This project deployed acoustic and Doppler radar sensors on railroad tracts to detect the presence, speed, and length of trains as they approached highway grade crossings. Sensor data was transmitted to the TransGuide Control Center where computer programs calculated and predicted the time and duration grade crossings at or near freeway exits would be blocked. This information was posted on freeway dynamic message signs to warn exiting drivers of possible delays.

The system was found to work well; however, at the time of the study low levels of congestion in the network did not warrant the implementation of an advanced warning system to avoid long delays. Simulation techniques were used to show the potential benefits of the system and to identify a decision-point at which congestion levels would warrant such a system in the future.

RESULTS

With a 25 percent increase in congestion and a 15 percent driver compliance rate, the simulation model indicated the warning system would yield positive benefits in terms of reducing total network delay. With a 20 percent driver compliance rate the total network delay would decrease by 6.7 percent and the crash rate would decrease by 8.7 percent.