HUMAN-ROBOT TRUST: Is Motion Fluency an Effective Behavioral Style for Regulating Robot Trustworthiness?

Abstract

Finding good behavioral styles to express robot trustworthiness will optimize the usage of robots. In previous research, motion fluency as behavioral style was studied. Smooth robot motions were compared with trembling robot motions. In a video experiment an effect of motion fluency on trust was found, while in an Immersive Virtual Environment (IVE) experiment, no effect was observed [1]. In this research, we explored the question whether the motion fluency effect is present in a short version of an IVE task and disappears when the task is longer. Results indicate this is not the case. Several explanations for this null-effect are discussed and several recommendations for further human-robot trust studies are provided.