Towards Intuitive Human-Robot Cooperation

Abstract

Human-robot cooperation calls for the treatment of human-machine communication channels, especially if humanoid robots are involved. In this paper, we consider implicit nonverbal channels given by recognizing the partner's intention and proactive execution of tasks. We propose a method that keeps the human in the loop and allows for the systematic reduction of uncertainty inherent in implicit cooperation. We present a benchmark scenario as well as preliminary implementation results.