Citation

Qian Y, Wang Y, Rizos DC, Farkas C, Huang CT. U.S. DOT Federal Railroad Administration. Washington, D.C.: U.S. DOT Federal Railroad Administration, 2022.

Copyright

(Copyright 2022, U.S. DOT Federal Railroad Administration)

The full document is available online.

Abstract

Railroad crossings may cause unexpected blockages and considerable traffic delays, especially in urban areas. However, there is no practical system available to share potential crossing blockage information to the public. An unexpected blockage threatens first responders while they respond to emergencies. This research project has developed an Intelligent Crossing Assessment and Traffic Sharing System (i-CATSS) that can detect and predict highway-rail blockages at grade crossings and provide first responders with real-time information of traffic conditions at crossings.

From August 2019 through February 2021, researchers at the University of South Carolina developed an innovative Intelligent Crossing Assessment and Traffic Sharing System (i-CATSS) that can detect and predict highway-rail blockages at grade crossings and provide first responders with real-time information of traffic conditions at crossings. The system evaluates the total expected delay time due to both passing trains and vehicle congestion in front of the railroad crossing.

The research team developed a graphic user interface to display the estimated arrival time of the train and the estimated departure time for the monitored crossing, given the information shared from the partner railroad. Traffic status is consistently updated when new incoming information is received. The team also developed an artificial intelligence (AI) model to detect the number of vehicles waiting in front of the blocked crossing. The system automatically starts whenever a train is detected within the area of interest by the surveillance camera. The correlations between the number of the waiting vehicles and the delay time are established based on the AI model. The model training and validation are performed using the surveillance videos recorded at the crossing of interest.

First responders from Columbia, SC, offered opinions on the impact of the unexpected railroad blockages through a survey. They also assisted in system development. The South Carolina Department of Transportation for the City of Columbia assisted in location identification and video collection. Industry partner CSX provided the train operation information for system development and improvement.