National Center for Supercomputing Applications TRANSIMS Visualization

Background

TRANSIMS, a transportation simulation code, was developed by the U.S. Department of Transportation (USDOT) to move traffic modeling away from traditional approaches towards computing-intensive agent based simulations. The promise of this approach lies in the fact that such a code can track individual travelers and that travelers can be derived from a synthetic population. Based on surveys and zoning information, the travelers' activities can also be used to derive the need for individual trips. Being able to follow specific individuals during their individual trips allows the implementation of details that were previously unavailable to modelers. TRANSIMS covers car-based trips, as well as all other modes of transportation (bus, rail, biking, walking, and more).

Techniques were also developed to interpret the data in a visual form that integrates the results of detailed computations to make interpretation of critical results far easier for the analysts and decision makers.

The original TRANSIMS contained a visualizer component that had limited capacity for the size of the road network, as well as the time period that it could cover. When this visualizer was discontinued, USDOT began using the Balfour fourDscape application. This renders the simulation code's information in a rich environment, which includes satellite and aerial photography, buildings based on polygons, and even such details as trees.

The results from TRANSIMS simulations are extremely detailed. The sheer amount of data to be analyzed requires advanced visualization techniques to combine the computing power of modern visualization equipment with the human brain's unique ability to spot unusual patterns in visual data with relatively little effort. Snapshot data are available that covers the exact location of all travelers in a large metropolitan area on a second-by-second basis. Summary data can be extracted that shows the traffic density, congestion, and other traffic-related parameters for each individual road segment modeled in the code. On the most detailed level, the exact phase of every traffic signal is available throughout a full day of simulation.

Each of the visualization tools currently employed for TRANSIMS case studies has certain advantages and limitations. Most of the tools are commercial tools, such as the fourDscape software, which are not readily available to researchers because of their high licensing cost. As a rapidly growing open-source project, TRANSIMS is clearly in need of a more specialized open-source visualization platform that meets its users' needs and can evolve with projects as they develop.

TRACC Research Activities

TRACC is collaborating with the National Center for Supercomputing Applications (NCSA) at the University of Illinois, in the development of new techniques and approaches to portray the results of traffic simulations in metropolitan areas by means of TRANSIMS. NCSA is an international leader in developing and deploying artful visualizations of scientific phenomena.

The project's key objectives include development of visualization methodologies that can:

• Assist transportation modelers in the creation of large-scale metropolitan models,
• Efficiently aid the human visual system to detect errors in the model,
• Aid analysts in validating the models against observable information, and
• Provide easily interpretable visual results to transportation system models and other critical tools.