Research Agency: Cambridge Systematics, Inc.
Principal Investigator: Tom Rossi
We all know as travelers that we make our travel choices in response to many factors including destination, price, travel time, travel time reliability, convenience, status, parking availability, and information about current conditions. We also know that congestion can drive us to change travel behavior because it pushes us over some personal threshold. At that point we shift our route, shift departure time, join a carpool, take transit, work at home if allowed, or maybe move.
Public policies such as adding highway capacity, improving traffic operations, adding transit capacity, introducing priced roads, providing better traveler information or offering companies with tax benefits for transit subsidies further influence the choices that we make. Public agencies are continuously evaluating difficult policy options like these but the transportation modeling tools are not adequate for the job.
The essence of the problem addressed by Project C10B is that traveler behavior responds to network conditions and network conditions respond to behavior. The present generation of models is not sensitive to this dynamic interplay and, therefore, cannot properly analyze transportation alternatives. In other words, the planning representation of demand is not informed by operating conditions on the network at the time the travel occurs. In turn the representation of network operations is not informed by changes in demand. Project C10A integrated the supply and demand sides of transportation demand forecasting in order to test operational improvements to the highway system as well as capacity enhancements.
The primary objective of Project C10B was to make operational a dynamic integrated model–an integrated, advanced travel-demand model with a fine-grained, time-dependent network-and to demonstrate the model’s performance through validation tests and policy analyses. This integrated model system was necessary because most current travel models are not sufficiently sensitive to the dynamic interplay between travel behavior and network conditions, and are unable to reasonably represent the effects of transportation policies such as variable road pricing and travel demand management strategies.
The secondary objectives of this project were to: (1) produce a portable and transferrable product, process, and sample data set that can be adapted for use elsewhere or used for research, (2) incorporate SHRP 2 Capacity products from Project C04 (Improving Our Understanding of How Highway Congestion and Pricing Affect Travel Demand) and Project C05 (Understanding the Contribution of Operations, Technology, and Design to Meeting Highway Capacity Needs) into the model capabilities, (3) incorporate travel time reliability into the modeling capabilities, (4) demonstrate the application of outputs of the integrated model to estimate greenhouse gas emissions using EPA’s MOVES Model, and (5) demonstrate the dynamic integrated model set in a real-world environment on selected policies.
Project C10B developed an integrated, advanced travel demand model with a fine-grained time-sensitive network simulation for the Sacramento, California, region. Project C10B developed a methodology to reflect changes in the nature of demand, mode choice (including “new modes” such as work or shopping at home and nonmotorized travel), destination choice, timing, and route of travel as a response to transit service, highway network congestion, roadway management strategies, road pricing, parking policies, and other public policies aimed at reducing congestion. This was achieved by integrating the activity-based demand model DaySim and a Dynamic Traffic Assignment (DTA) model, DynusT and a transit network simulation model, FAST-TrIPs. All are open-source products. Integration meant that a feedback loop was built between the demand and network assignment model systems. All of the demographic, highway network, and transit service data required to run the model set were assembled, and the feedback between the demand model and the DTA was tested in a subarea of Sacramento and on the full urban network. The model sets and software Start-up Guides are available on the archive.
A companion report and model set are available for the application in Jacksonville, Florida through Project C10A. This work has the same objective and uses DaySim as the demand model but uses TRANSIMS for the highway network assignment. Project C10B was intended to serve communities with more mode choices than those in Project C10A by taking into account mode choice response to highway conditions.
In the future, the products of this research may be updated through implementation work of FHWA, AASHTO, and others. The archive is intended as a digital record of the products and supporting materials of all reliability related research and development that occurred during SHRP 2 which Congress mandated to end on March 31, 2015.
Note: This project contains artifacts, such as .zip or .xlsx files, under ‘Non-Datasets’ but may contain data of interest embedded in an analysis tool/spreadsheet. For researchers interested in this additional data, it is suggested to download these artifacts individually from the ‘Non-Datasets’ section.