Therapist project

Our objective

The objective of this proposal is the development of THERAPIST, a socially interactive robot for neuro-rehabilitation assistance. Our initial hypothesis is that patients will get consistently engaged in a therapeutic non-physical interaction with a social robot, facilitating the design of new robot-based therapies that will improve the patient recovery time and reduce the overall socio-economic costs. In support of this hypothesis, recent pioneering studies have shown that paediatric patients can be easily driven into highly attentive and collaborating attitudes by letting them interact with socially interactive robots. These initial works have established challenging directions of research that have yet to be further explored. This project will advance in this direction, testing the initial hypothesis by establishing a road book for designing, building and validating socially interactive robots, used as therapist assistants in real hospital rehabilitation scenarios. One critical factor to achieve these goals is to offer the patient effective social interaction. The level of engagement that the patient will commit to a socially assistive robot, largely depends on its ability to gain the patient's confidence and maintain their interest. To this end, the human-robot interaction (HRI) process should be natural and intuitive, and enhanced with non-verbal multimodal communication channels (gestures, facial expressions and other means of embodied communication). Thus, as part of this project we will design and build a cognitive architecture for THERAPIST that will include automated behaviour detection and activity recognition, pro-active ability, task-dependent attention and failure recognition. It will also provide automated understanding of human emotions, through vision and voice processing, and believable generation of emotions through multiple channels, such as voice synthesis, and face and body movements. This reactive, socially pro-active behaviour of THERAPIST should be always modulated by the unfolding of a specific therapy. We do not forget that this is our main objective and that interaction is a tool to improve the motivation and adherence of the patient to the treatment.

Proposed use case:
motivation, design and evaluation.
read more
Inner models: a tool
for prediction, evaluation and selection.
read more
Deliberative level:
automated planning, monitoring and learning.
read more
Common, model-update
and task-oriented
read more