Transportation

1
Transportation Supply Chain Systems

• John H. Vande Vate
• Spring, 2001

2
Who am I?

• John H. Vande Vate
• Professor and EMIL Director
• Office 433
• Phone [phone]
• Office Hours
• Tuesday, Thursday 11-12 or
• By appointment

3
Administrative Details

• Class Home Page
• www.isye.gatech.edu/jvandeva/Classes/6203/syllabu
  s2001.htm
• Keep up with information here!
• Text Logistic Systems Analysis by Daganzo
• You may also want to read
• Business Logistics Management by Ballou

4
Grading

• Exams 60
• Mid-term March 1, 2001
• Final
• Projects 30
• Groups of 3 or 4
• Class Participation 10
• What you contribute to the class

5
Objectives

• Issues and Tools for Design and Analysis of
  Logistics Systems
• Daganzo Analytical models based on summary data
• Class Discussions Mathematical Programming
  models incorporating more detailed data

6
Motivation

• Analytic models
• Identify trade-offs
• Communicate arguments
• Ball park answers
• Numerical Models
• Consider the details
• Manage the complexity
• Refine solutions

7
Algebraic Modeling Languages

• AMPL
• www.ampl.com
• AMPL A Modeling Language for Mathematical
  Programming, Fourer, Gay and Kernighan
• No full version generally available
• XPress-MP
• Academic version available for download at
  www.isye.gatech.edu/class/xpress/
• Full version available in the graduate and
  undergraduate labs.

8
More Options

• AIMMS
• This application is supported in the Lab.
• Student version??
• MPL or related tools
• Can be downloaded free from www.maximal-usa.com/do
  wnload/
• No full version available

9
Good and Bad of EMIL

• Good
• Executive Masters in International Logistics
• CIOs, VPs, Directors and Managers from
• Agilent, Baxter, Fedex, Ford, GM, Hapag-Lloyd,
  Intel, Lucent, Milliken, Motorola, Ryder, UPS,
• Meeting with Execs from
• BMW, DaimlerChrysler, HP, Infineon, Quelle,
• ...

10
Good and Bad of EMIL

• Bad
• I will be out of the country
• January 19 to February 3
• Possible trip to Asia in April
• Still wanted to do this class.

11
Resolution

• Weeks of
• January 22
• January 29
• No class. Your assignment is
• Get access to and master a modeling language
• Complete a small case to test your skill.

12
Project 1 Due February 6

• Build and solve a mathematical programming model
  for determining the best strategy for shipping
  the goods in The example of Section 1.2 of
  Daganzo
• Comments
• Use the data given in the text and the distances
  given on the class home page
• Your model must be linear (integer).
• The plants and warehouse stagger shipments.

13
Comments Contd

• Shipments between plants and and from plants to
  the warehouse are full truck load.
• Each DC receives all supply of each part from one
  plant. (Different parts may come from different
  plants)
• If your model is too large for the student
  version of the solver, drop the last dcs until it
  fits.

14
To Turn In

• Model
• with clear but concise explanation
• Solution
• Summarize intelligently
• Compare with an Analytic model
• What are the key differences and why are they
  there.

15
Tentative Schedule
16
Illustrative Example

• Develop a distribution strategy to minimize
  inventory and transportation costs
• Purpose
• Illustrate the analytical modeling approach
• Outline issues covered in the course

17
Overview

• Products
• Computers CPU, Monitor, Keyboard
• Televisions TV and Console
• Distribution Centers
• 100 across the US
• Sell 10 TVs and 10 computers per day

18
An Illustrative Example
Facts
19
Strategies to Consider

• All direct shipments in full truckloads
• All shipments via Indianapolis in full truckloads
• All direct shipments at the optimal frequency
• All shipments via Indianapolis at the optimal
  frequency
• .

20
Simplification

• Shipments on the order of 1,000 miles

Facts
21
Inventory

• At the DCs?
• 1/2 truckload or 500 Consoles at 100 each
  50,000
• 1/2 truckload or 3,000 CPUs at 300 each
  900,000
• 1/2 truckload or 1,500 Monitors at 400 each
  600,000
• 1,550,000
• Carrying cost 15 of 1,550,000 or 232,500
• Total Carrying cost at the DCs 23,250,000!

22
Inventory at the Plants

• Two Extremes
• Simultaneous shipments
• Staggered shipments
• Reality?

23
Simultaneous shipments

• Green Bay ships once every 2.4 years
• Inventory is 1.2 years demand or 3,000 CPUs
• Inventory value is 900,000
• Inventory carrying cost is 135,000
• Denver ships 2.5 times per year
• Inventory is 20 of annual demand or 500 Consoles
• Inventory value is 50,000
• Inventory carrying cost is 7,500
• Indianapolis ships 1.6 times per year
• Inventory is 5,000/3.2 or about 1,500 items
• Inventory value is 600,000
• Inventory carrying cost is 90,000

24
Staggered Shipments

• Each plant holds 1/2 a truck load
• Total plant inventory like one more DC

25
Total Cost

• Transportation Costs 460,000
• Inventory Costs at DCs 23,250,000
• Inventory Costs at Plants 232,500
• Total 23,942,500!
• The Next Step
• Trade-off Inventory and Transportation

26
Costs, Inventory and Transport
27
The Big Picture Finance
28
Extending the Analysis
29
Simple Location Models
30
Back to Our Example

• Using Transshipment to Reduce Inventory
• Strategy 2 Assemble Products in Indianapolis and
  distribute by truckload from there

31
Via Indianapolis
Facts
32
Inventory Effects

• Inventory at a DC
• Visited by truck 4.6 times per year
• Same as before
• So inventory costs at DC are the same as before,
  right?

33
Wrong!

• Total Inventory Carrying Cost at DCs
• 4,891,300
• compared with
• 23,250,000
• under earlier strategy. Explain!

34
The Difference
Other resolutions?
35
Inventory at the Plants

• Denver
• 1/2 truckload of Consoles
• Same as before, right? 7,500
• Green Bay
• 1/2 truckload of CPUs
• Same as before 135,000
• Indianapolis
• 1/2 truckload of Consoles 7,500
• 1/2 truckload of CPUs 135,000
• 1/2 truckload of TVs and Computers 48,900
• 
  191,400

36
With Transshipment

• Transportation Costs 751,800
• Inventory Costs at DCs 4,891,300
• Inventory Costs at Plants 333,900
• Total 5,977,000
• Without Transshipment
• Transportation Costs 460,000
• Inventory Costs at DCs 23,250,000
• Inventory Costs at Plants 232,500
• Total 23,942,500!

37
Other Advantages/Disadvantages

• Disadvantages
• Capital Cost of Warehouse
• Extra handling at Warehouse
• Advantages
• Economies of Scale in Assembly
• Reduced demands on DC labor

38
1-Many Distribution with Transhipment
39
1-Many Distribution
40
Vehicle Routing
41
Transshipment
42
Vehicle Routing
43
Many-to-Many Distribution
44
Many-to-Many Distribution
45
Other Issues

• Workload balancing UPS case
• Combinatorial Bidding Home Depot Case
• Yield/Demand Management
• .

46
(No Transcript)