UC San Diego

Robots are increasingly being used to support human groups and teams. Action teams (such as in emergency healthcare or disaster response) experience high workloads, must work quickly, and must make decisions under uncertainty. They often perform critical tasks where errors or delays can lead to grave harm. Therefore, robots in these teams must be designed and contextualized to not contribute to errors or interrupt human team workflow. In this study, we conducted a behavioral analysis of human-robot action teams as they collaborated with each other to complete tasks in escape rooms to better understand how a robot's actions influenced intra-team dynamics and how team characteristics affected human-robot teaming. Our findings highlight the importance of robots' functional and social contributions for acceptance within teams, the adaptive nature of teaming behavior in response to perceived robot capabilities, and the significance of the robot's nonverbal cues in shaping human expectations. These insights offer valuable implications for designing effective human-robot interactions in collaborative environments.