Professional Documents
Culture Documents
This presentation has been made in accordance with the provisions of Part VB of the copyright
act for the teaching purposes of the University.
Copyright© 2005 by Heng-Soon Gan
Optimisation = Efficiency + Savings
• Kellogg’s
– The largest cereal producer in the world.
– LP-based operational planning (production, inventory, distribution)
system saved $4.5 million in 1995.
• Procter and Gamble
– A large worldwide consumer goods company.
– Utilised integer programming and network optimization worked in
concert with Geographical Information System (GIS) to re-engineering
product sourcing and distribution system for North America.
– Saved over $200 million in cost per year.
• Hewlett-Packard
– Robust supply chain design based on advanced inventory optimization
techniques.
– Realized savings of over $130 million in 2004
Source: Interfaces
Mathematics in Operation
Real Practical Problem
Mathematical (Optimization) Problem
x2
Computer Algorithm
Human Decision-Maker
Decision Support
Interface
Information Systems
A Team Effort
Users
Interface
Information Systems
Mathematical
Mathematical (Optimization) Problem Programming
Childcare Centre
Human Decision-Maker Manager
The Mathematical Problem
• Modelled as an Integer LP
x
i 1
ik 4, k D Skill Demand
Workplace Regulation
Xpress MP
• Database Management:
– Staff Profile (Name, Category)
– Annual leave
– Shift preferences
– Reserve staff
– Roster
– etc….
• Project aim: support the Army (Force Design Group) with their
capability options development and analysis, seeking
– What types of forces should be maintained?
– What force strength is required?
to ensure forces are effective in achieving defence objectives
Choose forces
(STRATEGIC) $ $ $ $ budget
Force
configuration
Deploy forces e e e e e e e max effectiveness
(TACTICAL)
Objectives
The Mathematical Model
• An integer LP-based prototype decision
support tool has been developed.
• Future directions
– database management
– integrated military systems – Military Information
System
The ForceOp Tool
• Before this tool,
– force design was carried out manually
– a lengthy and laborious process, based on intuitive-
reasoning (no quantitative basis).
– difficult to assess effectiveness or compare quality of
solutions
LP as a “What-If” Tool
The Facility Location Problem
• LP-based techniques can be used to locate
– manufacturing facilities,
– distribution centres,
– warehouse/storage facilities etc.
taking into consideration factors such as
– facility/distribution capacities,
– customer demand,
– budget constraints,
– quality of service to customers etc.
using Operations Research techniques such as
– linear programming,
– integer linear programming, and
– stochastic programming.
Customer
10 000 W-3
180 000 36km
Warehouse
W-4 (W)
C D
220 000 Assume:
180 000
• Transportation cost:
$20/km/unit
B W-5
E
• Warehouses have the same
10 000 units W-2
O/H cost
A • Warehouse has very large
F capacity
W-1 W-6 Problem modelled as an
integer linear program, and
10 000
solved using XpressMP.
The Mathematical Model
n n d
Minimise f x W
i 1
i i
i 1 j 1
ij yij
s .t .
d
y
j 1
ij Ci xi , i 1 n
y
i 1
ij Dj , j 1 d
xi 0 ,1
yij is int eger
yij
j
i
xi
Scenario 1
• Scenario 1: 10 000 W-3
180 000
Warehouse O/H W-4
cost is very small as C D
compared to 220 000
180 000
transportation cost
– Warehouse O/H:
B W-5
E
$6 000 000
W-2
– Transportation cost:
10 000 units
$20/km/unit
– proximity dominates
A
F
– operate the
W-1 W-6
warehouse closest
to each customer 10 000
Scenario 2
• Scenario 2: Warehouse 10 000 W-3
O/H cost is very large 180 000
as compared to W-4
transportation cost C D
– Warehouse O/H: 220 000
180 000
$1 800 000 000
– Transportation cost:
$20/km/unit
B W-5
E
– too expensive to 10 000 units W-2
operate a warehouse
– hence, the most A
centralised warehouse F
selected (based on
demand & distance) W-1 W-6
10 000
Scenario 3
• Scenario 3: Both 10 000 W-3
180 000
warehouse O/H and W-4
transportation costs C D
are competing 220 000
180 000
– Warehouse O/H:
$60 000 000
B W-5
E
– Transportation cost:
$20/km/unit 10 000 units W-2
– solution is not
obvious; too many A
possibilities
F
W-1 W-6
10 000
Scenario 4
• Scenario 4: Both 10 000 W-3
180 000
warehouse O/H and W-4
transportation costs C 150 000
D
are competing AND 220 000
180 000 10 000
warehouse capacity 30 000
• Possible extensions
– limitations to the number of distribution centres
– warehouse-customer distance constraint
– complex cost functions
– uncertain demand
Other OR Applications
• Other areas where OR techniques have been proven
to be useful include
– Inventory control
– Warehouse design, storage and retrieval, order picking
– Vehicle routing
– Delivery transport mode selection
– Capacity and manpower planning
– Production scheduling