Overview
The design of social robots is still a serious scientific challenge. Effective Human-Robot Interaction (HRI) in which robots can learn from, teach and be able to anticipate what humans will be doing and how to react appropriately are still distant. The still uncrossed frontier in robotics design is the ability for the robot to modulate its pace of interaction to the emotional state of the user. To be effective, social robots need to understand the emotional state of the human partner and modulate their actions accordingly. The overall goal of the AIRCA project is to endow robotic agents with the capability of choosing and modulating their actions to maximize the engagement of their human partner in a joint action scenario characterized by a sequence of complementary actions. To this aim AIRCA robot will be endowed with a personality that will allow on one side avoiding stereotyped behaviour, on the other it will support the Human partner to reach and maintain a specific emotional state (e.g. happiness, surprise, fear ). AIRCA robot is of particular interest for dealing with children with Autism Spectrum COndition (ASC). One of the main characteristics of ASC is a social difficulty that impairs effective communication with others and the capability to establish proper joint actions. Ministero dell'Università e della Ricerca MUR - BANDO 2022 Accumulated evidence shows that humanoid robots effectively engage children with ASC in games that require cooperation with them, therefore providing a very powerful support therapy based on joint activity that is at the core of treating social impairment in children with ASC. Starting from these previous studies, we will develop games based for instance on building block constructions to be played with a social robot whose actions (movement, gaze direction) and voice will be continuously modulated in time starting from the ecological estimate of the child's emotional state based on automatic voice processing, gaze tracking and gesture analysis. This technological solution has the potentiality to greatly improve the treatment of children with ASC as it will allow to avoid the occurrence of meltdown states, i.e. the tendency of children with ASC to close in themselves. Furthermore, the interaction with the AIRCA robot can improve the social competence of children with ASC and may facilitate establishing a positive relationship also with another person such as the therapist. At the same time, given the difficulty of these children to interact with others, it will be an ideal testing scenario to define the most suitable modalities of action of a social robot that make its interaction with the Human partner most natural and rich. This would allow the development of the next generation of soc+ial robots that can sustain prolonged interaction with their Human partners in the most effective and natural way.
Project partially funded by PRIN2022 Grant Number G53D23002860006 in the field of Life Science.
For further information, you may contact Prof. N. Alberto Borghese, project coordinator