Skip to Main Content
Peer Review Program
HomeUnited States Department of Transportation, HomeContact UsSite Map
Travel Model Improvement Program - TMIP
About TMIPTMIP ServicesClearinghouse, selectedConferences and CoursesContactsTravel Model DiscussionsTRANSIMSLinks

BROWSE

SEARCH

ORDER DOCUMENTS

NEWSLETTER

Clearinghouse

TMIP Newsletter Archive

TMIPConnection
The Travel Model Improvement Program Newsletter

Issue 2, January 1995

The Travel Model Improvement Program is sponsored by:
U.S. Department of Transportation
Federal Highway Administration
Federal Transit Administration

Table of Contents
Fort Worth Conference
Land Use Modeling Conference Scheduled
Upcoming Conferences Related to TMIP
Frequently Asked Questions Regarding TMIP
FY 1993-94 TMIP Accomplishments
Current TMIP Activities
Listing of Available Reports
TRANSIMS Travelogue

Fort Worth Conference

The Travel Model Improvement Program (TMIP) sponsored a workshop conference August 14 - 17, 1994 in Fort Worth, Texas. Over 100 members of the travel forecasting, environmental and transportation planning, consulting and academic communities participated. The conference program was designed to stimulate discussion on the current travel modeling needs and issues, with the overall objective of establishing an agenda for the TMIP. The conference brought together experts and practitioners to:

The first day of the conference focused on Track B, improvements to existing travel models. Research that had been or was currently being conducted under this Track as well related research being conducted by others was presented during the morning session. Following those presentations, participants divided into workshops to discuss the short term research needs to improve existing models and analytical techniques.

Day two of the conference was directed toward research in Track C, longer term model improvements. During the morning participants heard presentations on TRANSIMS, the new model approach undertaken by the Los Alamos National Laboratory. The Tuesday workshops then focused their discussions on TRANSIMS and other longer term model research needs that must be addressed to perform current and anticipated future planning and policy analyses.

Wednesday workshops concentrated on the issues of deployment, dissemination and education associated with Track A, Outreach, and on the data needed to support continued research in Tracks B and C as well as to meet other data needs outlined under Track D, Data Collection.

The workshops developed extensive recommendations for research to improve the precision, accuracy, and confidence of forecasts from demographic, economic, land use and urban development models.

Research is also needed on data collection strategies and coordination, including panel surveys, stated preference surveys, use of geographic information systems and other new technology, improved survey designs, better use of existing data sources, greater consideration of data on existing conditions for use in support or in lieu of forecasts, and data quality and decay.

Other pressing needs are for increased and improved training and support, more and better documentation, improved dissemination of information on research activities and results, and increased coordination and funding of research.

The sponsors considered the conference highly successful. The TMIP support staff are using the results of the conference to develop work plans for all four tracks of the Program. Proceedings are being prepared as well and are expected to be available early in April 1995.

Summary Recommendations

Workshop participants prepared a summary of priority recommendations for the TMIP which include research to provide a better understanding of:

Table of Contents

Land Use Modeling Conference Scheduled

The Federal Highway Administration, the Federal Transit Administration, the Office of the Secretary of the Department of Transportation, and the Environmental Protection Agency are sponsoring a conference on land use modeling February 19 - 21, 1995 in Dallas, Texas. The conference will recommend actions to improve existing and develop new land use forecasting procedures. The recommendations will include the research and development needed to accomplish those improvements./

Improvements to existing travel models and the new generation of models being developed will require more sophisticated land use forecasts. Simply estimating the number of employees or residents in a zone is not likely to be adequate for activity based travel models, which require detailed information on household demographics and employment characteristics as well as precise locations. Many policy initiatives also require detailed and sophisticated land use/activity forecasts.

Table of Contents

Upcoming Conferences Related to TMIP

Table of Contents

Frequently Asked Questions Regarding TMIP

Who is funding TMIP?

Funding for the Travel Model Improvement Program is provided by the U.S. Department of Transportation, including the Federal Highway Administration, the Federal Transit Administration and the Office of the Secretary; the Environmental Protection Agency; and the Department of Energy.

What is the relationship between Track B, Short-term improvements, and Track C, Long-term improvements?

Track B projects will improve existing models and can be completed in a short time frame. Track C has a goal of producing a new generation of travel demand forecasting, microsimulation, and air quality models in 5 years.

What type of projects will be funded in Track D?

There are two major areas of necessary research: (1) data to support model improvement and development in Tracks B and C, and (2) data needs, collection methods, and analyses for transportation program applications of states and regional planning agencies. The Track D program is just getting underway and over the next few months a comprehensive and coordinated program will be developed to meet these general categories of need.

Why was Los Alamos National Laboratory chosen to develop the new generation of travel/air quality models?

The Los Alamos National Laboratory (LANL) was chosen because of their experience with microsimulation (from defense work) and air quality analysis and their demonstrated ability to apply that microsimulation expertise to transportation analysis. Improved air quality analysis and the need for better information on vehicle operations on the transportation networks to feed the air quality analysis are principal reasons why new travel models are needed.

What kind of computer hardware will be needed to run TRANSIMS?

This question has been a concern to potential users and the LANL project staff. They have recently developed a distributed hybrid microsimulation that can be run on either a cluster of computer workstations or a dedicated computer. For the case of the network cluster, LANL has shown that the computational workload can be dynamically redistributed across the computer network while the problem is running. This capability, when fully developed, will allow users to run TRANSIMS on microcomputers linked by a local area network.

What kind of data will be necessary to run TRANSIMS?

It is still too early in the development of TRANSIMS to answer this question. The LANL project staff is beginning with a very large set of possible data needs, including those used by existing 4-step models as well as others not currently used. Sensitivity analysis and other techniques will be used to determine the most relevant and useful data. This rigorous screening process will no doubt result in a different list of data than currently used. However, at this time, it is not known what or how much data will be required. LANL is sensitive to the need for early notification of the data needs and will provide interim information as available.

Will land use forecasts be prepared for TRANSIMS?

Land use forecasts will be prepared outside of TRANSIMS. Small area land use forecasts will be used to produce a synthetic set of households and population tied to particular household locations.

When will TRANSIMS be available for use? Are there going to be intermediate products from TRANSIMS?

The complete TRANSIMS modeling structure will be complete in 1999, however, LANL is not responsible for producing a version of TRANSIMS which is ready for commercial distribution. Methods for distributing TRANSIMS to the user community are now being researched. Every effort will be made to distribute and train users on intermediate products as they are available.

How can I find out more about TMIP?

An informational brochure which describes TMIP is available. Additionally, this newsletter is published periodically and is designed to provide regular updates on the Program and other related research. A mailing list is being maintained for persons interested in receiving the newsletter. A TMIP Bulletin Board is also being implemented (see discussion under "Current Activities") to enable individuals or agencies to stay abreast of current developments and to exchange information with others regarding travel demand forecasting improvements.

Table of Contents

FY 1993-94 TMIP Accomplishments

Track A – Outreach

Based on the input received at the TMIP conference in Fort Worth, an initial work plan for Track A has been developed. The TMIP Review Panel will be commenting on this work plan in January.

Track B – Near Term Model Improvements

Several projects have been performed or are underway as part of Track B. These include:

Identification of Short Term Travel Model Improvements
This project identified existing methods and applications to improve current urban travel models in the near term. The identified improvements are methods and procedures that have been implemented in some urban areas, but may not be well known. The results of this project are documented in the report "Short Term Travel Model Improvements", August 1994. (see Listing of Available Reports section for availability)

The Effects of Land Use and Travel Demand Management Strategies on Commuting Behavior
This project examined the effects of urban design and land use characteristics on individual transportation choices. Results of the study are documented in the report "The Effects of Land Use and Travel Demand Management Strategies on Commuting Behavior", July 1994. (see Listing of Available Reports section for availability)

Freight Modeling Conference
Contracts for the development of a quick response freight forecasting manual and sponsorship of a freight modeling conference (See "Upcoming Conferences") have been let, and work is underway.

Track C – Long Term Model Improvements

A new model, TRANSIMS, is being developed at the Los Alamos National Laboratory (LANL) to link the previously separate travel forecasting, microsimulation, and air quality models. LANL staff have visited six MPOs to gain an understanding of their needs and responsibilities and to solicit comments on the proposed TRANSIMS structure and design. Both ISTEA and CAA legislation and supporting documents have been reviewed, and the report, "TRANSIMS Model Requirements as Derived From Federal Legislation", provides an overview of the requirements TRANSIMS must satisfy. A report, "TRANSIMS: Operational Description of the Software Framework", summarizes the requirements and provides a high order structure for the LANL staff.

Other accomplishments by LANL include study of current air pollution dispersion models and the integration of those models and the current group of meteorological models with the TRANSIMS traffic model.

Track D – Data Collection

The Travel Survey Manual update is underway is to develop a new manual of travel survey techniques based on the advances made during the past two decades and the experiences of a number of recent surveys conducted around the U.S. The update is expected to be available in April 1995 from the U.S. Department of Transportation.

Table of Contents

Current TMIP Activities

TMIP Clearinghouse

A clearinghouse focusing on travel modeling research efforts, including recently completed projects, those currently underway, and anticipated projects, is currently being organized. Each reference will include the title, contact name, phone number and a 1 or 2 paragraph description. Papers, research projects, work programs, survey model design and training programs, will all be included in this clearinghouse. This resource will be available in a variety of formats to reach the greatest number of people possible.

The formats will include distribution on the bulletin board (see description below), as a machine readable literature database, as a printed report, and highlights that will be published in the TMIP newsletter.

Bulletin Board

The information collected for the clearinghouse will provide the starting point for the bulletin board. A systems administrator will be available to help users find information. Topic folders, keywords, and other searching techniques will be used to ease the task of finding relevant information.

Table of Contents

Listing of Available Reports

Addresses for report contacts:

Kim Fisher
c/o Federal Highway Administration
400 7th St SW, HEP-22, Room 3232
Washington, DC 20590
Phone: (202) 366-4054, Fax: (202) 366-3713

Lynette Engelke
Texas Transportation Institute
201 E. Abram Street, Suite 600
Arlington, TX 76010
Phone: (817) 277-5503, Fax: (817) 277-5439

Norman Paulhus
Technology Sharing Program
U.S. DOT
400 7th Street, SW, M-453
Washington, D.C. 20590

Table of Contents

TRANSIMS Travelogue – January 1995

TRANSIMS TRAVELOGUE describes current activities within the TRANSIMS project.

What is TRANSIMS?

The TRansportation ANalysis and SIMulation System (TRANSIMS) is one project in the multi-track Travel Model Improvement Program sponsored by the U.S. Department of Transportation and the Environmental Protection Agency. The TRANSIMS project is a major effort to develop new, integrated transportation and air quality forecasting procedures to satisfy the Intermodal Surface Transportation Efficiency Act and the Clean Air Act and its amendments. Development of TRANSIMS is being led by the Los Alamos National Laboratory (LANL).

TRANSIMS is a set of integrated analytical and simulation models and supporting data bases. The TRANSIMS methods deal with individual behavioral units and predicts trips for individual households, residents and vehicles rather than for zonal aggregations of households. TRANSIMS also predicts the movement of individual loads of freight. A regional microsimulation executes the generated trips on the transportation network and predicts the performance of individual vehicles and the transportation system. Motor vehicle emissions are estimated using traffic information produced by TRANSIMS. A major advantage of TRANSIMS for air quality analysis is the detail it provides regarding individual motor vehicle operation.

Previously microsimulation at the detail performed by TRANSIMS required very fast, high capacity computers. The TRANSIMS team has employed computational techniques to overcome such limitations and has demonstrated significant performance gains through distributed computing and other methods. Furthermore, it is expected, that by the time development of TRANSIMS has been completed, technical capabilities will have advanced sufficiently for greater computing capacity to be more readily available.

Requirements Analysis

After researching the Clean Air Act Amendments of 1990 and the Intermodal Surface Transportation Efficiency Act of 1991, The LANL team prepared wrote a high-level issues and requirements document entitled "TRANSIMS Model Requirements as Derived from Federal Legislation." That document translates the analytical requirements stipulated in the recent federal legislation into TRANSIMS design requirements. The key requirements are: representation of individual travelers and freight items; databases and simulations that support behavioral and choice/utility models that are sensitive to travel time constraints, preferences, travel behavior, cost, and perceived quality of service; environmental models, and system simulations. Furthermore, component models should be sensitive to detailed representations of environmental and vehicle characteristics correlated to both the geographic and demographic distribution of travelers. The models must be capable of simulating continuous traffic, transit, freight, bike, and pedestrian travel over the course of extended periods (both peak and non-peak), different days of the week, months, and seasons.

The LANL team is now identifying requirements pertaining to the software and hardware environments in which the computer code will be implemented. They are also working on an overall plan for systematic identification of requirements from other sources.

TRANSIMS Operational Description

The LANL teams has prepared the document entitled "TRANSIMS: Operational Description of the Software Framework," which sets the context for selection of the application projects from which to derive detailed software functional requirements, the associated software specification documents, and the detailed design documents for components of the TRANSIMS environment. This document presents examples of analyses that will be supported by TRANSIMS, describes its conceptual foundation and analytical rationale, and describes the major high-level components of the TRANSIMS architecture. The major components are: the Household and Commercial Activity Disaggregation Module, the Intermodal Route Planner, the Traffic Microsimulation, and the Environmental Simulation. These components are supported by a Land Use and Demographic Representation and a Transportation Network Representation. Interfaces also must be provided for a Land Use and Demographic Forecast capability and a Transportation Infrastructure Change capability, both outside the scope of the TRANSIMS effort. Finally, an Analyst Toolbox supports the user in problem and analysis design, statistical analysis, sensitivity analysis, and visual presentation of the results.

High-Speed Microsimulation

The LANL team has examined techniques to overcome the computational difficulties of calculating the second-by-second travel behavior of each model traveler throughout the day. Methods for attacking this problem include: higher performance computers, distributed processing, critical complexity, hybrid simulation, and fast algorithms.

When and where possible, the team has considered computational algorithms for highly parallel processor computers. Because most TRANSIMS users do not have access to such capabilities, methods for partitioning the microsimulation computational workload among the distributed processors of a local area network and even more broadly on worldwide networks have been developed. Furthermore, the team has improved the performance by incorporating load balancing that shifts the effort when one or more processors are over- or under-utilized.

Critical complexity, an overriding concern throughout the TRANSIMS development, determines the level of model detail, and no more, necessary to produce the emergent behavior being studied. The work thus far is exemplified by the characteristic traffic dynamic behavior that emerges from very simple cellular automata models for vehicular traffic. Cellular automata traffic models divide the transportation network into a finite number of cells. For example, each cell could be approximately the length of a vehicle. Then at each time step each vehicle is moved from one cell to another according to a simple rule set.

In a hybrid simulation the simulation detail changes between regions. Thus, most of a metropolitan area may be modeled with the fast running cellular automata model that interfaces with a much more detailed analysis of a specific corridor or locale. A major difficulty being addressed in hybrid simulations is matching the detailed simulation requirements with the leanness of the fast-running simulation.

An example of the fast algorithms examined is replacing rule-based driver behavior algorithms with trained neural networks that rapidly compute a driver's next action.

TRANSIMS Program Plan Changes

After visiting six metropolitan planning organizations (MPOs) (Dallas-Ft. Worth, Boston, Portland OR, Oakland, Chicago, and Denver) and observing firsthand the differences in analytic procedures and needs, the team determined that the initial approach of using two major applications to drive TRANSIMS development would not provide timely interaction and feedback from the user community. Therefore, plans were altered to pursue the same end result but with more interim products, capabilities, and applications.

The new approach is to develop an interim operational capability (IOC) for each major TRANSIMS component described previously in the TRANSIMS Operational Description. When each IOC is ready, a complete specific case study will be conducted to confirm the IOC features, applicability, and readiness. Specific case studies will be conducted with the collaboration of the staff of a selected MPO. This approach should provide quicker feedback from the user community.

The Traffic Microsimulation will be the first IOC, with the goal of having it ready for testing in August 1995. As this IOC is developed, studies that the IOC should support will be identified. The second IOC will integrate the air quality analysis capability of TRANSIMS with the microsimulation. Again, the IOC development will be driven by studies identified as important to the users and will be followed by a specific case study. The team intends to issue subsequent IOCs for the Intermodal Route Planner and for the Household and Commercial Activity Disaggregation modules. These IOCs may be standalone modules, but will be integrated with the other TRANSIMS modules. The case studies will demonstrate the integrated package.

This change in approach does not alter the goal of an integrated framework for predicting individual travel behavior and for supporting transportation planners from travel demand forecasting to assessments of transportation system performance.

First IOC: Microsimulation

Several microsimulation versions that successfully modeled traffic behavior have been developed. In the Albuquerque demonstration, the team simulated traffic on two Interstate Highways and at their interchange using vehicle objects on a continuous road network. For the IVHS Incident Detection Testbed this capability was extended to include many lanes, signalized intersections, incidents, and additional driver behavior. A single-lane cellular automata (CA) simulation exhibited traffic congestion, shock waves, and roadway capacity. An enhanced CA version runs on distributed processors 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 micro-simulation experience places the LANL team in an excellent position to use the best of what has been learned to develop the first traffic microsimulation IOC.

To the extent possible the first traffic microsimulation IOC will rely on data currently existing at MPOs. Techniques will be developed to incorporate these data readily into the TRANSIMS methodology. The IOC will let the user adjust the input as necessary for analysis. Similarly the user will be supported with several output options. These features will be developed with input and feedback from the potential users.

NCTCOG: First Case Study Region

The overall TRANSIMS approach was presented to the six MPOs, and information was obtained 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. This and other information are being used to develop detailed requirements and specifications for the TRANSIMS architecture and design.

When the TRANSIMS approach was changed, sufficient information was available from the MPOs to decide which regional area to use for the 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 the team decided to work with the North Central Texas Council of Governments (NCTCOG) (Dallas-Fort Worth) for the first case study, the Traffic Microsimulation IOC. NCTCOG has enthusiastically supported this decision.

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

Environmental Modeling

Air quality efforts have been devoted to five major tasks: the study and improvement of the current vehicle emissions models; the study of the current dispersion models for vehicle air pollution sources; the integration of the current suite of meteorological and air pollution dispersion models with the TRANSIMS Traffic model, the emissions model, and possibly a microscale air flow model; the preparation of a summary of literature relevant to air pollution modeling of vehicle sources; and the preparation of a TRANSIMS Environmental Modeling briefing.

To model air pollution concentrations successfully, TRANSIMS traffic output parameters (e.g., vehicle acceleration, speed, engine temperature) must be correlated to vehicle emission rates. Numerous articles on the current US EPA emissions model (Mobile 5), vehicle emissions experiments, and vehicle driving cycle studies have been collected. Various experts on vehicle emissions have been contacted (e.g., John German of the Environmental Protection Agency, Pablo Cicero-Fernandez of the California Air Resources Board). A protocol for implementing improvements to the current emissions model have been defined.

In the process of investigating on-going U.S. work on calculating real-time emissions for vehicles, the team obtained (1) data from the more relevant California Air Resources Board measurements of real-time emissions from vehicles operating under realistic high engine speed and load conditions; (2) a report on a South Coast Air Quality Management District project at UC Riverside to develop a program for obtaining traffic flow patterns, describing driving behavior, and estimating vehicle emissions; (3) a copy and a user's manual of a Radian Corporation program to calculate automobile evaporative emissions; (4) a Radian paper describing a method to calculate cold start emissions; and (5) a number of other driving behavior studies for specific conditions, such as freeway on-ramps and in different cities. The team has also explored the possibility of Los Alamos involvement in the National Cooperative Highway Research Program's three-year study of real-time vehicle emissions. They have an agreement in principle with Sierra Research Institute, who studied driving behavior in California, to obtain their code to calculate real-time light-duty vehicle emissions and use it as a preliminary vehicle emissions module of TRANSIMS. To obtain a thorough understanding of current vehicle air pollution issues, they have produced a summary paper of articles concerning highway and urban-canyon models and measurements, emissions measurements, and driving cycle experiments.

As a first cut, the team will use current vehicle emissions studies to correct the Mobile emissions model for high accelerations, cold engine starts, etc. Currently, several research groups (e.g., US EPA, CARB, SAI, GM, Ford, Georgia Tech) are actively engaged in experimental research in this area. They intend to improve our formulations as results become available. Note that the driving cycle will be obtained directly from the TRANSIMS traffic model.

Further Information

For further information about the TRANSIMS program, please contact:

Dr. LaRon L. Smith
Los Alamos National Laboratory
Mail Stop F606
PO Box 1663
Los Alamos, New Mexico 87545
Phone: 505-665-1286
Fax: 505-665-0879
E-mail: llsmith@lanl.gov

Table of Contents

To subscribe to this free newsletter send an e-mail to TMIP@tamu.edu or contact Gary Thomas at (ph.) (979) 458-3263, (fax) (979) 845-6001, (mail) Gilchrist, Room 112, Texas Transportation Institute, Texas A&M University System, 3135 TAMU, College Station, TX 77843-3135