Gaze detection based driver modelization in an electric vehicle using virtual force field and Steer by Wire system

Even if the reliability of automobiles was greatly improved, number of accidents and casualties on the road remain quite high. Those incidents are most of the time due to the behaviours of drivers, which could be avoided using driving assistance. Therefore this paper proposes a new driving assistant structure, based on a virtual potential fields approach. By that mean, driver´s state and intention are estimated through an analysis of facial features, and collision free trajectory detections become possible. This driving assistant is implemented in an electric vehicle using a Steer by Wire system and the validity of the proposal is verified through simulation.