A Curve Warning System to Enhance Transportation Safety    

ABSTRACT 

Previous studies have shown that crashes on curve sections of rural roadways have a higher severity than straight sections.  Recent research at Clemson University using 2019 crash data indicated that crashes are more than four times more likely to be fatal on curve sections compared to straight sections.  The research also found that one of the primary contributing factors to fatal crashes on curve sections is excessive speed.  This poster presents a rural road warning system application designed to alert drivers if they are driving too fast to safely maneuver a curve.  The application combines curve and vehicle characteristics to determine the maximum safe speed of vehicles about to enter a curve.  By comparing the actual vehicle speed to the maximum safe speed, an appropriate warning message can be identified and sent to the vehicle via infrastructure-to-vehicle communication.   The poster presents underlying principles and application components, including the sensing approach (computer vision-based machine learning), sensor specifications and placement, edge computing, communication needs, and power requirements.