Cooking Up Trust: Eye Gaze and Posture for Trust-Aware Action Selection in Human-Robot Collaboration

Abstract

In Human-Robot Collaboration (HRC), trust is an essential factor that can change over time, and robots capable of estimating a human’s trust and using that information to select their actions can improve the quality of interaction. In this paper, we present our early-stage research on a trust-aware policy for HRC applied to an assistive cooking scenario. We propose to study how physiological signals, such as eye gaze and posture, can be used to estimate the human’s trust level. We describe our experimental platform and the evaluation scenarios used to test the effectiveness of the policy.