Teaching Commonsense and Dynamic Knowledge to Service Robots.

Incorporating commonsense and coping with dynamic knowledge are key capabilities of service robots to efficiently interact with humans. In the presented system, we demonstrate how to equip service robots with commonsense knowledge and the dynamic reasoning capabilities of Answer Set Programming (ASP). We investigated the response of our system to basic human needs and evaluated the viability and scalability of the combination of the commonsense knowledge database ConceptNet 5 and the ASP solver Clingo. Our results show the flexibility and versatility of our approach. Further, we identified the need for research on scalability in case of environments that are abundant with objects.