Socio-Contextual Constraints for Human Approach with a Mobile Robot

The efficiency of human-robot collaborative task completion benefits from seamless interactions. Robots that act in a shared environment with humans, can improve the initiation of such interactions by externalizing the intention of accompanying motions. Especially within robot locomotion the ability to convey intentions non-verbally is beneficial for interaction initiation and thus for mutual collision avoidance and the reduction of interferences. This work examines specific trajectory features that provide the required natural movements and support intention conveyance. The named correlations are investigated in the social context of robot locomotion for a pro-active robot-to-human approach. Trajectories that aim to clearly convey this intention of the robot are planned based on optimal control principles. Within a subject study different features like path shape and orientation for robot-to-human approach trajectories are evaluated in their performance of supporting intention conveyance. Results show that the perceived naturalness and the comprehensibility of robot-to-human approach motions is significantly higher for trajectories with human-like features.