Proceedings of the SREC (Social Robots in Therapy and Care) Workshop at HRI 2019

TL;DR

This paper discusses the challenges and requirements for social robots in therapy and care, emphasizing personalization, context awareness, and adaptive behaviors to effectively support diverse user needs.

Contribution

It highlights the key technical and design challenges for social robots in therapy, focusing on perception, personalization, and adaptive interaction strategies.

Findings

Robots need to perceive social cues and emotional states accurately.

Personalization of robot behavior enhances therapy effectiveness.

Adaptive interaction strategies are crucial for diverse user needs.

Abstract

Robot-Assisted Therapy (RAT) has successfully been used in Human Robot Interaction (HRI) research by including social robots in health-care interventions by virtue of their ability to engage human users in both social and emotional dimensions. Robots used for these tasks must be designed with several user groups in mind, including both individuals receiving therapy and care professionals responsible for the treatment. These robots must also be able to perceive their context of use, recognize human actions and intentions, and follow the therapeutic goals to perform meaningful and personalized treatment. Effective interactions require for robots to be capable of coordinated, timely behavior in response to social cues. This means being able to estimate and predict levels of engagement, attention, intentionality and emotional state during human-robot interactions. An additional challenge for social robots in therapy and care is the wide range of needs and conditions the different users can have during their interventions, even if they may share the same pathologies their current requirements and the objectives of their therapies can varied extensively. Therefore, it becomes crucial for robots to adapt their behaviors and interaction scenario to the specific needs, preferences and requirements of the patients they interact with. This personalization should be considered in terms of the robot behavior and the intervention scenario and must reflect the needs, preferences and requirements of the user.