Theory of Mind Based Assistive Communication in Complex Human Robot\n Cooperation

Abstract

When cooperating with a human, a robot should not only care about its\nenvironment and task but also develop an understanding of the partner's\nreasoning. To support its human partner in complex tasks, the robot can share\ninformation that it knows. However simply communicating everything will annoy\nand distract humans since they might already be aware of and not all\ninformation is relevant in the current situation. The questions when and what\ntype of information the human needs, are addressed through the concept of\nTheory of Mind based Communication which selects information sharing actions\nbased on evaluation of relevance and an estimation of human beliefs. We\nintegrate this into a communication assistant to support humans in a\ncooperative setting and evaluate performance benefits.\n We designed a human robot Sushi making task that is challenging for the human\nand generates different situations where humans are unaware and communication\ncould be beneficial. We evaluate the influence of the human centric\ncommunication concept on performance with a user study. Compared to the\ncondition without information exchange, assisted participants can recover from\nunawareness much earlier. The approach respects the costs of communication and\nbalances interruptions better than other approaches. By providing information\nadapted to specific situations, the robot does not instruct but enable the\nhuman to make good decision.\n