TRANSIMS Is Coming!
TRANSIMS is a series of integrated transportation and air quality analysis and forecasting models being developed at the Los Alamos National Laboratory (LANL). The principal components of TRANSIMS are an activity generator, an intermodal route planner, a traffic microsimulation, and an environmental analysis module.
The TRANSIMS design has been driven by legislative and analytical requirements that exceed the capabilities of current models and methodologies. To address these concerns, TRANSIMS will offer transportation planning agencies increased policy sensitivity, more detailed vehicle-emission estimates, and improved analysis and visualization capabilities.
The U.S. Department of Transportation (DOT) has provided most of the funding for the development of TRANSIMS. Additional funding has come from the Environmental Protection Agency (EPA).
The Los Alamos TRANSIMS team has incorporated work from researchers across the country into the new models. The team has also received invaluable assistance from two metropolitan planning organizations -- the North Central Texas Council of Governments in Dallas and Portland (Ore.) Metro. These two organizations have provided the planning environments, data, and experience necessary to develop TRANSIMS.
TRANSIMS is part of the Travel Model Improvement Program (TMIP), a multiyear, multi-agency program designed to improve both the analytical tools and the integration of these tools into the planning process. TMIP has focused on both short-term and long-term improvements to the models and planning procedures. The short-term improvements concentrated on changes to the existing four-step modeling process. TRANSIMS is the long-term effort to redesign the modeling process from the ground up.
As the completion of this new integrated travel demand forecasting and air quality analysis tool nears, the exciting and daunting tasks of producing a commercial version of the software, training users, and deploying TRANSIMS are getting underway. TRANSIMS and the Early Deployment Program, designed to accomplish these tasks, are described in this article.
TRANSIMS' Four Modules
The activity-based travel demand module estimates the number, characteristics, and locations of activities in which individuals will participate during the simulation period. Activities include work, shopping, and recreation. These activity estimates are based on the characteristics of individuals, their households, and available vehicles.
The intermodal trip planning module computes combined route and mode trip plans to accomplish the desired activities. The planner tries to accommodate all the desired activities so secondary activities, such as shopping, may be planned during the routing of a principal trip, such as coming home from work. TRANSIMS maintains the identities and characteristics of individual travelers and of the personal vehicles or transit vehicles used throughout their trips. Activity locations, including home, work, and shopping locations, are identified by specific geographic points of origin and destination.
The traffic microsimulation module uses the intermodal paths developed in the earlier module to perform a regional microsimulation of vehicle interactions. The microsimulation continuously computes the operating status, including locations, speeds, and acceleration or deceleration, of all vehicles and engines throughout the simulation period. Every motor vehicle in the study area is monitored in this manner to identify traffic congestion and emission concentrations.
The emission module provides the data for the air quality analysis. Using the vehicle information generated in the microsimulation module, the emission module forecasts the nature, amount, and location of motor vehicle emissions. The emission information is then used in urban air shed models to predict urban air quality. Data from the emission module will be compatible with the EPA's Models-3 air shed model.
Finally, the selector manages the feedback of information among the first three modules of TRANSIMS. "Manage" in this context refers to decisions such as what percentage of the regional trips should be fed back between modules, which trips should be fed back, how far back the trips should go for replanning, and when to stop iterating.
The modules developed for TRANSIMS contain many significant advances beyond the four-step models:
- Disaggregate Models. TRANSIMS tracks individuals, households, and vehicles, not zonal aggregation of households and employment. TRANSIMS also estimates travel second-by-second throughout the entire day rather than total travel for various periods.
- Simulation. The regional microsimulation uses vehicle interactions to produce operating speeds, intersection operations, and vehicle operating conditions for each vehicle in the system instead of deterministic equations.
- Integrated System. Underlying the development of the TRANSIMS modules has been the effort to tightly couple the functions and data flow among the four modules. While it is currently possible to add a microsimulation or an air quality emission calculation into current forecasting models, such capabilities require considerable postprocessing. Feedback and the careful scrutiny of policy or infrastructure actions are currently difficult to achieve. By integrating all these capabilities in one model, TRANSIMS overcomes many of the limitations that transportation professionals currently experience.
- Highly Detailed Vehicle Emission Estimates. Current emission models use average speed estimates from the current models. The resulting emission estimates are insensitive to traffic conditions and lack precision. In contrast, the TRANSIMS emission estimates are based on the operation of individual vehicles as they interact in roadway traffic.
- Operational Tools. The microsimulation module permits very detailed analyses of traffic operations on the transportation network. This capability could be used to evaluate improvements such as traffic signal plans and ramp metering.
Status and Next Steps
The development of TRANSIMS' core capabilities is complete; however, a case study, set in Portland, Ore., will explore the sensitivity of TRANSIMS software to different types of data.
The work plan for the case study is still being developed, and two sensitivity tests are being considered. The first test would determine the effects of generating synthetic local streets instead of realistically coding every single street in the region. The second test would identify the effects of synthesizing traffic signal plans. Both of these data collection and coding efforts have been very time-consuming and difficult. To test these and other model sensitivities, the Portland staff assembled the actual local street and traffic signal plans to compare with the results of the synthesized data.
In addition, a demonstration of TRANSIMS' capability to simulate intelligent transportation systems (ITS), will take place. The ITS evaluation will require changes and additions to the core software. For a limited number of ITS tools, the LANL staff will show how to design the tests, represent the ITS tools, and assess the impacts on congestion and other measures of effectiveness. Representing the ITS tools will be a critical design and model operation issue.
The Early Deployment Program
Section 1210 of the Transportation Efficiency Act for the 21st Century (TEA-21), passed in May 1998, calls for an Advanced Travel Forecasting Procedures Program. Recognizing the progress made in the development of TRANSIMS, this legislation directs DOT to move to implementation.
The goals of the Early Deployment Program are to:
- Implement TRANSIMS in diverse settings, including areas with differing population, geography, and planning issues.
- Demonstrate TRANSIMS' capability to analyze a wide range of policies.
- Develop a base of experienced TRANSIMS users, including consultants, universities, and local agency planning staff, to provide support as additional areas convert to TRANSIMS.
- Identify problems and opportunities for preparing data sets for TRANSIMS applications.
- Field test early commercial versions of the TRANSIMS software.
In addition to the completion of TRANSIMS, funding was allocated for three specific tasks: (1) development of user-friendly and commercially viable TRANSIMS software, (2) training and technical assistance on the use of TRANSIMS, and (3) financial support for converting to TRANSIMS. These three tasks are still ongoing.
Commercially Viable Software
LANL developed the basic TRANSIMS technology, including the functional modules, the system framework, and examples of its application.
The commercialization process will include licensing TRANSIMS to a vendor/developer team, contracting with that team to modify TRANSIMS, and providing financial support for the modifications. The modified software will use the basic TRANSIMS technology with user-friendly enhancements and other additions to make it more accessible for transportation professionals.
The first step in this effort was the TRANSIMS Commercialization and Deployment Opportunities workshop in Santa Fe, N.M., June 28 to 30, 1999. At this workshop, organizations with an interest in the commercialization process received a technical overview of TRANSIMS; learned the requirements for the commercial software; met with the LANL development team, potential end users, and other organizations with which they could team; and, if they signed a restricted license, received TRANSIMS software for evaluation.
A request for proposals (RFP) was issued by LANL in August 1999, and a pre-bid conference was held in September 1999 to answer any questions on the RFP. The contracting process should be completed by the spring of 2000.
Training and Technical Assistance
TRANSIMS will be accepted if potential users believe that the additional capabilities outweigh the additional work required to use TRANSIMS.
Just as importantly, present and future transportation modeling professionals must receive sufficient training to confidently use TRANSIMS. As TRANSIMS or TRANSIMS-based software becomes a commonly used and accepted analysis tool, this training will naturally be offered by universities around the country, institutes and other training organizations, and software vendors and other private firms. However, for now, the only TRANSIMS training program is being developed by the Texas Transportation Institute (TTI).
TTI's multilevel training program will lead students through an increasingly detailed course to the level of understanding necessary for their interaction with TRANSIMS. The program is funded by DOT, and it is a collaborative effort of TTI, LANL, the academic community, and other transportation professionals. TRANSIMS training will begin in 2000.
Financial Support for Conversion
The final and most critical task in the Early Deployment Program is the selection of participating local planning agencies and their conversion to TRANSIMS. DOT will select local planning agencies to receive financial support to convert from existing travel forecasting procedures to TRANSIMS. That funding could be used for computer purchase, data collection, or hiring a consulting firm to assist with implementing TRANSIMS. Local transportation agencies with responsibility for regional transportation forecasts will be eligible to participate in the program.
Agencies applying to participate should keep in mind that TRANSIMS is an entirely new approach to travel forecasting. Conversion will require training in the basic theories and use of TRANSIMS and in the development of detailed land use data and multimodal networks.
The selection of participants will take place in two phases. In the first phase, agencies will be asked to submit a brief description of their analytic capability, data availability, implementation plans, proposed application of TRANSIMS, coordination with local and state agencies, and their commitment to help disseminate information on their experiences in converting. From these descriptions, DOT will select a smaller group of applicants to participate in the second phase. The second-phase applicants will need to provide additional information, cost estimates, and funding plans. The final selectees will be chosen to represent a wide range of population size, type of transit and auto use, population density, geographic location, and policy concerns.
The participating agencies will be selected at the same time, but not all will be converted simultaneously. This staging will be necessary to ensure that adequate technical support is provided as each agency starts working with TRANSIMS. Staging will also permit those being converted later to take advantage of the experiences gained with each successive deployment.
The final group of program participants will be selected in June 2000.
For more information about TRANSIMS research, the Early Deployment Program, and the Travel Model Improvement Program, visit the TMIP Web site at http://tmip.tamu.edu or call Kim Fisher at (202) 366-4054. At the Web site, you can order TMIP publications and contact members of the TMIP and TRANSIMS staffs via e-mail.
Kimberly M. Fisher is an assistant research scientist in the Urban Analysis Program of the Texas Transportation Institute (TTI). Currently, she is working onsite at the Federal Highway Administration on the Travel Model Improvement Program, the TRANSIMS model development, and the Early Deployment Program. She has been working in transportation planning and travel-demand forecasting for more than 10 years. Prior to working for TTI, she worked for the Denver Regional Council of Governments; URS Engineers in Colorado Springs, Colo.; King County Transportation Planning in Seattle, Wash.; COMSIS Corp., and the Urban Land Institute. She has a master's degree in civil engineering from the University of Maryland.