Public defence: Marieke van Otterdijk

Main research findings

Unlike humans, robots find it difficult to deal with the uncertainty and unpredictability of everyday life. This shortcoming is problematic if we would like to use robots as autonomous assistants. Humans use their intuition to deal with the unpredictability of situations. Intuition allows people to adapt their behavior to a situation and make choices with minimal effort. If robots had this skill, they would be better able to support us. To successfully model intuition for robots, we must first understand how human intuition works. This dissertation examines how human intuition works by focusing on immediate perception, reactions, and mental effort to model intuitive cognition for robots. Using eye tracking, interviews, and observations, we have investigated how people understand, interpret, and respond to the non-verbal behavior of robots in different situations. These studies gave us insight into the human intuitive thinking process. Furthermore, we discovered which components of observing a robot's behavior affect the user's intuitive understanding and how emotions and social context affect intuitive understanding. We combined these findings into a framework for intuitive cognition for robots.

Gallery