We're in This Together

Abstract

Naturalistic studies of human-robot interaction (HRI) in various domains of everyday life, including healthcare, domestic assistance, education, transportation, and other services, show that natural exchanges between people and robots often take place in the broader context of small groups, organizations, and communities. While initial research in HRI focused largely on evaluating the one-on-one interactions between humans and robots, the field is increasingly turning its attention to interactions involving multiple people and/or multiple robots. This talk is motivated by this 'group turn' in HRI, by theories and studies of group interaction in the social sciences, and by our own initial observations of group human-robot interaction (HRI) in public spaces. I will discuss how we can go beyond a dyadic understanding of HRI to incorporate group, organizational, and community goals and experiences into robot design.Group effects are well established in social psychology, and suggest that intergroup interactions are more aggressive and negative than ingroup interactions. Translating this to HRI means that interactions between humans and robots could be more negative if robots are seen as outgroup members, and more positive if robots are seen as ingroup members. Observations of open-ended interactions between people and robots also suggest that people interact with robots differently when they are in groups compared to when they are alone. Over the past five years, we have explored whether group effects from human interaction transfer to HRI in terms of people's reactions to multiple (as opposed to single) robots; characteristics of robot groups (such as synchrony or appearance) that may influence people's responses; tests of interventions that have been demonstrated to reduce prejudice or intergroup bias in humans (such as perspective taking); and tests of other theoretical predictions drawn from work on human intergroup behavior. While some of our findings confirmed theoretical expectations from social psychology, there were many exceptions.Many of our day-to-day small group interactions are in turn situated in various organizations (e.g. schools, hospitals) and even broader communities. The design and application of social robots should therefore incorporate an understanding of the social dynamics, goals, and other factors which are salient to the diverse actors in these broader contexts, and which affect HRI. Along with lab and field research on small group interactions, we also study how robots are designed, adopted, used, and perceived within organizational and community settings. This work suggests that the success of a particular robot is not tied only to the characteristics of the robot itself or the experiences of individual users, but depends on various organizational factors: work load and flow, available financial resources, the needs of and dynamics between diverse stakeholders, and community goals and values. This suggests that the design of robots for organizations and communities should take into consideration these broader factors, and involve community partners in the early stages of designing social robots that can address their shared needs. I will discuss several studies that use this approach.