Joe Monahan Josh Onuska Back Story NS Largest rail shipper of automobiles Scope Exclusive deal with Ford 19 million sold in 2010 Facilities operating 24 hours 7 days East Coast distribution ID: 535605
Download Presentation The PPT/PDF document "Multi-Commodity Flow" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Multi-Commodity Flow
Joe Monahan
Josh OnuskaSlide2
Back Story
N.S. - Largest rail shipper of automobilesScopeExclusive deal with Ford (1.9 million sold in 2010)Facilities operating 24 hours / 7 days East Coast distributionRisksTerrorism or warLabor union strikesSevere weatherDerailing / rail failureSlide3
Why Multi-Commodity?Slide4
Nodes and Arcs
Nodes5 assembly plants4 mixing centers25 distribution facilitiesArcs22,000 miles of rail30,000 Ford vehicles shipped weeklyCapacity is more than sufficient to handle demandSlide5
N.S. Rail SystemSlide6
Automotive Shipping FacilitiesSlide7
Assembly Plants & Mixing CentersSlide8
Assembly Plants
Location Supply Model Cleveland 5,000 Econoline Van
Chicago 5,000 Taurus, ExplorerDetroit 10,000 Mustang, F-150, Focus Kansas City 5,000 F-150, Escape, Contour Louisville 5,000 SuperDuty,Expedition,NavigatorSlide9
ContainerizationSlide10
Mixing Centers
Location Capacity RegionChicago 5938 NE/NWLouisville 3510 SE/EToledo 2800 NEKansas City 2538 NW/MWSlide11
Chicago Mixing CenterSlide12
Node OverlaySlide13
Node OverlaySlide14
Model
Multi-commodity: Sorted (5) / MixedSupply/Demand DrivenSupply = 30,000Demand = 30,00060,000 units in motionAccounts for 75% of Ford’s monthly U.S. salesNetwork RisksArcs unusableWeatherAccident
UnionsBridgesSlide15
Cleveland(Avon Lake)
ChicagoDetroit (Dearborn/Wayne)
LouisvilleKansas City
Toledo
Chicago
Louisville
Kansas City
Mixed Source
Mixed
Sink
Assembly Plants
Mixing Stations
Assembly Sinks
-30,000
30,000
-5000
-5000
-5000
-5000
-10000
D
C
L
K
A
5000
5000
5000
5000
10000Slide16
Network Characteristics
Network GoalMeeting demand at the minimum costDNM: Current vs Future Profit lossMeasure of effectiveness Average cost to deliver vehicle to distribution facilityMOE = Total Cost / # of cars shippedTotal Cost = (.25)(Rail Miles) + (DNM)($2700)Slide17
Assumptions
30,000 cars moved per week (From Mid-West Plants)20,000 to East Coast / 10,000 to West Coast$0.25 per rail mile for each car moved No cost incurred for movement between co-located facilitiesDNM results in lost salesSlide18
Assembly Plants & Mixing CentersSlide19
Assembly Plants to Mixing Facilities
Mixing Facility
Assembly PlantSlide20
Mixing Facilities to Distribution SitesSlide21
Initial Results
Base line – zero attacks:Total Mileage: 25,511,500 Rail milesTotal Cost: $6,377,875.00Cost per vehicle: $212.00Slide22
Results w/ Interdiction
1 Interdiction23,219,000 Miles; $11,204,750; $386.37/car5 Interdictions14,460,000 Miles; $49,515,000; $2,303.02/car10 Interdictions8,010,000 Miles; $88,402,500; $6,314.46/car17 Interdictions5,010,000 Miles; $139,600,500; $31,872.60/carSlide23
Impacts of InterdictionSlide24
Resiliency Curve
Cost over profit margin
Attack SaturationSlide25
Charts: Cost of Interdiction
Map of varying paths after each interdictionAnd/or histograms of cost at each interdictionSlide26
Resiliency
How do we make the system more robust and resistant to catastrophic collapse?Add arcs to isolated nodes Back up plan (trucks)Can we move nodes around (mixing stations or distribution points) to get a more efficient solution? Moving or even eliminating mixing stations improves performance.Slide27
Optimal Attack HistogramSlide28
Frequency of Attack LocationsSlide29
Additional Analysis
Option to eliminate one of the four mixing facilities:Bypassed Chicago$194,375 in weekly savingsSlide30
Analysis Limitations
Lots of assumptionsChanges in Consumer Demand by geographic locationSlide31
Conclusion
What did we learn from the model?Hub and spoke affectHow does the model react to interdiction?Greatest impact from interdiction seen in DNMCritical vulnerabilities?Isolated nodes, close proximity of Assm. PlantsHow can we improve the system to lower cost?Eliminate Mixing FacilityDiminishing returns after 17 attacksSlide32
Questions