Utilizing dynamic hazard knowledge for risk sensitive action planning of autonomous robots

Autonomous robots are required to perform tasks in complex and dynamic environments. For this class of systems, traditional safety assuring methods are not satisfying due to the unknown effects of the interacting system with an open environment. Briefly speaking: What is not known during the development phase can not be adequately considered. In order to tackle this problem, it is proposed to extend the safety measures with the so-called dynamic risk assessment. Therefore, the anticipatory capability of a Cognitive Technical System, the so-called mental action space, is utilized. The mental action space, a learned internal representation for possible courses of action, is dynamically assessed. The proposed dynamic risk assessment module provides this functionality. The core are quantitative risk models, so-called `safety principles', which can be specified during the system's design stage without losing the possibility to be adjusted or extended during the system's operating time. Finally, an exemplary application of the approach shows a real robot, enabled to safely plan and perform its tasks concerning risks arising due to interaction of robot and environment.