Robot mediated therapies for improvement in social interaction of ASD children

Abstract

Children are typically divided into three major categories. They are: autistic, down syndrome and typically developed children. In this research, we will be dealing with autistic children. Autistic children are biologically fit, and they usually do not have chromosomal disorder in them as down syndrome have, so their mental and physical development is approximately synchronized and better as compare to down syndrome children. Since this is not true in all autistic children, so few of them face issues in their interaction with people living around them. It is observed that children are more deviated towards toys and especially robots. Three major categories of robots are: aerial, mobile and humanoid robots. Research also shows that appearance of robots also affects autistic children. Closer the appearance of robot to a human being, larger will be the impact on these subjects, which shows their tendency towards humanoid robots. Researchers are introducing humanoid robots in education, improving social interaction, elimination of different disorders of life and many other related things. This thing falls in the category of Human Robot Interaction (HRI) in autism research filed. Different deficits which are common in these children are: lack of joint attention, lack of verbal and nonverbal communication, imitation, and motor movements. In this work we are designing two different robotic therapies, using two NAO robots. First therapy is for measuring joint attention and then it’s improvement and other one is for improving imitation skills based on joint attention, measured and improved in previous therapy. In first intervention, we are designing a multi humanoid robot based joint attention measurement system first and then integrating it with different types of reinforcement stimuli for transforming it into joint attention improvement system. These reinforcement stimuli are arranged from least effective to most effective and their effectiveness varies depending upon the behavior and sensitivity of each child. In second intervention, we are designing a multi humanoid robot based imitation system actuated by joint attention of ASD child. In both interventions, we are accessing the cognitive state of brain of ASD child as well. Accessing cognitive state of brain, before and after each therapy, is added just to note the impact of therapy on ASD child. We are noticing the mental presence of child via accessing his / her cognitive state of brain. We are also utilizing skeletonization for tracking motions of different body parts of child during both interventions.