Social Robots that can Sense and Improve Student Engagement

Abstract

It is highly likely that classrooms of the future will feature robots to assist the human teachers. Tutor robots will be valued for their capacity to motivate learners and to provide affective support during learning activities, which will require from them to be able to understand the students' affects and behaviours, and to respond to these through appropriate expressive motions. In this paper, we investigate the impact a robot teacher's behaviour has on the students' level of engagement. We outline research work we have carried out to tackle four of the challenges inherent to the effective deployment of social robots in classrooms: (1) sensing and understanding learners' affective states and behaviours in class; (2) combining affect and behaviour understanding to capture classroom's dynamics; (3) knowing what gestures a social robot should use as a learning facilitator; and (4) equipping the tutor robot with expressive and motivational capabilities.