Recent technological advancements have led to the emergence of supportive robotics to help children with developmental disabilities become independent. In conventional research, in robot therapy, experiments are often conducted by operating the robot out of the subject's sight. In this paper, robot therapy using a system that can autonomously recognize the emotions of a child with developmental disabilities and provide feedback was developed. The aim was to quantitatively infer emotional changes in children using skin conductance (EDA) during robot therapy. It was demonstrated that the robot could recognize emotions autonomously and provide feedback to the subjects. Additionally, a quantitative evaluation was conducted using EDA. By analyzing the symptoms related to developmental disorders, it may be possible to improve the recognition rate and tailor therapy based on symptoms.
Keywords: EDA; HRI; children with autism; emotions; robot therapy; social robot.