Considerations about social norms compliance in a shared elevator scenario

In this position paper, we present our ongoing research on socially acceptable robot navigation for an indoor elevator sharing scenario. We highlight the current challenge of designing interactions for a robot behavior both effective in accomplishing its tasks, but not intrusive and at risk of breakdown. We discuss the advantages and limitations of modelling these behaviors on the base of a full humanlike approach. In particular, we discuss the risk that a full human-like approach presents of creating the illusion of social competence. It has been observed that this illusion often leads to breakdowns when the technology is faced with complex and potentially ambiguous social situations. We propose the principle of “machine-like yet humanfriendly” behavior as a way to address the risks of the full human mimicking approach. We believe that this approach can provide more understandable and less disruptive behaviors for routine integration into human spaces. We conclude discussing the need for a multi-layer experiment set-up to evaluate and validate this approach.

Poster