CRII: CHS: Examining and applying robot sociality to enhance intergroup human-robot interaction

机器人正在成为人们生活中越来越常见的增强功能。他们辅导学生，帮忙做家务，在工厂工作，等等。要想让机器人技术发挥最大的作用，重要的是要了解人们如何最好地与它互动。在这个项目中，研究人员应用社会心理学和群体动力学的理论来改进人-机器人合作。此前的研究表明，人类与他们视为队友的机器人(称为“内部组”)的合作比与他们视为对手的机器人(称为“外部组”)合作得更好。然而，内部群体和外部群体是社会范畴，还不知道机器人必须表现出怎样的社会行为才能从内部群体偏爱中受益。相反，非社交机器人能否避免被视为群外成员的缺点也是不确定的。在这个项目中，研究人员将发展对这些关键因素的理解，并将它们应用到真实世界的人类-机器人团队中，例如用于建筑。这将促进心理科学的进步，展示心理群体动力学的原理在多大程度上延伸到非社会主体，以及这些原理何时崩溃。研究结果将导致机器人设计建议，以改善人与机器人的合作。例如，这项研究可能表明，当机器人的行为一致时，人类和类似机器的机器人可以最有效和顺利地一起工作。为了检查组间的人与机器人的交互，研究人员将建立合作和竞争的环境来引发组效应，并确定机器人必须有多大的社会性才能像组内成员或组外成员一样被对待(研究1)。然后，研究人员将确定社会群体原则，如机器人群体的实体性(凝聚力)，是否会影响与他们的互动，即使机器人是非社会性的(与现实世界中工厂创造的建筑机器人相似；研究2)。由于人类群体将与机器人互动(例如，家庭成员与家用机器人、施工人员互动)，研究人员将研究人类群体类型(例如，朋友、同事)如何影响对机器人的反应(研究3)。由于群体比个人更具竞争力，研究人员将在与机器人进行群体互动之前检查与它们的一对一接触，以减少对它们的偏见(研究4)。最后，研究人员将把研究1-4中的研究成果应用于现实世界，并与他们来自工程学院的合作者一起引入建筑机器人(研究5)。这一研究将为理解人-机器人交互中的群体动力学奠定基础。根据研究结果，研究人员将就哪些机器人形式和组内行为最适合组间互动提供设计建议。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Robots are becoming increasingly common enhancements in people's lives. They tutor students, help around the house, work in factories, and more. For robotic technology to be the most helpful, it is important to learn how people best interact with it. In this project, the researchers apply theory of social psychology and group dynamics to improve human-robot teaming. Previous research has shown that humans work better with robots that they view as teammates (called 'ingroup') than robots that they view as opponents (called 'outgroup'). However, ingroup and outgroup are social categories, and it is not known how social a robot must behave to benefit from ingroup favoritism. Conversely, it is uncertain if nonsocial robots can avoid the drawbacks of being viewed as outgroup members. In this project, researchers will develop the understanding of these critical factors and apply them to real-world human-robot teams, such as for construction. This will promote the progress of psychological science, demonstrating how far principles of psychological group dynamics extend to nonsocial agents and when these principles break down. The results will lead to robot design recommendations to improve human-robot teaming. For example, this research might indicate that humans and machine-like robots can work most efficiently and smoothly together when the robots behave uniformly.To examine intergroup human-robot interaction, the researchers will set up cooperative and competitive contexts to elicit group effects and determine how social the robots must be to be treated like ingroup or outgroup members (Study 1). Then the researchers will determine if social group principles, such as entitativity (cohesiveness) of robot groups, affect interactions with them even when robots are nonsocial (paralleling factory-created construction robots in the real world; Study 2). Because human groups will interact with robots (e.g., family members interacting with a home robot, a construction crew), the researchers will examine how human group type (e.g., friend, coworker) affects responses toward the robots (Study 3). Because groups are more competitive than individuals, the researchers will examine one-on-one contact with robots before group interaction with them to reduce prejudice against them (Study 4). Finally, the researchers will apply their research findings from Studies 1-4 to the real world and the introduction of construction robots with their collaborators from the College of Engineering (Study 5). This research will lay the foundation for understanding group dynamics in human-robot interaction. Based on the findings, the researchers will provide design recommendations about what robot form and intragroup behavior will be best suited for intergroup interaction.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.