TMIPConnection
Issue 10, December 1999

The Travel Model Improvement Program Newsletter

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

Table of Contents

TRANSIMS Travelogue

What is TRANSIMS?

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

TRANSIMS is a set of integrated analytical and simulation models and supporting databases. The TRANSIMS methods deal with individual behavioral units and proceed through several steps to estimate travel. TRANSIMS predicts trips for individual households, residents and vehicles rather than for zonal aggregations of households. TRANSIMS also predicts the movement of individual freight loads. A regional microsimulation executes the generated trips on the transportation network, modeling the individual vehicle interactions and predicting the transportation system performance. Motor vehicle emissions are estimated using traffic information produced by TRANSIMS.

The Portland TRANSIMS Studies

Los Alamos is designing the Portland studies to refine and demonstrate the core TRANSIMS technologies and to investigate the effects of the TRANSIMS input data on the output. The study preparations and execution will prepare TRANSIMS-LANL for the TRANSIMS-DOT commercialization process and deployment of the TRANSIMS technologies to metropolitan planning organizations (MPOs).

One goal of the Portland TRANSIMS studies is to complete and demonstrate the basic TRANSIMS technology. A second is to develop methods and techniques that ease the TRANSIMS input data requirements for both forecasts and base year studies. Both efforts are critical to the successful deployment of the TRANSIMS technology, in which MPOs will exercise the methodologies in a forecast setting such as a regional transportation plan on a 15 to 20 year horizon.

The base TRANSIMS technology has the capability to model and simulate the activities and movements of a synthetic representation of every individual in a metropolitan area. This base set of technologies includes modules to develop a population and an activity list, route travelers between activities, and simulate their second-by-second movements on the transportation infrastructure. Feedback methods for stabilizing the simulation and modeling some transportation characteristics are part of the basic TRANSIMS framework. Some base technology characteristics are: multi-modal and intermodal travel, shared rides, high-occupancy-vehicle (HOV) lane usage, populations aligned in age with the study year, generic models of intelligent transportation system technologies, and the capability to represent itinerant travelers and freight as background traffic using given trip tables. Post processing the microsimulation output data allows for estimates of vehicle emissions and accident probabilities; equity analysis or network reliability studies as demonstrated in the Dallas study; “validation” of traffic counts by comparisons with existing data or models; and estimates of traditional transportation statistics such as VMT (vehicle miles traveled).

During the next year, Los Alamos will complete a base year Portland “validation” study for the years 1996-1997. Portland Metro is completing a transportation network and a set of land use data that reflect these years. The Portland study will demonstrate all modules of TRANSIMS including feedback. It will attempt to “duplicate” traffic conditions in Portland for a typical day in 1996. Featured will be multi-modal and intermodal trips (walk, auto, bus and light rail), actuated signals, shared rides, and feedback to stabilize travelers’ activity, mode and route choices. Vehicle emissions and accident probabilities will be estimated for the base year. Because HOV lanes or Intelligent Transportation System (ITS) technologies were not present in Portland in 1996, microsimulation of HOV lanes and feedback to model ITS technologies will be implemented in a second study.

The complete plan for HOV and ITS studies will be developed over the next year. A modified base network will be used in this study. It will include technologies for inter- and intra-household shared rides, and a methodology in the microsimulation for multiple occupancy vehicles to use HOV lanes. At a minimum,ITS technologies to be developed, will include automated tolls and ramp metering.

Study Features

This base-year study will be the first time that the entire system, including the environmental module, is exercised on a real problem. Reports will describe detector technologies to drive actuated signals, iteration schemes for rapid stabilization of the system, methodologies to align the populations, methodologies for actuated signals, the characteristics of multi-modal and intermodal travel, and the realism of the emissions estimates. Some details on the specific technologies to be developed in this study are:

A methodology will age the baseline synthetic population from 1990 to 1996-1997 using projected summary demographics commonly available in planning organizations.

We will feed information from the microsimulation to the intermodal route planner to adjust the travelers’ plans (mode choice, route plans, and their associated travel times) to match their actual executed travel more closely. We also will use feedback to refine their activity times and locations.

The three main TRANSIMS modules include methods to handle both multi-modal and intermodal travel. As in existing transportation models, the TRANSIMS framework permits modeling mode choice in the activity list. The router does this by trying all possible modes and using a rational choice model to decide the best mode, or by feedback with a rational choice model to select travelers for mode reassignment. We will investigate combinations of these three to determine realistic mode choices quickly.

  • We are developing a generic detector technology to control actuated signals. At a minimum this detector will measure flows, densities, speeds, counts and their variability, and will allow noise to be included in any measurement. By changing the detector parameters, it will mimic any existing or future detector characteristics. This detector model is a precursor to the development of ITS technology methods.
  • We will add actuated signals as a feature to this study. If successful, the generic signals research discussed in the next section may encompass representations of such signal types.
  • We will develop some form of shared rides for each module. For the base-year study it may only be possible to consider intra-household shared rides. We will include inter-household shared rides in the HOV/ITS study.
  • We are developing methods to generate trips from generic time-of-day-dependent trip tables to produce background traffic that represents itinerant travelers and freight.
  • At a minimum we will complete the following output analysis of the base run. (1) We will estimate and display all forms of mobile emissions including those from both light- and heavy-duty vehicles and those caused by cold starts. (2) We will estimate and display the probability of accidents by transportation link and time of day. (3) We will collect, display and compare validation data with Portland counts or other model results.

Generic Representations

Like existing models, TRANSIMS in a forecast setting requires data that reflects the future land-use, population demographics, and transportation infrastructure. With this information, the synthetic population is aged to match the projected population, the land-use is used to reassign activity locations and times, and the future transportation infrastructure is converted to TRANSIMS transportation network data. A TRANSIMS analysis with this data constitutes a TRANSIMS “forecast.” TRANSIMS executions of such forecasts differ little from base-year applications or alternatives studies in which network detail may be reasonably well known. However, for a future-year network, the data would have to be estimated. Currently, a comprehensive model of a metropolitan region would include a network that represents all streets, along with the allowed movements between links on the network. It would have signals and signal timings that yield realistic traffic flows. It would have realistic transit schedules. The input preparation of such specific data for an approximated future network would be time and data intensive, while the output sensitivities to the input data uncertainties are unknown.

As stated previously, a second goal of the Portland TRANSIMS studies is to develop methods and techniques that diminish the TRANSIMS input data requirements, thereby making both forecasts and base-year studies with the technology much easier. In this series of studies, we will develop methods and techniques for generic representations in the TRANSIMS network, rather than the current data intensive representations that try to replicate the operational features exactly. Generic representations will be investigated for local streets, signalized intersections and transit schedules.

The work to create a network is fairly extensive, particularly if each local street is represented. In addition, this detailed representation increases the processing time for each TRANSIMS module. Los Alamos will lead a research effort with assistance from Portland Metro, to determine the minimum representation necessary to adequately simulate urban travel. As this local structure also accounts for land use and activity data, the proper representation is important. The effect of local streets, both their layout and their relationship to the surrounding major roadways, on the TRANSIMS output (including the emissions estimates) needs to be understood. This is important for forecast years where alternative street layouts may be placed on green fields, as well as for the simulation of a base year where simple local street representations could reduce data collection and execution times. This research may allow for generic local street representations across green fields. In many cases local streets may be represented simplistically, parametrically, or not at all.

Collecting traffic signal information for a network is time consuming. Methodologies must be developed for missing signal plans for the base year, and for both the placement and plans for the forecast years. Solving this problem, will require either a generic methodology for the creation of the input network, or a generic representation of traffic behavior at major intersections for the microsimulation. Portland Metro is taking the lead for characterizing the signal locations and timings. They will consider a rule-based approach to determine where to place signals. Possibilities include all arterial-arterial, arterial-collector and collector-collector intersections. Some local-collector and local-arterial intersections may also be considered.

Los Alamos will concentrate on the signal characterization in TRANSIMS and the effects of signal timings and placements on TRANSIMS output, iparticularly the emissions estimates. We are currently researching parameterized detectors and developing a probabilistic signal timing scheme that will allow for generic signal types. With this parameterization, traffic behavior (particularly that leading to emissions estimates) in the intersections can be calibrated. These parameterized intersections may mimic all signal types: pre-timed, actuated, adaptive, etc. Transit itineraries are needed for the forecast year. The current itineraries for the base study are from the Portland Tri-Met scheduling process (run-cutting and scheduling). This approach is probably not feasible for a future hypothetical system containing transit stops in what are currently green fields. Working together, Portland Metro and Los Alamos will develop a simpler approach.

TRANSIMS Commercialization

On June 28-30, 1999, Los Alamos hosted the TRANSIMS Opportunity Forum in Santa Fe, New Mexico. Approximately 50 attendees not directly associated with the TRANSIMS project learned about the TRANSIMS technologies and our plans to commercialize them. Neil Pedersen, Michael Morris, Keith Lawton, Tom Kane, Keith Killough, Annette Liebe, Robert Jones, John Simon, and Chris Barrett presented transportation planning market information which reflect an excellent understanding of the TRANSIMS technology, supported the need for TRANSIMS, and helped potential bidders for the TRANSIMS commercialization process. During sessions set aside for interactions with the TRANSIMS team, team members answered questions about prepared posters illustrateing aspects of TRANSIMS and its modules. Compact disks containing a test and evaluation version of TRANSIMS-LANL and its documentation were provided to those attendees who signed the appropriate license agreement.

The TRANSIMS request for proposals (RFP) was posted August 13, 1999 on the Los Alamos National Laboratory procurement website. The RFP seeks offers to:

  • Transform the TRANSIMS technologies into commercially viable software products and/or other Information Technology (IT) services that are available and useful to the transportation planning market. The resultant products and/or IT services shall be designated “TRANSIMS-DOT.”
  • Deliver the TRANSIMS-DOT products and services to as many as six early deployment transportation planning organizations in support of their TRANSIMS transition process and plans.
  • Be prepared to conduct all or selected elements of the TRANSIMS-DOT transition process for those early deployment transportation planning organizations requesting such assistance.
  • Establish a commercially viable “TRANSIMS” business infrastructure to facilitate the long-term application of the TRANSIMS technologies by the transportation planning market. This support is considered critical to the ultimate success of the TRANSIMS Commercialization Process.

On September 9, 1999, Los Alamos National Laboratory held a Pre-Proposal Meeting as a convenience for prospective bidders to ask questions and for the Laboratory to clarify its needs.

The proposal due date is November 19, 1999. Additional information may be found on the TRANSIMS website.

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: 505-665-5249
E-mail: llsmith@lanl.gov

Table of Contents

MPO Solicitation Plan

In September FHWA distributed their plan for soliciting interest from transportation planning agencies to participate in the TRANSIMS Early Deployment Program (EDP). That plan may be obtained from state transportation departments or state FHWA offices and is on the TMIP WorldWide Web site. The FHWA asked such agencies to indicate whether they expect to respond to the formal solicitation, currently anticipated to be issued inJanuary. The plan describes requirements for the solicitation to help agencies decide whether to respond.

The Solicitation Plan describes TRANSIMS and the EDP. The EDP was established to support the implementation of TRANSIMS. As part of the EDP, several agencies responsible for regional transportation planning will be selected to implement TRANSIMS. That implementation will serve to test and demonstrate TRANSIMS. FHWA will provide funding for part of the cost to implement TRANSIMS, to be matched by the selected agencies.

TRANSIMS is an entirely new approach to travel forecasting, and converting to TRANSIMS will require considerably greater commitment than simply switching software. Selected agencies will have to provide more data, have greater computing capability and provide more training than previously experienced with conventional travel forecasting programs. Agencies participating in the EDP will be selected in two phases. Phase One will begin in December, when interested agencies will have to submit specific information as well as a formal expression of interest and financial support. The FHWA will evaluate the technical and fiscal capability of responding agencies and select several of those for further, more rigorous consideration. The selected sites will be those judged best qualified for and most likely to be successful implementing TRANSIMS. Analytic capability, availability of needed data, and fiscal resources to match federal funds will be key considerations in the selection process.

It is expected that the final selections will be announced in mid-2000. For more information on the EDP and the agency selection process, contact Bob Radics at the FHWA Southern Resource Center: 404-562-3692 or robert.radics@fhwa.dot.gov. For information about TRANSIMS or TMIP, contact Kim Fisher: 202-366-4054 or kim.fisher@fhwa.dot.gov

All MPOs and State DOTs attending the TRB Annual Meeting are invited to a question and answer session on the TRANSIMS Early Deployment Program solication process:
1:30 on Thursday, January 12, 2000
Caucus Room – Washington Hilton

Table of Contents

TRB Annual Meeting

There will be a session devoted to TRANSIMS at 2:30-5:30 on Wednesday, January 12, 2000. The session will be at the Washington Hilton Hotel; the room is the Lincoln West. Topics to be covered are:

  • The TRANSIMS Early Deployment Program
  • Status Report on TRANSIMS Development
  • Accident Probabilities with TRANSIMS
  • TRANSIMS Feedback Research
  • Signal Research with TRANSIMS
  • TRANSIMS Treatment of Emissions
  • National Network Analysis Capability

Presenters are from Los Alamos National Laboratory, Frederick Ducca, Federal Highway Administration, and Kimberly Fisher, Texas Transportation Institute.

Table of Contents

New Documents

Two new documents are available for persons desiring more information about TRANSIMS and the Early Deployment Program.

TRANSIMS, Volume O – Overview is a high level, conceptual overview of the TRANSIMS software package. This document was prepared by Los Alamos National Laboratory and is intended as a primer for potential TRANSIMS users. It explains TRANSIMS capabilities and describes the major data that must be prepared, the major software methods that use these data, and the major output data produced by the software. It contains descriptions of each TRANSIMS module and explanations of several new analytical techniques possible with TRANSIMS. This document does not provide details of individual data items or data structures used by TRANSIMS, nor does it provide a detailed explanation of software algorithms. Persons desiring such detailed information can obtain full text documents via the document page at the TRANSIMS web site, http://transims.tsasa.lanl.gov/. The TRANSIMS Overview is available on the TRANSIMS Web page, in the document section under the June 1999 heading.

The second document, Early Deployment of TRANSIMS – Issue Paper, was prepared by the TMIP program as an introduction to the Early Deployment Program (EDP). The EDP is designed to move TRANSIMS from research to application. This is the first in a series of papers that will describe the EDP and TRANSIMS. This document describes TRANSIMS capabilities and status, and how transportation planning agencies can be involved in the EDP. The information provided on the EDP schedule, administrative details, and funding descriptions may be particularly helpful. This paper provides a fairly general description of the modeling details and technical specifications of TRANSIMS to introduce it to technical, management, and transportation policy professionals. Subsequent documents will provide more detail; those documents will be announced on the TMIP and TRANSIMS web pages when they are available. Hard copies of both documents are available from Kim Fisher, 202-366-4054 or fax 202-366-7660. Full text versions of both documents are also available on the Internet. The Issue Paper is available on the TMIP web page, http://tmip.tamu.edu, in the What’s New Section.

Table of Contents

TMIP Freight Track

TMIP Track F freight planning coordinators are reviewing a draft of the FHWA “Freight Data Handbook (FDH) for States and MPOs”. After revisions resulting from that review, the handbook will be published early in 2000. Also, the 1997 Commodity Flow Survey (CFS) will be released for metropolitan areas and rest-of-state regions in December,1999. The CFS will provide data for the top 50 cities, the remainders of states having one or more of the top 50 cities, and the entirety of states not having one of the top 50 cities, all of which account for 107 CFS regions.

The FDH will provide the location and description of databases, where to find them and how to use them. The FDH will also include a listing of names, phone numbers and e-mail addresses of freight planners and others knowledgeable about freight databases for MPO, regional and statewide freight modeling. More detail about the databases can be obtained by contacting appropriate persons in the listing. Persons having difficulty accessing, acquiring or using freight databases are encouraged to contact Russ Capelle at 202-366-5685. Information about any freight planning studies and commercial vehicle surveys done for MPOs, regions or states should be sent to Rus Capelle for inclusion in the bibliography in the FDH.

Table of Contents

New TMIP Seminars

TTI is sponsoring professional training seminars on three topics as part of the Travel Model Improvement Program (TMIP). The purpose of these seminars is to "deliver" the results of TMIP research to the target audiences, thereby building capability at those agencies to implement travel forecasting model improvements developed in TMIP sponsored research. Each of the seminars will be presented three times at different locations around the United States. The seminars are part of the TMIP training program in which previous training manuals were prepared, describing how to conduct travel forecasting. Material from two manuals and recent course graphics will be incorporated in these seminars:

  • Travel Model Calibration and Validation will describe, explain and demonstrate the process and procedures for estimating, calibrating and validating regional travel forecasting models with base year data. Procedures applicable to each of the standard 4-step travel models will be described. Procedures for performing checks for reasonableness of travel forecasts will also be described. The seminar will be lead by Tom Rossi.
  • Forecasting Land Use Activities will describe, explain and demonstrate methods and data sources for forecasting the nature and amount of land using activities, specifically population, dwelling units, basic employment and service employment. The procedures addressed will include estimating control totals for an urban region and distributing activities among subregional analysis areas. This seminar will include a rudimentary description of land use models and other allocation procedures. Procedures for assimilating needed data from various sources and checking the reliability of that data will also be described. The seminar will be led by Paul Waddell.
  • Vehicle Availability Forecasting will describe, explain and demonstrate key concepts, data sources and procedures for forecasting vehicle ownership and availability. These procedures will be consistent with and capable of producing data readily usable with the standard 4-step travel forecasting models. This seminar will be led by Jerry Faris.

Each seminar will be conducted completely in one day. The seminars will accommodate 30 participants, first come-first served. Procedure manuals developed in TMIP will be the basis of the seminars, and those manuals will be given to attendees. The seminar presentations will describe recent advances in the state of the practice for the seminar subject.

The seminars will be free of charge and conducted by consultants experienced in the respective seminar topics. The final schedule and locations of the seminars have not been determined; they will be announced on the TMIP Web site (tmip.tamu.edu) and the TMIP e-mail service. More information on the seminars can be obtained by contacting Lisa Day (817-462-0530).

Table of Contents

TMIP Training Courses

Introduction to Travel Forecasting

An innovative training opportunity sponsored by TMIP is nearing completion. TTI and a training consultant are developing a self-taught course on basic travel forecasting. The course will be available on CD/ROM and hard copy. Both the CD and the manual will be stand alone and will include exercises and provide background and practice examples to serve as references for practicing travel forecasting personnel. The course will be a major revision of a course offered for some time by the National Highway Institute (#15254).

The course will include new topics, as well as updated and new examples of traditional 4-step travel forecasting procedures. A real community will be used throughout the course as the basis for examples and workshop exercises. The students will be able to follow the forecasting process with realism, seeing how the procedures apply to the real community. The course includes a one-day compter workshop that reinforces theories and concepts learned in class.

The students will experience real issues of obtaining data needed for the forecasting process and interpreting output of models and the meaning of those results. The course will be designed to help both novices and personnel with limited experience. It will help planners desiring a better understanding of the principles and techniques of travel forecasting.

For information on scheduling of this course and availability of the CD/ROM and manual, contact Michael Culp at 202-366-9229. The CD/ROM is expected to be available in May 2000.

Advanced Travel Forecasting Course

A new, advanced travel forecasting course is now available from the National Highway Institute. The course was designed to help experienced travel modelers assimilate new procedures developed in TMIP sponsored research and elsewhere. The course provides practical training on many of the most recent advances in travel forecasting including land use modeling, tour based modeling, time-of-day forecasting and ITS analyses. The course also addresses TDM and TCM analyses, transportation/land use policies and transportation system performance indicators. Persons taking this course should have had at least 3 full years of hands-on experience using 4-step travel models. The first advanced course will be given in Milpitas, California, April 11-14, 2000. More information on this course and scheduling of future sessions can be obtained on the TMIP web site.

Schedule for the Introduction to Travel Forecasting Courses follows:

January 18-21, 2000
Columbus, OH
Mark Byrom
614/466-7825

February 22-25,2000
Salt Lake City, UT
Lorrie Lau
415/744-2628

March 20-24, 2000
Lansing, MI
Dave Bell
517/335-2961

April 24-28, 2000
Baltimore, MD
Brian Betlyon
410/962-0077 Ext. 3076

May 8-12, 2000
Austin, TX
Marylin Dell
512/483-3628

May 22-26, 2000
Atlanta, GA
Ben Williams
404/562-3671

Table of Contents

Conferences

Transport Systems - Organization and Planning
3rd KFB Research Conference
June 13-14, 2000, Stockholm Sweden

Call for Abstracts

Themes – Preliminary Presentation

  1. Transport and location modelling
  2. Logistics and supply chain management
  3. Infrastructure economics and development
  4. Transport policies and impact analysis for sustainable development
  5. Urban transportation systems
  6. Institutions and processes of transport and traffic planning
  7. IT and transport markets
  8. Deregulation and organization
  9. Traffic simulation
  10. Public transport: management and organization

Abstracts Deadline
November 30, 1999.
Send your abstracts to the Scientific Secretariat. Please use e-mail or disk. The Scientific Committee reserves the right to reject abstracts. Acceptance from Scientific Committee not later than January 15, 2000.

Final Papers
Deadline March 1, 2000.
Acceptance from Scientific Committee not later than April 1, 2000.
Time - JUNE 13-14, 2000

Conference Site
Stockholm School of Economics, Sveavägen 65, Stockholm, Sweden

SCIENTIFIC SECRETARIAT
Katja Berdica ktj@infra.kth.se
KTH, Royal Institute of Technology
Department of Infrastructure and Planning
Division Transport and Location Analysis
SE-100 44 Stockholm, Sweden
Phone: +46 8 790 9416,
Fax: +46 8 790 67

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