Doctoral colloquium

Abstract

In humanhuman interaction, social signals or unconscious cues are sent and received by interaction partners. These signals and cues influence the interaction partner wanted and unwantedsometimes to achieve a distinct goal. To be aware of those signals and especially of implicit cues is crucial when interaction between a robot and a human is modelled. This project aims to use implicit body and machine movements to make HRI smoother and simpler. A robot should not only consider social rules with respect to proxemics in communication or in encounter people. It should also be able to signal and understand certain spatial constraints. A first spatial and situational constraint, which this research project currently focuses on, is avoiding each other. This includes not only passing by but also especially making room for each other. Consider a narrow place, e.g. hallways, door frames or a small kitchen. A robot might block the way or drives towards you, pursuing its own goal like you. Humans do not even speak to each other in order to pass by and avoid bumping into each other even if the space is narrow. A first study is currently conducted to find out which behaviour is the most appropriate avoiding strategy and how participants express their wish to pass by. Therefore, a variety of defensive and more offensive avoiding strategies of the robot are applied in experiments. The results of the study will be used to equip the robot with spatial concepts to make interaction faster and more appropriate.