Getting acquainted with a developing robot

Abstract

Two factors that been suggested to influence the ways in which people interact with robots, namely users' initial expectations on the one hand and their increasing acquaintance with their robotic partner due to repeated interaction over time on the other. In the current study, eight participants interacted with a humanoid robot in five different sessions. Between the sessions, the robot was trained on the linguistic material presented to it by its human tutor in the preceding session, and thus the robot exhibits increasingly more knowledge of the domain. The results uncover the interaction between users' preconceptions and feedback-driven interactional effects that shape human-robot interactions. While considerable differences between users can be observed, all users respond to the robot's feedback and increasing linguistic capabilities in comparable ways.