A flexible method for criticality assessment in driver assistance systems

This paper presents an efficient algorithm to compute time-to-x (TTX) criticality measures (e.g. time-to-collision, time-to-brake, time-to-steer). Such measures can be used to trigger warnings and emergency maneuvers in driver assistance systems. Our numerical scheme finds a discrete time approximation of TTX values in real time using a modified binary search algorithm. It computes TTX values with high accuracy by incorporating realistic vehicle dynamics and using realistic emergency maneuver models. It is capable of handling complex object behavior models (e.g. motion prediction based on DGPS maps). Unlike most other methods presented in the literature, our approach enables decisions in scenarios with multiple static and dynamic objects in the scene. The flexibility of our method is demonstrated on two exemplary applications: intersection assistance for left-turn-across-path scenarios and pedestrian protection by automatic steering.