Infrastructure-based vehicle maneuver estimation at urban intersections

Accident-prone spots in urban areas are mainly located at signalized intersections, where road users are often confronted with complex situations that are sometimes hard to interpret and deal with properly. The objective of this paper is to present a model for estimating the maneuvers of the vehicles approaching an urban intersection. This estimation is part of the Intelligent Cooperative Intersection Safety system (IRIS), which assists the road users at urban intersections. The system uses vehicle-to-infrastructure communication and extended road side sensing to track and predict the movements of individual road users. The estimation of the future situation at the intersection is an input for the identification of safety-critical situations. Once the system detects a critical situation, it sends a warning message to the road users via short-range communication. The system was successfully tested at a real intersection in the city of Dortmund, Germany.