Antecedents of trust in human-robot collaborations

Abstract

Robotic systems are being introduced into military echelons to extend warfighter capabilities in complex, dynamic environments. While these systems are designed to complement human capabilities (e.g., aiding in battlefield situation awareness and decision making, etc), they are often misused or disused because the user does not have an appropriate level of trust in his or her robotic counterpart(s). We describe a continuing body of research that identifies factors impacting a human's level of trust in a robotic teammate. The factors identified to date can be categorized as human influences (e.g., individual differences in terms of personality, experience, culture), machine influences (e.g., robotic platform, robot performance in terms of levels of automation, failure rates, false alarms), and environmental influences (e.g. task type, operational environment, shared mental models). A framework for human-robot team trust was constructed, which is evolving into a working model contingent upon the results of an on-going meta-analysis.