Transcript Document

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
Outline of This Presentation:
1. Problem Statement
2. Some Trends That Need Addressing
3. A Multilevel Modeling Framework
4. Summary
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?
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)
Implication:We’ll need a good deal more freight handling
capacity
* Source: FHWA, Office of Freight Management & Operations, July 2001.
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
Trend Towards Customer (Demand) Driven Logistics Systems
Information
System
Transport System
“PUSH”
METHODS OF
CONTROL
(relative
importance)
Inventory
Transport
System
Inventory
5
“PULL”
METHODS OF
CONTROL
(relative
importance)
Source: Based on a U.S.Department of Transportation,
Secretary’s Office of Intermodalism
Presentation
Information
System
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/vessels
-- 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, multistep product supply chains.
-- The emergence of freight intermediaries, including large
3PLs & (global) 4PLs
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.
An Example of Trying To Get “FROM SOUP TO NUTS”:
“Enterprise-Wide Simulation and Analytic Modeling
of Comprehensive Freight Movements”
NSF Grant No. CMS-0085720
(PIs: Kitty Hancock*, Anna Nagurney* & Frank Southworth**)
*University of Massachusetts at Amherst
**Oak Ridge National Laboratory
Purpose: To develop a methodology that links:
a) business logistics (notably financial transactions) to
b) freight movements (i.e. physical transactions), and
c) knowledge-based information flows
within a single analytic (and network-based)framework.
Major Components of the Regional Goods Movement Simulator
I-O Interface
Databases
(Static and Dynamic)
Supply Chain Decision
Making Simulator
(SSDMS)
Freight Traffic
Simulator (FTS)
Physical
Transportation
Network
Multilevel
Logistical/Financial/
Informational
Network
Real-time Information
Simulator (RTIS)
Supply Chain -Transportation Supernetwork Representation
Raw material sources
Distribution centers
Retail Markets
Transaction cost
information
Plant
Financial
Network
Raw material sources
Retail Markets
Demand or order
information
Plant
Logistical
(Product Supply
Chain)
Network
Travel time
information
Unexpected
issues
information
Physical Transportation Network
Two-way information exchanges between specific decision-makers
Information
Network
Freight Traffic Simulator
I-O Interface
Databases
(Static and Dynamic)
Supply Chain Decision
Making Simulator
(SSDMS)
Freight Traffic
Simulator (FTS)
Physical
Transportation
Network
Multilevel
Logistical/Financial/
Informational
Network
Real-time Information
Simulator (RTIS)
Logistical Network: Simulating the Supply Chain
Three-Tier Supply Chain Representation Used
in Model Development
1
i
m
1
j
n
1
k
o
Tier 1:
i = 1…m firms
(manufacturing plants)
Tier 2:
j = 1…n retailers
Tier 3:
k =1 …o demand markets
Multilevel Network Structure of the Supply Chain Decision Making Simulator
Flows are information on
prices and movements
(bi-directional)
Financial Network
p11
p1i
p1m
1
i
m
1
j
n
Logistical Network
1
i
m
q11
1
qmn
j
Informational
Network
n
qno
1
k
p2n
1
k
p31
p3o
o
o
Flows are commodity movements
(top-down, supply driven)
Flows are prices (bottom-up,
market driven payments)
* From: Nagurney, Ke, Cruz, Hancock & Southworth, 2001,“Dynamics of Supply Chains:
A Multilevel (Logistical/Informational/Financial) Network Perspective”
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
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.
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 retailer’s marginal production + transaction
costs.
Supply Chain Simulator -- System Dynamics
Consumer demands drive the supply chain (based on market demand
functions)
If a retailer’s price plus transaction costs exceeds a market’s willingness to
pay, the volume of goods moved between that retailer-demand market
pair decreases. If the retailer’s 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 retailer’s
charged price and its marginal costs.These costs include transaction
costs and the price charged for the commodity by the producing firm.
Freight Traffic Simulator
I-O Interface
Databases
(Static and Dynamic)
Supply Chain Decision
Making Simulator
(SSDMS)
Freight Traffic
Simulator (FTS)
Physical
Transportation
Network
Multilevel
Logistical/Financial/
Informational
Network
Real-time Information
Simulator (RTIS)
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.
Principal Steps in the Complete Freight Traffic Micro-Simulator
(under construction)
The ORNL North American Truck-Rail-Water Intermodal Network
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.
An Intermodal Freight Routing Example
transfer
local
terminal
terminal
access road
local rail
spur line
transfer
terminal
rail line haul
Railroad #2
Railroad #1
interline
highway network link(s)
origin
notional local access link
to highway network
Route Impedance = modal line-haul travel costs
+ intra-terminal transfer costs
+ inter-carrier (interlining) costs
+ local network access and egress costs
+ network-to-terminal local access costs
destination
Freight Traffic Simulator
I-O Interface
Databases
(Static and Dynamic)
Supply Chain Decision
Making Simulator
(SSDMS)
Freight Traffic
Simulator (FTS)
Physical
Transportation
Network
Multilevel
Logistical/Financial/
Informational
Network
Real-time Information
Simulator (RTIS)
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.