Automatic Scenario Generation for Advanced Traffic Management Systems, 10-9150

 Principal Investigator
Steven W. Dellenback, Ph.D., PMP

Inclusive Dates: 07/01/99 - 09/30/00

Background - Throughout the nation, congestion is increasing on highways and arterial roadways. The state Department of Transportations (DOTs) and local municipalities simply cannot increase the number of lane miles available to keep pace with the increasing number of vehicle miles being logged each year. Two major factors prevent the building of highways, funding and land availability. The cost of building roadways continues to escalate, and there are simply not enough funds to build the necessary roadways to alleviate congestion. Secondly, obtaining rights-of-way and overcoming the various environmental regulations cannot be accomplished in many metropolitan areas (e.g., Austin).

One solution to both these problems is Intelligent Transportation Systems (ITS). The primary goal of ITS is to engage technology to better use the existing capacity of the nation's highways. A particular class of ITS systems, termed Advanced Traffic Management Systems (ATMS) has been the focus of SwRI's Automation and Data Systems Division. The local San Antonio ATMS, the Texas Department of Transportation (TxDOT) TransGuide system, is widely recognized as the most advanced ATMS system in the nation. One functionality area lacking in ATMS systems is incident management (i.e., traffic congestion due to an accident, debris, or weather). When an incident occurs, an ATMS uses a scenario to determine how the ATMS field equipment should be modified (i.e., what messages to display, what lanes to close). This project addressed the development of an Automatic Scenario Generation concept that provides SwRI with a core technology that can be leveraged into new and extended project activities.

The motivation behind this research was the limitation of the two methods currently in use by traffic management organizations to create scenarios:

• Predetermined. Hundreds of hours are spent developing solution scenarios manually (based on a set of runs provided by the traffic management engineer and local knowledge of the highway infrastructure). After these scenarios are developed, the traffic management engineer must cross-check them. Then they must be inserted into the operational system. If a traffic incident occurs for which no solution scenario exists, one must be created dynamically. The advantage to this approach is that existing scenarios can be implemented within seconds of an incident occurring.

• Ad hoc. Solution scenarios are not created a priori, rather, when an incident occurs, the response is created dynamically (again this response is based on a set of operational rules). This approach is significantly slower than the first because it takes time to "piece together" the equipment that will be in the solution.

Approach - This project developed a set of algorithms that created solution scenario rules (termed a Constraint Language) and coupled these rules with the roadway geometry [available from Geographical Information System (GIS) data files] to create solution scenarios in a dynamic fashion. Roadway geometry is a significant factor in the creation of a scenario, and the concept is novel in its approach to integrate scenario generation rules and a GIS database. The scenarios that are created were validated against currently existing traffic management guidelines.

Accomplishments - The success criteria established in the proposal was that the scenarios generated must mimic the static scenarios (those scenarios created by hand) 95 percent of the time. The algorithms developed during this program produced scenarios that matched the static scenarios.

Intelligent Systems, Advanced Computer and
Electronic Technology, and Automation Program
2000 IR&D Home SwRI Home