TUM and BMW develop AI system for real-time traffic hazard response
Researchers at the Technical University of Munich and BMW Group have created an AI system that combines sensor data from vehicles, traffic lights, and overhead sign gantries to generate safety recommendations in less than 100 milliseconds. The system creates a digital representation of traffic with up to ten centimeters positional accuracy and continuously predicts the movement of cars, bicycles, and pedestrians. According to Professor Alois Knoll, the system allows drivers to see obscured areas and react more quickly to hazards, such as when children run onto the road or cyclists come around a bend. The researchers have filed patents for construction site remapping and a new approach to data fusion in distributed systems.
