To Err Is Robot: How Humans Assess and Act toward an Erroneous Social Robot

Abstract

We conducted a user study for which we purposefully programmed faulty behavior into a robot's routine. It was our aim to explore if participants rate the faulty robot different from an error-free robot and which reactions people show in interaction with a faulty robot. The study was based on our previous research on robot errors where we detected typical error situations and the resulting social signals of our participants during social human-robot interaction. In contrast to our previous work, where we studied video material in which robot errors occurred unintentionally, in the herein reported user study, we purposefully elicited robot errors to further explore the human interaction partners' social signals following a robot error. Our participants interacted with a human-like NAO, and the robot either performed faulty or free from error. First, the robot asked the participants a set of predefined questions and then it asked them to complete a couple of LEGO building tasks. After the interaction, we asked the participants to rate the robot's anthropomorphism, likability, and perceived intelligence. We also interviewed the participants on their opinion about the interaction. Additionally, we video-coded the social signals the participants showed during their interaction with the robot as well as the answers they provided the robot with. Our results show that participants liked the faulty robot significantly better than the robot that interacted flawlessly. We did not find significant differences in people's ratings of the robot's anthropomorphism and perceived intelligence. The qualitative data confirmed the questionnaire results in showing that although the participants recognized the robot's mistakes, they did not necessarily reject the erroneous robot. The annotations of the video data further showed that gaze shifts (e.g., from an object to the robot or vice versa) and laughter are typical reactions to unexpected robot behavior. In contrast to existing research, we assess dimensions of user experience that have not been considered so far and we analyze the reactions users express when a robot makes a mistake. Our results show that decoding a human's social signals can help the robot understand that there is an error and subsequently react accordingly.

On the impact of different types of errors on trust in human-robot interaction

“…The presence of errors has been reported to result in increased anthropomorphism and likeability, despite a reduced task performance (Salem et al, 2013). Mirnig et al (2017) found no significant impact of errors on a final perceived level of trust in a robotic assistant but also found an increase in likeability. Guznov, Lyons, Nelson, and Woolley (2016) also found no statis-tically significant impact on self-reported trust levels in automation despite manipulating both error type and severity.…”

“…Whilst others conclude the perceived intelligence of the robot increased after having made a mistake and attempting to put it right, but only when the new method was error-free (Lemaignan, Fink, & Dillenbourg, 2014;Hamacher et al, 2016). Mirnig et al (2017) allowed for the classification of real errors into types; social norm and technical. They also highlighted that all robotic errors could be classed as technical from a roboticists point of view in contrast with a naive participant.…”

“…as of yet, perform flawlessly, and as such investigating the effect of robot errors on the resulting trust levels is a well researched topic in human-robot interaction (HRI) (Hancock et al, 2011;Mirnig et al, 2017). Salem, Lakatos, Amirabdollahian, and Dautenhahn (2015) suggest that there is a lack of adequate definitions of trust, specifically within HRI.…”

The perception of a robot partner’s effort elicits a sense of commitment to human-robot interaction

“…Previous research has shown that humans' patience towards a robot that performs suboptimally can be increased if the robot employs a mitigation strategy such as seeking human assistance and/or adapting its approach (Lee et al, 2010;Brooks et al, 2016;Mirnig et al, 2017), or expresses a negative emotional reaction and attempts to rectify its mistake (Hamacher et al, 2016). The present study aimed to build upon this research by addressing the more general question of how to sustain human interactants' willingness to persist in interacting with a robot partner despite increasing boredom or frustration --irrespective of whether that boredom or frustration arises from errors on the part of the robot or from the nature of the interaction itself.…”

Leolani: A Reference Machine with a Theory of Mind for Social Communication

“…People make mistakes; but machines err as well [14] as there is no such thing as a perfect machine. Humans and machines should therefore recognize and communicate their "imperfectness" when they collaborate, especially in case of robots that share our physical space.…”