Danesh Tarapore

This project investigates how social and collaborative robots can adapt their behaviour to individual users during human–robot interaction. Rather than assuming that all users have the same preferences, tolerances and interaction styles, the research develops user-aware decision-making frameworks that learn from both explicit feedback and implicit behavioural signals.

The work combines online learning and constrained decision making with multimodal sensing of human behaviour, including facial expression, body pose and verbal interaction. A particular focus is understanding and reducing user frustration while enabling robots to learn individual preferences and provide effective assistance.

The research has been evaluated through human–robot interaction studies involving TIAGo and Pepper robots across collaborative manipulation, assistive localisation and social interaction scenarios. The broader aim is to develop autonomous robots that adapt their behaviour during interaction in response to the needs and reactions of individual users.