How can human-robot interaction be improved by making robots more socially intelligent? This is the key question at the heart of socially assistive robotics (SAR): a new field of intelligent robotics that focuses on developing machines capable of assisting users through social rather than physical interaction, in order to encourage people to have the drive and motivation to do their own work, for improved health and wellness.

Prof Matarić’s research brings together engineering, neuroscience, health, social, developmental and cognitive sciences to create robots that can serve as coaches, motivators and companions. This requires personalising human-robot interaction through appropriate speech, gesture and body language; the embodiment is the most important aspect.

Successes include coaching stroke patients to perform rehabilitation activities, helping children with autism to learn social skills, encouraging teens at risk of type-2 diabetes to exercise, motivating first graders to make healthy food choices, and helping elderly patients with Alzheimer's disease to stay engaged.

This talk described those projects and the associated research into embodiment, modelling and steering of social dynamics. Illustrated with many videos, it looked at long-term user adaptation for SAR.