BTS Logo BTS Navigation Bar Contents References Contacts
NTL Logo NTLNTLNTL
U.S. Department of Transportation
NTL Browse Reference Contributors Future Plans


TRANSIMS: Project Summary and Status MAY 1995








                         TRANSIMS: 
ansportation ANalysis and
                                   SIMulation System


                         PROJECT SUMMARY AND STATUS


                         May 1995


                         Prepared by

                         LaRon Smith,
                         Richard Beckman
                         Keith Baggerly
                         Doug Anson
                         Michael Williams
                         Los Alamos National Laboratory


                         Prepared for
                         U.S. Department of Transportation
                                   Federal Transit Administration
                                   Federal Highway Administration
                                   Office of the Secretary
                         U.S. Environmental Protection Agency



                               TRANSIMS:

             TRansportation ANalysis and SIMulation System

             LaRon Smith, Richard Beckman, Keith Baggerly,
                     Doug Anson, Michael Williams

                    Los Alamos National Laboratory

Abstract

This document summarizes the TRansportation ANalysis and SIMulation
System (TRANSIMS) Project, the system's major modules, and the
project's near-term plans.  TRANSIMS will employ advanced
computational and analytical techniques to create an integrated
regional transportation systems analysis environment.  The simulation
environment will include a regional population of individual travelers
and freight loads with travel activities and plans, whose individual
interactions will be simulated on the transportation system, and whose
environmental impact will be determined.  We will develop an interim
operational capability (IOC) for each major TRANSIMS module during the
five-year program.  When the IOC is ready, we will complete a specific
case study to confirm the IOC features, applicability, and readiness.

Introduction

The TRansportation ANalysis and SIMulation System (TRANSIMS) is part
of the multi-track Travel Model Improvement Program sponsored by the
U.S. Department of Transportation. and the Environmental Protection
Agency.  Los Alamos National Laboratory is leading its development. 
TRANSIMS will address issues resulting from the Intermodal Surface
Transportation and Efficiency Act of 1991, such as considerations of
land use policies, intermodal connectivity, and enhanced transit
service.  It will support analyses of potential responses to the
stringent air-quality requirements of the Clear Air Act Amendments of
1990.

The TRANSIMS Project objective is to develop a set of mutually
supporting realistic simulations, models, and data bases that employ
advanced computational and analytical techniques to create an
integrated regional transportation systems analysis environment.  By
applying forefront technologies and methods, it will simulate the
dynamic details that contribute to the complexity inherent in today's
and tomorrow's transportation issues.  The integrated results from the
detailed simulations will support transportation planners, engineers,
and others who must address environmental pollution, energy
consumption, traffic congestion, land use planning, traffic safety,
intelligent vehicle efficacies, and the transportation infrastructure
effect on the quality of life, productivity, and economy.


Click HERE for graphic.


                               May 1995



TRANSIMS: TRansportation ANalysis and SIMulation System       May 1995

The previous figure illustrates the TRANSIMS architecture.  The
TRANSIMS methods deal with individual behavioral units and proceed
through several steps to estimate travel.  TRANSIMS predicts trips for
individual households, residents, freight loads, and vehicles rather
than for zonal aggregations of households.  The Household and
Commercial Activity Disggregation Module creates regional synthetic
populations from census and other data.  Using activity-based methods
and other techniques, it produces a travel representation of each
household and traveler.  These two submodules, Synthetic Populations
and Activity Demand and Travel Behavior, are described separately in
the following pages.

The Intermodal Route Planner involves using a demographically defined
travel cost decision model particular to each traveler.  Vehicle and
mode availability are represented and mode choice decisions are made
during route plan generation.  The method estimates desired trips not
made, induced travel, and peak load spreading.  This allows evaluation
of different transportation control measures and travel demand
measures on trip planning behaviors.

The Travel Microsimulation executes the generated trips on the
transportation network to predict the performance of individual
vehicles and the transportation system.  It attempts to execute every
individual's travel itinerary in the region.  For example, every
passenger vehicle has a driver whose driving logic attempts to execute
the plan, accelerates or decelerates the car, or passes as appropriate
in traffic on the roadway network.

The Travel Microsimulation produces traffic information for the
Environmental Models and Simulations to estimate motor vehicle fuel
use, emissions, dispersion, transport, air chemistry, meteorology,
visibility, and resultant air quality.  The emissions model accounts
for both moving and stationary vehicles.  The regional meteorological
model for atmospheric circulation is supplemented by a model for local
effects.  The dispersion model is used for directly emitted
contaminants and handles both local and urban scale problems.  The air
chemistry model includes dispersion, but is designed to deal with
secondary pollutant production on larger scales.

The following pages describe each TRANSIMS module in greater detail. 
The last page describes our interim operational capability approach to
TRANSIMS development.


Synthetic Populations

Purpose

The Synthetic Populations submodule creates a regional population
imitation whose demographics closely match that of the real
population.  The imitation's households also are distributed spatially
to approximate the regional population distribution.  The synthetic
population's demographics are provided to the Activity Demand
submodule to derive individual and household activities requiring
travel.  The household locations determine travel origins-and
destinations.

Background

The underlying TRANSIMS theme is that individual behavior and their
interactions, as constrained by the transportation system, generate.
the transportation system's performance.  To effect that performance
in a

                                Page 2



TRANSIMS: TRansportation ANalysis and SIMulation System       May 1995

simulation, individual behavior must be modeled.  The Synthetic
Populations submodule begins the process of creating those
"individual" models.

Approach

The 1990 census data, including the Census Standard Tape File 3 (STF-
3) and the Public Use Microdata Sample (PUMS), are used to develop a
baseline population.  We create distinct households for each-census
tract or block group area (we use census tracts, but the same methods
could construct populations for block group areas).  The procedure
involves four stages.  First, for the census tract in question the
census summary tables from STF-3 and the corresponding PUMS sample are
grouped by family and non-family households.  Second, for each
household type, we construct a multiway table of all demographics
available from STF-3.  In the third step we create households by
random selection (according to the constructed multiway table
probabilities) of similar households in the PUMS sample.  The last
stage ages the population to the desired date.  A more detailed
procedure follows for family households.  The procedure is similar for
nonfamily households and group quarters.

The STF-3 demographic summary tables for family households are (1)
householder race by household class (a combination of household type
and presence and age of children), (2) householder age, (3) family
income, and (4) the number of workers in the family.  We use these
four summary tables and the corresponding PUMS sample to create the
five-dimensional multiway table of probabilities for each combination
of the five demographic variables.  The methods used to construct the
multiway table include iterative proportional fitting, maximum
likelihood and minimum chi-square.  Because the PUMS is a sample from
multiple census tracts, iterative proportional fitting is not a
statistically correct procedure for this construction.  However,
preliminary studies indicate that this procedure performs rather well
and may be more practical than trying to use either maximum likelihood
or minimum chi-square, each of which require optimizing hundreds of
parameters.

The multiway table constructed for the five family household
demographics is sparse.  For example, the table for census tract 1.07
from Albuquerque, NM has 5880 cells but the corresponding PUMS
contains only 2213 families.  Therefore most constructed multiway
table cells contain zero counts and hence zero probabilities.  Small
probabilities, representing a fractional count, could replace these
zero probabilities and may improve the results.  Additionally,
demographics other than the five in the constructed table, such as the
number of persons in each household, could be imputed from the PUMS. 
Imputation of additional demographic variables should be investigated
to determine if the resulting population of family households better
mimics that of the census tract.

Two options exist to select the number of families for each multiway
table category.  First, the actual number of families in the census
tract may be multiplied by the table cell probabilities and rounded to
determine the number of families with each demographic combination. 
Or, the families may be drawn at random according to the multiway
table probabilities.

For the last stage, we are collaborating with Professor R. Kitamura to
age the population.

We must develop methods to validate and verify these household
construction procedures.  We can check the constructed population of
households with existing census data.  For example, STF-3 contains a
summary table for the total number of persons in family households. 
Because the total number of people is not

                                Page 3



TRANSIMS: TRansportation ANalysis and SIMulation System       May 1995

controlled in the construction of family households, we can use this
summary as one validation of the resulting population.  For validation
and verification of household characteristics that are not. in STF-3
(for example the number of vehicles by the number of people in the
households), we can construct a synthetic collection of "census
tracts" and corresponding synthetic "PUMS" samples by considering PUMS
samples as "census tracts" and combining approximately 20 neighboring
PUMS samples.  The resulting population of approximately 100,000
people then is sampled to create the "PUMS" for the constructed
"census tracts".


Activity Demand and Travel Behavior

Purpose

The purpose of the Household and Commercial Activity Disaggregation's
Activity Demand submodule is to generate household activities,
activity priorities, activity locations, activity times, and mode and
travel preferences.  The activities and preferences are functions of
the household demographics created by the Synthetic Population
submodule.  The Intermodal Route Planner uses the activities and
preferences to determine individual's and load's trip plans for the
region.

Background

Households and businesses have activities that must be, or are desired
to be, performed during the day.  Many of these activities require the
transportation system to move a load (individual or freight) to a
certain place at a certain time.  Thus, activity demand generates
travel demand.  These activities and how they are performed depend on
the demographics of the household and its individuals, or on the
nature of the business.

At this time we are concentrating on traveler activities; modelling
commercial load activities will be difficult until more data is
available on shipping.  Traveler activities are aggregated at a low
level into household activities.  Household activities then are
estimated from probability distributions dependent on the household
demographics.  These demographics include the ages of the inhabitants,
the household family type (single, married couple, married couple with
small children, married couple with older children, etc), the
household income, the number of cars on the household, and the members
of the household who can drive.  These demographics are produced by
the Synthetic Populations submodule.

Approach

Models will be developed for the generation of activities from
household demographics.  There is a considerable literature pertaining
to such models, and we will use the existing research findings as much
as possible.  Similarly, models will be developed for generating mode
preferences from individual demographics, again using existing
research where appropriate.  Local surveys will be necessary to
calibrate these models.

Based on the input demographics, a list of travel activities will be
produced for each household.  These activities will be designated as
"household" or "individual" activities, the distinction being that
"individual" activities require the participation of a specific person
in the household (for example, going to a workplace), whereas
"household" activities may be divided or shared among the people in
the household (for example, going shopping).  Associated with each
activity is a set of parameters defining the activity importance, the
activity duration, and a time interval during which the activity must
be performed, if it

                                Page 4



TRANSIMS: TRansportation ANalysis and SIMulation System       May 1995

is performed at all (for example, work is mandatory, so a work trip
must be made, but a shopping trip is typically not as important and
may be skipped on a given day if the scheduling is too difficult). 
Locations, such as the household address and the workplace and school
addresses, will be provided for mandatory activities.  Locations of
other activities (shopping) are not specified- the planner will choose
these from a list for the locality.

We also will model travel behavior, that is, the propensity of
individuals or households with given demographic characteristics to
exhibit certain travel patterns or behavior.  Travel behavior is more
global than is travel demand.  Hence we may use models given in the
literature that are based on global data sets such as the National
Personal Transportation Survey.  For example, the choice of driver in
a family household given the household's socioeconomic characteristics
may be similar (in a probabilistic sense) across all sections of the
country.  At a minimum, we will model mode preferences for every
individual and the travel characteristics in a family household.  In
particular, we need to model carpooling and travel arrangements (for
example, which person shops) within family households.  We will
consider models of car availability rather than ownership.

Although extremely difficult to accomplish because of insufficient
data, we must develop additional activity demand models for non-
passenger trips, including freight, service, and commercial trips.

All models will be probabilistic in nature in that for a given set of
demographics, a distribution of activities will be produced.. The
actual activities passed to the planner will be chosen by randomly
selecting from this distribution.  Logistic regression, classification
and regression trees and neural nets are examples of tools that will
produce these distributions.


Intermodal Route Planner

Purpose

The Intermodal Route Planner generates regional individual activity-
based travel demand.

Background

A "load" is a traveler or a commodity.  A trip plan is a sequence of
modes, routes, and planned departure and arrival times at the origin,
destinations), and mode changing facilities projected to move the load
to its activity locations.  We assume that travel demand derives from
a load's desire or need to perform activities.  The HCAD provides the
Planner with disaggregated activity demand and travel behavior.  The
Planner assigns activities, modes, and routes to individual loads in
the form of trip plans.  The individual trip plans are input to the
Travel Microsimulation for its analysis.

Trip plan selection is related directly to a load's desire to satisfy
individual (or in the case of freight, corporate) goals.  Goals
measure a trip plan's acceptability and depend on the load's
socioeconomic attributes and trip purpose.  Typical goals include
cost, time, and distance minimization, and safety and security
maximization.  The load's objective is to minimize the deviations from
these goals.

Mode and route preferences also are important in the Planner.  A
preference is the inherent partiality or bias a load has for a
particular mode or route.  Typical preferences include departure time,
origin-destination directedness, and

                                Page 5



TRANSIMS: TRansportation ANalysis and SIMulation System       May 1995

congestion avoidance.  These preferences reduce the Planner's
(activity, mode, and route) solution space and offer significant
computational savings.

Approach

The travel demand problem is formulated as a mathematical program
based on a multi-goal objective function.  The Planner's solution
method has four phases as shown in the figure.  In the first three
phases, performed independently for each load, the individual's travel
behavior preferences are adjusted iteratively to satisfy the travel
goals.  After every load has a feasible, or reasonable, trip plan, the
fourth phase superimposes all trip plans on one another in space and
time.  The network characteristics then are updated based upon the
projected interaction of all trip plans.  The method then returns to
the individual load trip planning phases and the entire process is
repeated.  The iterative process terminates when either all trip plans
are feasible or after some criteria are satisfied.


Click HERE for graphic.


The trip plans are evaluated with respect to the individual's travel
goals.  For some loads, all travel goals are satisfied.  For others,
the plans may not satisfy some or any of the individual's travel
goals.  In these instances, those plans that minimize the goal
deviations will be retained.  The result will be trip plans that
represent regional travel demand and its activity, mode, and route
choice variability.

The Planner accounts for latent travel demand by travel goal
deviations and unplannable activities.  Goal deviations measure the
load's dissatisfaction with the transportation system.  If one must
travel longer than desired to get to work, the deviation from one's
goal travel time represents an unsatisfied demand.  If a
transportation infrastructure change is considered, its value can be
measured by the reduction in the population's travel goal deviations. 
Unplannable activities are those that the load must forego because of
the transportation system's deficiencies.  The Planner will attempt to
route the load's low priority activities after all high priority
activities have been scheduled.  These activities will be planned only
if the current and subsequent high priority trip plans are feasible. 
The second latent demand measure is these unplannable low priority
activities.  Thus, a measure of effectiveness of any proposed facility
change is the change in unplannable activities and the resulting trip
plans.  Loads plan their trips based on activity requirements,
knowledge of the transportation system, travel goals, and assumptions
about other load trip plans.  The travel-planning decisionmaking
process can be achieved iteratively with the Planner and
Microsimulation.  Feedback loops between both modules mimic
individual's real travel process of plan, execute, and replan.

The Intermodal Route Planner model generates activity-based travel
demand at the individual load level.  By (1) receiving activity demand
and travel behavior from the HCAD, (2) providing individual trip plans
to the Travel

                                Page 6



TRANSIMS: TRansportation ANalysis and SIMulation System       May 1995

Microsimulation, and (3) obtaining feedback from the Microsimulation
from the trip plan execution, the Planner forecasts regional travel
demand over multiple time scales.


Travel Microsimulation

Purpose

The TRANSIMS Travel Microsimulation module mimics the movement and
interactions of travelers thoughout a metropolitan region's
transportation system.  For this discussion, traveler refers both to
human travelers as well as freight loads, etc.  The Intermodal Route
Planner provides a trip plan to each traveler that he then attempts to
execute on the transportation network.  In the process he interacts
with other travelers and the transportation system.  The combined
traveler interactions produce emergent behaviors such as traffic
congestion.

Background

The TRANSIMS Travel Microsimulation models many transportation modes
including automobiles, trucks, buses, light rail, commuter rail,
bicycles, and pedestrians.  Thus, the microsimulation includes
roadway, transit, rail, bikeway, and pedestrian networks.  In the
following discussion, we illustrate the TRANSIMS microsimulation with
roadway transportation examples because of its high use, complexity,
and importance to air quality.  The roadway network includes freeways,
highways, streets, ramps, turn lanes, grades, and intersections
(signalized or unsignalized).  In executing their trip plans, vehicle
drivers accelerate, decelerate, turn, change lanes, pass, and respond
to other vehicles and signs and signals.  Drivers exhibit behavior
between aggressive and passive.  Vehicles have weight and acceleration
and deceleration characteristics.  Analysis requirements determine the
necessary microsimulation detail.

Increasing the microsimulation's detail increases its behavioral
representation of real transportation systems, but it also increases
its computational burden.  The representation quality is called the
model's fidelity.  One goal is to find the minimum computational
detail necessary to produce the fidelity needed for specific analyses. 
This minimum computational detail is called critical complexity.  A
hybrid technique uses high-fidelity microsimulations for areas where
detailed results are needed and low-fidelity, fast-running
microsimulations for areas where there is less interest.  This hybrid
microsimulation requires matching the microsimulations at their
boundaries.

Approach

We are studying two approaches to the microsimulation.  Applications
and investigations with the two approaches will form the basis for
deciding which approach will be used in later TRANSIMS versions.

In the first approach, the links (roadway segments) of the network
representation of the transportation system are a continuous domain. 
A vehicle can be positioned along any point on the segment.  The
vehicle driver evaluates the current situation and decides his next
action that advances the vehicle to a new position.  The vehicle and
driver objects retain their characteristics as they move through the
network.

The second approach is to use a cellular automata (CA)
microsimulation.  CA traffic models divide the transportation network
into a finite number of cells.  For example, each cell could be
approximately the length of a vehicle.  At each

                                Page 7



TRANSIMS: TRansportation ANalysis and SIMulation System       May 1995

time step of the simulation, each cell is examined for a vehicle
occupant.  If a vehicle is present in the cell, the vehicle is
advanced to another cell according to a simple rule set.  A CA
microsimulation is low fidelity, but provides a means to simulate
large numbers of vehicles and maintain a fast execution speed. 
Increasing the fidelity by decreasing the cell size, adding vehicle
attributes, and expanding the rule set results in slower computational
speed.  We will explore the fidelity and performance limits of the CA
microsimulation to establish the computational detail necessary to
meet the analysis requirements.

The primary Travel Microsimulation output is the second-by-second
location of each traveler.  Analysis of the primary output yields
additional information such as velocities, accelerations,
decelerations, average speeds, average travel times.  Plots such as
travel time vs traffic density, traffic flow vs traffic density,
vehicle positions vs time; and animation such as vehicle movements on
network segments also can be generated.  Data on positions,
velocities, accelerations, decelerations, and vehicle total travel
time also is input to the emissions model to determine effluents at
spatial locations thoughout the region.


Environmental Models and Simulation

Purpose

The purpose of the environmental module is to translate traveler
behavior into consequent air quality, energy consumption, and carbon
dioxide emissions.  The environmental module will use information from
the planner and the microsimulation and it will support the analyst's
toolbox.  It also could provide information on fog to the
microsimulation.

Background

Transportation systems play a significant role in urban air quality,
energy consumption and carbon-dioxide emissions.  Recently, it has
been found that current systems for estimating emissions of pollutants
from transportation devices lead to significant inaccuracies.  When
these inaccuracies are coupled to air quality models and limited
meteorological data, it is difficult to tell whether the most
appropriate path is being taken to achieve air quality goals.  Most
existing emission modules use very aggregate representations of
traveler behavior and attempt to estimate emissions on typical driving
cycles.  However, recent data suggests that typical driving cycles
produce relatively low emissions with most emissions coming from off-
cycle driving, cold-starts, and evaporative emissions.  Furthermore,
some portions of the off-cycle driving such as climbing steep grades
are apt to be correlated with major meteorological features such as
downslope winds.  These linkages are important, but they are not
treated systematically in the current modeling systems.

Approach

We plan to develop a system of linked modules including: (1) emissions
modules, (2) regional-scale meteorological models, (3) microscale
meteorological modules (street-canyon), (4) dispersion and transport
modules, and (5) air chemistry (airshed) modules.  The development of
these modules will build upon efforts already underway at Los Alamos
and the larger air quality community.  At Los Alamos we have
experience with regional scale (urban metropolitan area)
meteorological models that can describe airflow and turbulence driven
by terrain and land use without the requirement of many local
measurements.  We have dispersion and transport models that can take
the information from the meteorological model and describe the
dispersion that occurs in complex terrain.

                                Page 8



TRANSIMS: TRansportation ANalysis and SIMulation System       May 1995

We also have available an airshed model (through collaboration with
investigators at Carnegie Mellon University) that uses the
meteorologicalmodule outputs with emissions to describe the air-
chemistry in a metropolitan area.  We currently are developing an
appropriate emissions model and a simplified microscale meteorological
module.

The emissions module must take the information on individual traveler
behavior and produce NOx, CO, aerosol, and hydrocarbon emissions for
input to both the airshed and dispersion models.  We will have
individual vehicle motion available on one-second intervals from the
Travel Microsimulation.  We also will have traveler's plans that
describe when vehicles are used and when and where they are
stationary.  In our preliminary emissions module we will integrate an
existing modular emissions model, VEHSIME, with existing models for
evaporative emissions, cold-start emissions, and high-emitting vehicle
emissions.  We also will develop a preliminary micro-scale
meteorological model based on adding street-canyon eddies to our
dispersion module.  At the same time we are adding a capability to
treat fog and clouds to our meteorological model.

In the longer term the preliminary emissions module will be replaced
by a physics and chemistry-based model being developed at the
University of Michigan with additional modifications being developed
by other studies.  We also will develop or acquire an aerosol
emissions model to be used for both diesel and gasoline engines.  The
air chemistry model will be extended to address organic aerosol
production.

Interim Operational Capabilities

We visited six metropolitan planning organizations (MPOs) (Dallas-Ft. 
Worth, Boston, Portland OR, Oakland, Chicago, and Denver).  We
presented the overall TRANSIMS approach and obtained information on
their responsibilities, transportation and air quality issues,
processes for carrying out their activities, potential applications
for TRANSIMS, their resources, and user feedback on what TRANSIMS
should do for them.  We are using this and other information to
develop detailed requirements and specifications for the TRANSIMS
architecture and design.

To provide greater, more timely interaction and feedback from the
TRANSIMS user community, we have formulated an approach for TRANSIMS
development in which we will develop an interim operational capability
(IOC) for each major TRANSIMS module during the five-year program. 
When the IOC is ready, we will complete a specific case study to
confirm the IOC features, applicability, and readiness.  We will
complete the specific case study with the collaboration of the staff
of a selected MPO.  This approach should give us quicker feedback from
the user community and provide interim products, capabilities, and
applications.  This approach maintains our goal of an integrated
framework for predicting individual travel behavior and for supporting
transportation planners from travel demand forecasting to assessments
of transportation system modifications.

The Travel Microsimulation will be the first IOC, with the goal of
having it ready for testing in August 1995.  As this IOC is developed,
we will work with the selected MPO to identify studies that the IOC
should support.  The second IOC will integrate the air quality
analysis capability of TRANSIMS with the Travel Microsimulation. 
Again, the IOC development will be driven by studies identified as
important to the users and will be followed by the specific case
studies.  We intend to issue subsequent IOCs for the Intermodal Route
Planner and for the Household and Commercial Activity Disaggregation
modules.  These IOCs may be

                                Page 9



TRANSIMS: TRansportation ANalysis and SIMulation System       May 1995

standalone modules, but will be capable of integration with the other
TRANSIMS modules.  The case studies will demonstrate the integrated
package.

We have developed several microsimulation versions that successfully
modeled traffic behavior.  In the Albuquerque Demonstration, we
simulated traffic on two interstates and at their intersection using
vehicle objects on a continuous road network.  For the IVHS Incident
Detection Testbed, we extended this capability to include many lanes,
signalized intersections, incidents, and additional driver behavior. 
Our single-lane cellular automata (CA) simulation exhibited traffic
congestion, shock waves, and roadway capacity.  An enhanced CA version
runs on distributed processers and dynamically redistributes the
computational load during execution.  Another CA version with multiple
lanes and freeway interchanges simulated critical traffic volumes on
the 48,000 km of the entire German Autobahn using a 64-node partition
of the Intel Paragon parallel processor machine.  This broad
microsimulation experience places us in an excellent position to take
the best of what we have learned to develop the first Travel
Microsimulation IOC.

To the extent possible the first Travel Microsimulation IOC will rely
on data currently existing at MPOs.  We will develop techniques to
incorporate these data readily into the TRANSIMS methodology.  The IOC
will be supported with the capability to let the user adjust the input
as necessary for his analyses.  Similarly the user will be supported
with several output options to support his analyses.  We will develop
these features with input and feedback from the potential users.

From our MPO visits we had sufficient information to decide which
regional area to use for our first case study.  We considered numerous
factors, but the major ones were: staff and management interest, staff
capabilities, and data availability.  The ratings were very close and
the decision was difficult, but we decided to work with the North
Central Texas Council of Governments (NCTCOG) (Dallas-Fort Worth) for
the first case study using the Travel Microsimulation IOC.  NCTCOG has
enthusiastically supported this decision.

The input, feedback, and interest of all the MPOs we visited have been
very helpful in establishing the scope and requirements for TRANSIMS. 
We anticipate that the future IOCs will be developed in association
with case studies for other metropolitan regions.  We also anticipate
that there may be other supporting developmental efforts in which
collaboration with a metropolitan region would be helpful both to the
TRANSIMS effort and the MPO.

Acknowledgements

Although the authors accept responsibility for the TRANSIMS
descriptions in this document, there are numerous others who have
contributed, or are contributing, to the TRANSIMS Project.  Without
their support, ideas, and input, this document would not be possible. 
Chris Barrett and Darrell Morgeson share responsibility for the
TRANSIMS vision and initiative.  Other contributors include: Kathy
Berkbigler, Mike Brown, Brian Bush, John Davis, Deborah Kubicek, Verne
Loose, Mike McKay, Jack Morrison, Kai Nagel, Rob Oakes, Steen
Rasmussen, Marcus Rickert, Jay Riordan, Doug Roberts, Paula Stretz,
Steve Sydoriak, Gary Thayer, and Murray Wolinsky.  We also appreciate
the support and feedback of our sponsors at the Federal Highways
Administration, the Department of Transportation, and the
Environmental Protection Agency: Fred Ducca, Ed Weiner, Ron Jensen-
Fisher, Ron Giguerre, and Mark Simons.


                                Page 10





(466.html)
Jump To Top