Knowledge-based risk assessment for intelligent vehicles

In order to set up assistance systems in intelligent vehicles or to control an autonomous vehicle a number of cognitive functions has to be realized in an integrated architecture. In this paper we propose a knowledge-based risk assessment procedure in order to identify objects which might be dangerous for the own vehicle. Having an advanced vision system with gaze control in mind a knowledge-based risk assessment can determine where to concentrate the attention. The approach is evaluated by simulating different traffic scenes.