The Speakers

Recovery from Ethical Errors in HRI

Abstract: In this talk I will discuss ethical errors in HRI from two different perspectives. First, I will briefly discuss our lab's research on norm violation response. Then, I will discuss various ethical challenges our laboratory has faced over the past ten years, and the various strategies we have adopted for dealing with those challenges: from adopting new techniques, to abandoning research directions, to rethinking funding strategies.  

When Robot Errors Meet Human Biases

Abstract: Robot errors are often assumed to lead to immediate loss of trust, requiring mechanisms for trust repair. However, evidence suggests that people do not always take robot behaviour at face value. Instead, they frequently rely on social cues and stereotypes when forming and maintaining trust, even when these cues conflict with observable behaviour. 

In this talk, I combine findings from a series of experiments examining trust formation, violation, and repair in HRI. I will show how design features such as voice and joint attention influence trust, despite clear evidence of a robot's un/trustworthiness. I will also discuss strategies that robots can use to recover trust after errors and to support more appropriate long-term trust calibration.