Model 3: Single sourcing
Each market should be supplied by only one factory
Modify objective, constraints & decision variables as
follows:
yi = 1 if factory i is open, 0 otherwise
xij = 1 if market j is supplied by factory i, 0 otherwise
n n m
Min f i yi D j cij xij
i 1 i 1 j 1
subject to
x
i 1
ij 1 for j 1,..., m
m Note!
D x
j 1
i ij K i yi for i 1,..., n
xij , yi 0,1
Network Model 3
Present the formulation of the model
Open/
Closed Atlanta Boston Chicago Denver Omaha Portland
Baltimore Closed 0 0 0 0 0 0
Cheyenne Closed 0 0 0 0 0 0
Memphis Open 10 8 0 0 0 0
Wichita Open 0 0 14 0 7 0
Z= $49,717
Network Model 4: Multiple
Capacity Options
We need to decide both the location and capacity
allocation for each facility as earlier, however:
We have multiple capacity options for each location (in
particular a high and a low option)
Network Model 4
Modelling approach:
Treat each scenario option as a separate facility and
introduce binary variables for each one of them as
earlier. For example:
Binaryvariables Y1 and Y5 will refer to N. America, and will
denote the low and high capacity, respectively.
Add a constraint to make sure that only one capacity
will be realized!
e.g. Y1+Y5 ≤ 1
Network Model 4
Formulation of the model
Min Z = 6000Y1+…+ 4000Y5+
+ 9000Y6+…+6000Y10
+1675x11 + 400X12 + ... + 311X55 + 1797X56
+1675x61 + 400X62 + ... + 311X12,5 + 1797X10,6
s.t.
(X11 + X21 + ... + X51) + (X61 + X71 + ... + X101) = 10
…. Satisfy the demand
of each demand city
(X16 + X26 + ... + X56) + (X66 + X76 + ... + X10,6) = 7
Capacity of the plants (in ton per year) and demand from
customers is also known
Uniform flow restrictions among all arcs of the network
cij = cost of producing and shipping one unit from node i to node j
subject to
n n n n
D:400
across each arc.
S:200
6
There are production costs
1
and storage costs.
2 7
S:300
D:180
3 S:100 5 T:0
Network Model 5
Numerical Example (cont.)
s.t.
Node 1: (X12 + X13 + ... + X17) – (X21 + X31) = 200
Node 2: (X21 + X23 + ... + X27) – (X12 + X32) = 300
Node 3: (X32 + X31 + ... + X37) – (X13 + X23) = 100
Network model 5
Origin Destination Unit Cost Flow
1 2 5.0 0
1 3 3.0 180
1 4 5.0 0
1 5 5.0 0
1 6 20.0 0
T:0 1 7 20.0 0
4
2 1 9.0 0
D:400
2 3 9.0 0
S:200 6 2 4 1.0 300
2 5 1.0 0
1
2 6 8.0 0
2 7 15.0 0
2 7 3 1 0.4 0
S:300
3 2 8.0 0
D:180
3 4 1.0 0
3 5 0.5 280
3 6 10.0 0
3 S:100 5 T:0 3 7 12.0 0
4 5 1.2 0
4 6 2.0 300
4 7 12.0 0
5 4 0.8 0
Z: 2320 5 6 2.0 280
5 7 12.0 0
6 7 1.0 180
7 6 7.0 0
Network Model 6: Network flow
for multiple products
Minimum cost network flow model - Multiple Products
Same network structure as before (this can easily relaxed)
2 7
S1:300
D1:180
S2:100
D2:140
3 S1:100 T:0
5
S2:100
Assume that the unit shipping costs are the same for both
products (this can be easily relaxed)
Network Model 6
Mathematical Model
xijp = quantity shipped of product p from node i to node j
cij = cost of producing and shipping one unit from node i to node j
p n n
Min c x ij ijp
subject to
p 1 i 1 j 1 n n
x
j 1
ijp x jip Di for all p and i 1,..., n
j 1
s.t.
Node 1, P1: (X121 + X131 + ... + X171) – (X211 + X311) = 200
Node 1, P2: (X122 + X132 + ... + X172) – (X212 + X312) = 200
…
Node 7, P1: X761 - (X171 + X271 + ... + X671) = - 400
Node 7, P2: X762 - (X172 + X272 + ... + X672) = - 200
X121+X122 ≤ 300
…
X761+X762 ≤ 300
Xij 0
Solution of Network Model 6
Origin Destination Unit Cost Flow Product 1 Flow Product 2 Total flow Arc Capacity
1 2 5.0 0 0 0 <= 300
1 3 3.0 180 120 300 <= 300
1 4 5.0 0 20 20 <= 300
1 5 5.0 0 0 0 <= 300
1 6 20.0 0 0 0 <= 300
1 7 20.0 0 0 0 <= 300
2 1 9.0 0 0 0 <= 300
2 3 9.0 0 0 0 <= 300
2 4 1.0 100 0 100 <= 300
2 5 1.0 0 0 0 <= 300
2 6 8.0 200 100 300 <= 300
2 7 15.0 0 0 0 <= 300
3 1 0.4 0 0 0 <= 300
3 2 8.0 0 0 0 <= 300
3 4 1.0 0 180 180 <= 300
3 5 0.5 280 20 300 <= 300
3 6 10.0 0 0 0 <= 300
3 7 12.0 0 20 20 <= 300
4 5 1.2 0 0 0 <= 300
4 6 2.0 100 200 300 <= 300
4 7 12.0 0 0 0 <= 300
5 4 0.8 0 0 0 <= 300
5 6 2.0 280 20 300 <= 300
Z=5570 5
6
7
7
12.0
1.0
0
180
0
120
0
300
<=
<=
300
300
7 6 7.0 0 0 0 <= 300
Network Optimization Models and
Examples cont.
More enhanced Capacitated Plant Location Models
Model 7: Simultaneous location of plants and warehouses
Model 8: Accounting for Taxes, Tariffs, and Customer
Requirements
Model 9: Adding restrictions on the number of locations +
alternative objective functions (see excel file “AI's
Athletic.xls”)
Model 10: Production and shipping through transshipment
points with multiple modes of transportation + location
decisions (see excel file “MIP network flow with multiple
trans modes and location.xls”
Network Model 7
n t l n n t t m
Min Fi yi f e ye chi xhi cie xie cej xej
i 1 e 1 h 1 i 1 i 1 e 1 e 1 j 1
Network Model 7
subject to
x S h for h 1,..., l m
x
hi
i 1 ej We ye for e 1,..., t
j 1
l t
x
hi ie
h 1 e 1 ej D j for j 1,..., m
e 1
t
x K i yi for i 1,..., n
yi , ye 0,1, xej , xie , xhi 0
ie
e 1
n m
x –x
i 1
ie
j 1
ej 0 for e 1,..., t
Network Model 8
x
i 1
ij D j for j 1,..., m
Network Model 9
Y
i 1
ij 1 for j 1,..., J
Yij X i i I , j J
x
iI
i P
Network Models 9 & 10
AI’s Athletic excel file
MIP network flow with multiple trans modes and
location excel file
Network Optimization Models and
Examples cont.
Multiple products and modes of transportation?
DHL Supply Chain Case
Multiple products, modes of transportation +
location decisions but no fixed costs?
Logistics
Network Design Problem see paper of
Cordeau, Pasin and Solomon 2003
Multiple products, location decisions + fixed costs?
Mixed Integer Non-Linear Programming (see Floudas
1995)
Making Network Design Decisions In
Practice
Do not underestimate the life span of facilities
Do not gloss over the cultural implications
Do not ignore quality-of-life issues
Focus on tariffs and tax incentives when locating
facilities