Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration

Office of Planning, Environment, & Realty (HEP)
Planning · Environment · Real Estate

HEP Events Guidance Publications Glossary Awards Contacts

Southeast Michigan Council of Governments (SEMCOG) Travel Model Peer Review Report (2011)

4.0 Southeast Michigan Council of Governments TDFM Assessment and Discussion

Following presentations related to the development of the current SEMCOG TDFM, Version E5, and the in progress updates that will lead to a new version of the SEMCOG TDFM, Version E6, in March 2012, the peer review moved on to presentations about the analytical needs of SEMCOG and model enhancement recommendations. This section of the report covers those presentations given by Rick Donnelly of Parsons Brinckerhoff.

4.1 Analytical Needs

The analytical needs of SEMCOG with respect to travel modeling were categorized into three groups: 1) national issues and trends, 2) Federal requirements (e.g. model certification by FHWA, and applications for FTA New Starts funding), and 3) Local requirements (comprising strategic uses, tactical uses, and performance measures).

4.1.1 National Issues and Trends

The context for the discussion of analytical needs was set by referring TRB Special Report 288, Metropolitan Travel Forecasting: Current Practice and Future Direction (2007)[7] and specifically this quote:

"The committee therefore recommends development and implementation of new modeling approaches to demand forecasting that are better suited to providing reliable information for such applications as multimodal investment analyses, operational analyses, environmental assessments, evaluations of a wide range of policy alternatives, toll-facility revenue forecasts, and freight forecasts, and to meeting federal and state regulatory requirements."

4.1.2 Federal Requirements

The presentation of analytical needs noted several important aspects of the Federal requirements, in terms of both the FHWA certification process and the use of the travel model to support FTA New Starts funding applications. FHWA has established a certification process and checklist for MPO models to ensure that they are capable of informing federally mandated air quality and transportation planning requirements. Other than certain requirements that affect the use of model outputs for air quality modeling, the certification checklist is not definitive about the structure or capabilities of travel demand models. FHWA acknowledges that analytical requirements vary from one MPO to another. However, the process is designed to ensure that the models that are in place are adequate for current and anticipated applications of the model. Moreover, they outline a series of questions to assess how well the modeling program addresses analytical risks, the agency's technical capabilities, and documentation. The documentation considered during certification includes, but is not necessarily limited to, three major areas:

While FTA does not approve models on a set schedule similar to FHWA's certification process, they do review models as part of their forecast review during consideration of New Starts applications. In general, FTA prefers to see a model with the following features:

4.1.3 Local Requirements

Local requirements were presented in two main categories: strategic versus tactical uses of the model. Strategic uses include long-range transportation plans, transportation improvement plans, and analyses of region-wide impacts and opportunities. Tactical uses focus on specific projects or programs, whose effects are usually localized or concentrated in certain corridors or smaller study areas.

An additional local requirement that was discussed is the use of the TDFM to report performance measures. The transportation-related measures that the TDFM will likely be called upon to inform about include:

The presentation of local requirements assigned the strategic and tactical requirements (including the performance measures) to travel model capabilities that are traditional and those are non-traditional as shown in Figure 2.

Figure 2: Assessment of local requirements

 

Tactical

Strategic

Traditional

Subarea analysis

Selected link analysis

Transit corridor or project analyses

Traffic impact studies

Time and cost savings of projects

Performance Measures

Update RTP and TIP

Project prioritization and programming

Air quality conformity

Assessment of regional strategies

Trade-off analysis

Improving regional mobility

Demand management vs. capacity

TOD and 4D analyses

Evaluating new transport modes

Non-Traditional

Construction and detour analyses

Work zone staging and management

Linkage with traffic operations models

Signal timing optimization

Traffic impact studies

Network reliability

Equity and environmental justice impacts

Capacity reduction strategies

Market accessibility and competitiveness

Economic impacts of transport

Understanding freight and supply chains

Impact of fuel price increases

Equity and environmental justice impacts

While the traditional travel model capabilities could be supported by a best practice trip-based model and would suggest that a continuation of SEMCOG's E Series models was sufficient, Parsons Brinckerhoff presentations identified that a dynamic traffic assignment (DTA) model would be required to support the non-traditional tactical requirements, while an integrated land use-transportation model would be required to support the non-traditional strategic requirements. Together, these two model requirements suggest the need for an activity based (AB) travel demand model.

4.2 SEMCOG Model Enhancement Recommendations

Following on from the discussion of analytical needs, Rick Donnelly of Parsons Brinckerhoff presented a summary of model enhancement recommendations that would achieve those analytical needs. The recommendations were introduced with reference back to the three categories of analytical needs. While the current model meets FHWA certification requirements and could support FTA New Starts applications and many of the local requirements, advanced models are becoming the state of the practice and are needed for many of the local requirements. The concept of multi-scale travel modeling was introduced to describe the roles that a new SEMCOG advanced model would cover and its linkages to both larger and smaller scale models (Figure 3).

Figure 3: Parsons Brinckerhoff's vision for multi-scale travel modeling

The figure summarizes an approach developed by Parsons Brinkerhoff regarding multi-scale travel modeling. The image identifies the MDOT statewide model as a megascopic model, categorizes the SEMCOG Ax model as both a macroscopic and mesoscopic model, and classifies detailed ad hoc traffic analysis models as microscopic models.

The presentation described an incremental strategy for gradually moving from current trip-based models to advanced models. The incremental strategy would make use of shared components, such as travel behavior data (e.g. surveys), networks, and target and validation data. The gradual transition would allow for staff development, and maintain fully operational models at each stage. The following sections outline transition strategies for the demand side, supply side, and data programs. The model enhancement recommendations were summarized as:

4.2.1 Transition Strategy: Demand Side

Transitioning to an AB model for the demand side of the TDFM was described as providing several advantages, including deeper insights into travel behavior, more realistic representation of travel dynamics, making linkages to a DTA more straightforward, eliminating non-home based trips, and improving the ability to support equity and pricing studies. The following stages were laid out:

  1. Best Practice Model Implementation (Version E6/E7)
    1. Complete market segmentation by income
    2. Destination choice model
    3. Calibrated mode choice model
  2. Enhanced Trip-Based Model (Version A1)
    1. Linked trips (half tours) to reduce non-home based trips
    2. Tour analysis of travel surveys
    3. Trip frequency choice model
    4. Finer temporal allocation factors
    5. Sub-county validation targets
  3. Population synthesizer and daily travel activity patterns (Version A2)
    1. Adapt UrbanSim population synthesizer
    2. Integrate trip generation into daily activity pattern models
    3. Application of daily activity patterns to synthetic population
  4. Tour-based mode and destination choice models (Version A3)
    1. Primary tour destination and mode choice
    2. Stop frequency and location choice
    3. Trip mode choice
    4. Implementation in microsimulation framework
  5. Fully integrated model (Version A4)
    1. Time-of-day choice (activity scheduling)
    2. Time-space constraints
    3. Inter-household interactions and constraints
    4. Full integration with dynamic network models.

4.2.2 Transition Strategy: Supply Side

Transitioning to a DTA model for the supply side of the TDFM was described as providing several advantages, including enabling robust tactical solutions, and more realistic representation of travel dynamics. The following stages were laid out:

  1. Data development
    1. Expanded traffic counts (hourly by vehicle type)
    2. Probe data
    3. Advanced traffic management system data feeds
    4. Performance reporting
  2. "Planning level" DTA
    1. Revision of link capacity functions
    2. Quantifying network reliability
    3. Use in parallel with static user equilibrium model(s)
  3. Simulation-based DTA
    1. Intersection coding templates
    2. Expansion of network coding to include intersections
    3. Signal timing heuristics
    4. Network summarization and reporting tools
    5. Micro and macro-level validations.

4.2.3 Transition Strategy: Data Programs

The following steps were laid out to transition to a data program that would support the development of advanced models:

  1. Second generation MI TravelCounts program
    1. Move to a continuous data collection program
    2. Tour-building heuristics
    3. Tweak to better understand intra-household interactions
  2. External travel survey
    1. Retrospective long-distance survey as TravelCounts add-on
    2. Mode-specific visitor surveys
    3. Focus on Freight Analysis Framework Version 3 synergies
  3. Commercial travel survey(s)
    1. Collaborate with CMAP (follow protocol)
    2. Focus of Freight Analysis Framework Version 3 synergies
    3. Major freight facilities database
    4. Commercial vehicle tracking programs
  4. Network data program
    1. These are data development steps that would assist with the collection and maintenance of the more detailed highway and transit network data required to support more advanced representation of transportation supply in, for example, DTA models:
      1. Evolution of Network Master
      2. Build off of OpenStreetMap and Google
      3. Network design problem - hourly classification counts
      4. Intersection coding templates
      5. Signal timing heuristics
      6. Network summarization and reporting tools
      7. Micro and macro-level validations
  5. Passive tracking
    1. Passive data collection of traveler data using, for example, data collected from GPS enabled smart phones, is starting to become a viable approach
      1. Evolution from self-reporting of travel to self-describing passively reported travel
      2. Pattern recognition and artificial intelligence constructs to mine and analyze large quantities of passively collecting data.

[7] TRB Special Report 288 is available at http://onlinepubs.trb.org/onlinepubs/sr/sr288.pdf

Updated: 04/28/2014
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000