XYZ Company Supply Chain Optimization Project Network Optimization

1
XYZ CompanySupply Chain Optimization
ProjectNetwork Optimization
ISyE 6203 Transportation and Supply Chain
Management
Prepared By Jayson Choy Christie Williams Andy
Ang Thomas Ou Naragain Phumchusri Raghav
Himatsingka
Date 04/25/2006
2
Agenda

• Introduction
• Key Deliverables
• Data Analysis
• Mathematical Model
• Recommendation
• Sensitivity Analysis
• Conclusion

3
Introduction

• Locations in Florida and California
• Each location has Multiple Operations
• Suppliers across USA
• Supplier shipments may be parcel,
  less-than-truckload or full truckload, some must
  be frozen or chilled

Project goal Reduce inbound transportation costs
across the business while meeting customer
service requirements.
4
Timeline
Phase
Jan
Feb
Mar
Apr
Project kick-off Deliverables Rationalization
Data Cleansing Validation
Preliminary Modeling
Validation of Model
Generation of results sensitivity Analyses
5
Agenda

• Introduction
• Key Deliverables
• Data Analysis
• Mathematical Model
• Recommendation
• Sensitivity Analysis
• Conclusion

6
Key Deliverables

• Create a graphic illustration of Current North
  American Supply Chain network
• Document Current Volumes and Freight spend by
  mode to each location and in total
• Identify and recommend North American
  Consolidation Points for most efficient route and
  capacity utilization
• Create a graphic illustration of the Recommended
  New Supply Chain Network with all consolidation
  facility representations and conceptual lanes to
  each location

7
California and Florida Supplier Locations
8
Problem Definition
Florida
9
Problem Definition
13
LTL Greatest Opportunity for Savings
8
36
11
6
26
10
Problem Definition

• Focus on Consolidation of LTL shipments to
  Florida
• Eliminated Frozen and Chilled shipments from the
  Optimization model
• Included most FTL shipments by breaking them down

11
Agenda

• Introduction
• Key Deliverables
• Data Analysis
• Mathematical Model
• Recommendation
• Sensitivity Analysis
• Conclusion

12
Data Analysis
13
Data Analysis
14
Data Analysis
15
Agenda

• Introduction
• Key Deliverables
• Data Analysis
• Mathematical Model
• Recommendation
• Sensitivity Analysis
• Conclusion

16
Supplier Locations (LTL Florida)(Data
Aggregation by 3-Digit Zip Code)
17
To Consolidate LTL Shipments into FTL
Present Situation
Shipper
Shipper
CP
Shipper
Shipper
Proposed Solution
XYZ Company
18
Proposed 2-Step Model
To Generate Potential Consolidation Point
Candidates
Step 1 Set Covering Model (SCM)
Step 2 Network Design Model (NDM)
To Determine which Consolidation Points to Open/
Close
19
Step 1 Set Covering Model (SCM)

• Integer Programming Model
• Model will decide 30 Potential Consolidation
  Points within 300 mile Radius
• from Suppliers.

To Maximize the Number of Suppliers which can be
Covered by the CPs

• Maximize sum(i in Suppliers) yi
• s.t sum(i in Suppliers) xi lt 30 forall(i in
  Suppliers)
• yi lt (sum(j in Suppliers) Matrixi,jxj)

To Generate at most 30 Potential CP locations
To Ensure that CPs are within 300 mile radius
from Suppliers
20
SCM Results 30 Consolidation Points
Next Step CP Candidates will be fed into the
Network Design Model (NDM)
21
Step 2 Network Design Model (NDM)

• Model Objectives
• To Decide which Consolidation Points to open or
  close
• To Determine whether Suppliers should Ship
  Direct to the company
• To Assign Suppliers to Consolidation Points

To Open or Close?
To Open or Close?
To Open or Close?
To Open or Close?
To Open or Close?
To Open or Close?
To Open or Close?
To Open or Close?
To Open or Close
22

• Constraint I If Supplier is not a Candidate CP
• We either serve this 3 Digit zip via LTL
  shipments to the
• destination or via a consolidation point

LTL Shipment
CP
Shipper
Direct LTL Shipment
XYZ Company
23
Step 2 Network Design Model (NDM)

• Constraint II If Supplier is a Candidate CP
• Case 1 If NOT OPEN
• We send LTL direct or via a designated CP

Case 1
XYZ Company
CP
24
Step 2 Network Design Model (NDM)

• Constraint II If Supplier is a Candidate CP
• Case 2 If OPEN
• We consolidate at CP and send FTL direct

Case 2
XYZ Company
CP
25
Step 2 Network Design Model (NDM)

• Constraint III Load Factor of 0.8
• Total Inflow into CP lt 0.8 Total Truckloads
  

LTL
LTL
CP
LTL

ie at least 1 truckload per week

• Constraint IV Frequency
• Minimum Truckloads going through
• an Open CP per year

gt 52
26
Agenda

• Introduction
• Key Deliverables
• Data Analysis
• Mathematical Model
• Recommendation
• Sensitivity Analysis
• Conclusion

27
5 CP Locations
28
Assignment of Suppliers
29
CP Location I Charlotte, NC
30
CP Location II Atlanta, GA
31
CP Location III Los Angeles, CA
32
CP Location IV Gulfport, MS
33
CP Location V Jackson, KY
34
LTL Direct Shipments
35
Cost Savings
8 Reduction
Less than 8 Reduction
36
Summary
CP Location Truckloads Per Year Volume ('000 lbs) Number of Assigned Suppliers
Charlotte NC 190 3800  102
Atlanta GA 52 1040 48
Jackson KY 52 1040 53
Gulfport MS 52 1040 65 
Los Angeles CA 52 1040 28
Direct LTL - - 136
37
Agenda

• Introduction
• Key Deliverables
• Data Analysis
• Mathematical Model
• Recommendation
• Sensitivity Analysis
• Conclusion

38
Sensitivity Analysis
39
Sensitivity Analysis
Shipments Per Week Charlotte Atlanta LA Gulfport Jackson
1 ? ? ? ? ?
2 ? ? ? ? ?
3 ? ? ?    
4 ? ?      
5 ?        
6 ?        
7 ?        
40
Agenda

1. Introduction
2. Key Deliverables
3. Data Analysis
4. Mathematical Model
5. Recommendation
6. Sensitivity Analysis
7. Conclusion

41
Conclusion

• Key Learning
• Counter intuitive peculiarities of LTL cost
  structure (small volumes, backhauling etc)
• Moving Forward
• Milk run study on remaining LTL direct volumes
• Optimization of other shipment modes
• (e.g. parcel, frozen, chilled etc)
• Optimization of Florida bound shipments

42
Q A