Improving Request Compliance through Robot Affect

Abstract

This paper describes design and results of a human robot interaction study aimed at determining the extent to which affective robotic behavior can influence participants' compliance with a humanoid robot's request in the context of a mock up search and rescue setting. The results of the study argue for inclusion of affect into robotic systems, showing that nonverbal expressions of negative mood (nervousness) and fear by the robot improved the participants' compliance with its request to evacuate, causing them to respond earlier and faster.