Dynamics of Supply Chains: A Multilevel Network

1
Dynamics of Supply Chains A Multilevel
Network Perspective on Sustainable Freight
Movements Frank
Southworth STELLA Workshop
on Globalization, E-Economy and Trade
University of Siena, Italy
June 7- 9, 2002 On
Assignment at the Bureau of Transportation
Statistics, US Department of
Transportation, Washington DC 20590
2

Outline of This Presentation
1. Problem Statement 2. Some Trends That Need
Addressing 3. A Multilevel Modeling
Framework 4. Summary
3
The Problem Rapid Growth in The Freight Sector
Is Producing More Traffic Congestion-Induced
Delays More Fossil Fuel Consumption More
Environmental (Air, Water, Noise)
Pollution Greater Consumption of Land
How Are We Going To Handle This Growing
Demand For Goods Movement?
4
The Freight Sector Is Expected To Grow
Substantially
In the United States, for example

• Some 9.1 billion tons of freight, worth 9.4
  trillion, were moved into, out of, and within
  the USA in calendar year 1998
• Forecasts suggest that (from 1998-2020)
• US-Domestic freight will grow at 2.8 per year
• ( 87 cumulative increase)
• US-International freight will grow 3.4 per year
• ( 107 cumulative increase)

ImplicationWell need a good deal more freight
handling capacity
Source FHWA, Office of Freight Management
Operations, July 2001.
5

• The Freight Sector Is Also Changing Rapidly
• Demand Side Changes (Market Evolution)
• Advances in real time, high capacity information
  technology have encouraged/opened the way for
• -- Rapid growth in e-commerce
• -- Growing demand for just-in-time delivery
  services
• (transfers of costs up the supply chain)
  and a shift of emphasis
• towards demand-driven supply chains
• -- An increasingly global marketplace for low
  cost goods delivery

6
Trend Towards Customer (Demand) Driven Logistics
Systems
Information
System
Transport System
PUSH
METHODS OF
CONTROL
(relative
importance)
Inventory
5
7
Supply Side Changes Some trends in freight
handling and storage -- A growing
interest in intermodal and containerized forms
of transportation. -- The design and
use of larger, more cost-effective vehicles/vessel
s -- The emergence of large
consolidation/break-bulk facilities and
freight villages -- Interest
in/emergence of high volume freight corridors.
The emergence of new business relationships
-- The development of enterprise-wide,
integrated, multi-step product supply chains.
-- The emergence of freight intermediaries,
including large 3PLs (global)
4PLs
8
Implication To understand trends and policy
options in the freight sector we must understand
the SOUP to NUTS of freight transportation
logistics. To do this we need to
1) Treat transportation as part of a broader, and
increasingly global logistics exercise 2)
Recognize that ON-TIME IS MONEY and that
reliability of service is now a very high
priority with many shippers and receivers 3)
Develop a better understanding of freight
economies of scale 4) Develop sufficiently
comprehensive analytic frameworks for
modeling/simulating sustainable freight transport
solutions.
9
An Example of Trying To Get FROM SOUP TO NUTS
10
Major Components of the Regional Goods Movement
Simulator
Supply Chain Decision Making Simulator
(SSDMS)
Physical Transportation Network
Multilevel Logistical/Financial/ Informational Net
work
11
Supply Chain -Transportation Supernetwork
Representation
12
Freight Traffic Simulator
Supply Chain Decision Making Simulator
(SSDMS)
Freight Traffic Simulator (FTS)
Physical Transportation Network
Multilevel Logistical/Financial/ Informational Net
work
13
Logistical Network Simulating the Supply Chain
14
Three-Tier Supply Chain Representation Used
in Model Development
15
Multilevel Network Structure of the Supply Chain
Decision Making Simulator
From Nagurney, Ke, Cruz, Hancock Southworth,
2001,Dynamics of Supply Chains A
Multilevel (Logistical/Informational/Financial)
Network Perspective
16

• This Multilevel, Dynamic Supply Chain System has
  now been
• Formulated as a complex network systems problem
  (based on
• Nagurney and Dong, 2000)
• Formulated as a dynamical system with a set of
  behavioral rules
• that focus on the dis-equilibrium aspects
  of supply chains
• under competition.
• Translated into a variational inequality problem
  with a
• unique and stable equilibrium solution
• Solved algorithmically using an iterative,
  discrete time adjustment process (based on the
  scheme in Dupuis Nagurney,1993)
• Made operational as a FORTRAN program

17
Supply Chain
Simulator -- System Costs Firms (producers) have
production costs and transaction (with retailer)
costs Retailers have commodity purchase costs
(from producers), transaction costs (with both
producers and consumers) and product handling and
storage costs Consumers (market demand) have
purchase costs plus transaction (with retailer)
costs. Note Bene Transaction costs can be
quite general, and include transportation,
financial, and information gathering costs.
18

• Supply Chain Simulator --
  Behavioral Rules
• Producers and retailers are profit
  maximizers. Consumers are cost minimizers. The
  commodity is homogeneous.
• The system tends towards a spatial
  equilibrium of prices and
• commodity flows. ( Samuelson, 1953
  Takayama and Judge, 1971 Nagurney, 1999)
• It is assumed that a fair price for producing
  firms to charge for a commodity the marginal
  costs of production transaction costs.
• At equilibrium, commodity flows occur between a
  producer- retailer pair if the marginal cost of
  production plus the marginal cost of transaction
  and handling is equal to the price of the
  commodity at the retail outlet. If marginal cost
  exceeds price, no shipments occur.
• At equilibrium, the volume purchased from
  retailers exactly equals the demand at that
  product market. For shipments to occur the price
  paid at market must equal each retailers
  marginal production transaction costs.

19

• Supply Chain Simulator --
  System Dynamics
• Consumer demands drive the supply chain (based
  on market demand functions)
• If a retailers price plus transaction costs
  exceeds a markets willingness to pay, the volume
  of goods moved between that retailer-demand
  market pair decreases. If the retailers price is
  below willingness to pay the volume of goods
  moved increases in relation to this difference.
• If demand at a specific market location exceeds
  existing retail supply, then the price consumers
  are willing to pay at that location increases in
  relation to the size of this unmet demand.
• Prices charged at retail locations reflect both
  market demand conditions and producer supply
  conditions
• The volume of a commodity shipped between a
  producing firm and a retailer evolves according
  to the difference between the retailers charged
  price and its marginal costs.These costs include
  transaction costs and the price charged for the
  commodity by the producing firm.

20
Freight Traffic Simulator
Supply Chain Decision Making Simulator
(SSDMS)
Freight Traffic Simulator (FTS)
Physical Transportation Network
Multilevel Logistical/Financial/ Informational Net
work
21
Freight Traffic Simulator

• Micro-Simulation is used to assign the
  origin-to-destination shipment volumes estimated
  by the Supply Chain Simulator to the physical
  transportation network.
• Complete cargo movements from production point to
  final consumption point
• are to be modeled.
• This may involve different modes at different
  stages in the physical supply
• chain. These movements are simulated over the
  ORNL multi-modal
• freight transportation network (Southworth
  Peterson, 2000).
• Congestion-induced delay costs will be modeled on
  each link along a route,
• including congestion at intermodal terminals.
  These will include the economic
• cost of highly variable day-to-day travel
  times ( as a service reliability cost).
• Transportation cost calculations will eventually
  require vehicle/container type
• and cargo size as well as type of carriage
  (for-hire vs. private) to be modeled.

22
Principal Steps in the Complete Freight Traffic
Micro-Simulator (under construction)
23
The ORNL North American Truck-Rail-Water
Intermodal Network
24
The ORNL Trans-Oceanic Waterways Network
This network is functionally linked to the ORNL
Highway-Rail-Inland and Coastal Waterways
Network database to allow routing of foreign
imported and exported freight, including US
Land-Bridge Traffic.
25
An Intermodal Freight Routing Example
26
local rail spur line
27
Freight Traffic Simulator
Supply Chain Decision Making Simulator
(SSDMS)
Freight Traffic Simulator (FTS)
Physical Transportation Network
Multilevel Logistical/Financial/ Informational Net
work
Real-time Information Simulator (RTIS)
28
Useful Applications of the Simulator Include
Explorations of
1) The benefits of more accurate and timely
information on ---- at-market commodity
prices ---- the pattern of freight flows
---- fuel consumption and related
pollutants production ---- profitability,
and longer term facility (re)locations 2) The
benefits of alternative forms of supply chain,
including ---- the use of mixed,
multiple level business connections between
firms in different network tiers
(e.g.direct producer-to-consumer
purchases and deliveries, as well as transactions
through intermediaries). 3) The effects of
traffic congestion (and the value of its
avoidance) on ---- at-market commodity
prices ---- the pattern of freight flows
---- mode choice ---- fuel
consumption and related pollutants production
---- profitability, and longer term facility
(re)locations 4) The functional links between
different supply chain configurations and
different network/terminal configurations.