Responses to Human and Robot Errors in Human‒Robot Collaboration: an Fnirs Study

Fengyuan Liu,Yishu Ji,Xin Lei,Pei-Luen Patrick Rau

DOI: https://doi.org/10.1007/978-3-031-60913-8_19

2024-01-01

Abstract:Performance assessment in the era of human-robot collaboration poses new challenges. Will human managers display varying responses to the success and failure of human versus robot employees? This study aims to investigate people's responses to success and errors made by humans compared to those made by robots using self-report measures and neuroimaging techniques. Twenty-four participants were asked to imagine themselves as managers tasked with reviewing videos of human-robot collaboration and evaluating the human and robot employees in the video. Results showed that, when the employee performed correctly, participants assigned more credit to the employee and showed stronger positive emotions when the employee was a robot than a human. When the employee made an error and caused failure, participants attributed more blame to the employee and showed stronger negative emotions when the employee was a human than a robot. Additionally, employee errors resulted in decreased trust, and the trust damage caused by human errors was higher than that caused by robot errors. Furthermore, the functional near-infrared spectroscopy technique showed that viewing robot errors caused decreased activation in the prefrontal cortex. These findings enrich our understanding of attribution, trust, and emotions in human-robot collaboration from the perspective of human managers, providing practical managerial implications.