Design of a Parametric Model of Personal Space for Robotic Social Navigation

摘要

The design of socially acceptable behaviours is becoming one major issue for the development of robots that are able to interact with humans in unconstrained environments. In particular, social behaviours such as gazing, mutual positioning or gesturing allow robots to initiate and maintain an information exchange with humans. This paper focuses on (1) the study of mutual positioning between a small humanoid robot and a person through two psychometric experiments and (2) the design of a parametric model of the personal space based on the results of the two experiments. Results suggest that human-human interpersonal distances are shorter than human-robot interpersonal distances during a communication exchange, at least for the small humanoid robot used in our experiments. We also found that participants evaluate different directions of approach in a significantly different way.