AuRoRa Project Podcast Presentation

Erin Liebe Professor Tyler Frazier Evolving Solutions- DATA 150

Talking Robots: Kersten Dautenhahn- Therapy Robots for Autism

The podcast episode “Talking Robots: Kersten Dautenhahn- Therapy Robots for Autism” explores The AuRoRa Project, which is dedicated to the integration of robotic therapy into special education classrooms, especially in Autism programs. The children playfully interact with the robot, engaging in fun games that simultaneously develop their social awareness, nonverbal communication, and ability to cooperate with others. This research is more of an observational experiment and therapeutic exercise than it is a clinical study. Kersten Dautenhahn, the founder of the AuRoRa Project, says that they “are not autism researchers, but try to help and assist children with autism in dealing with the social world” by collaborating with psychologists and experts in robotics. However, I believe that by introducing machine learning and by collecting statistics on the interactions between the children and robots, this idea could evolve into an easily applicable tool to use in special education classrooms outside of their experimental groups.

Children with Autism are the target audience of robotic therapy because they tend to prefer being alone. These robots help develop social skills without the stress of interacting with another human. Autistic students often struggle to imitate the behavior of those around them, detect social cues, cooperate with others, and take turns, even in conversation. Simply talking to another person can be intimidating as the child has to forfeit their recluse, stable, and predictable environment for an environment involving an uncontrollable and unpredictable person. Their inclination to withdraw from social interactions is the most defining characteristic of an Autistic person.

The never-ending stressors of understanding social cues are eliminated when playing with a humanoid robot. For this reason, people with disabilities tend to be comfortable with technology and even fascinated by it, especially Autistic children. In the AuRoRa project, the robots are set in the classroom and the children can approach the robots in the classroom if they would like, and most do. They play an imitation game with the robot, where the student performs a movement and the robot will copy it, or vice versa. Not only does this foster the child’s comfort with imitation, but Dauenhahn and her team even observed that if a child moves in a way that the robot cannot imitate, the children become creative and think of new movements to adapt to the needs of the robot. This also allows the child to practice turn taking which is usually a weakness for Austistic people.

The robots allow students to improve the basic social interaction skills that prepare them for pleasant human interactions. By implementing machine learning into robotic therapy for the disabled, we can “facilitate a gradual increase in the complexity of interaction, according to the individual child’s needs and abilities.” If we program the robot to consider the amount of time it takes for the child to initiate or imitate actions, we can make the robot adapt to the abilities of the child. When a child takes a long time to imitate the robot’s action, or does not imitate it at all, the robot could adjust and select from a level of simpler movements to make the game more inclusive to that child. In addition, creating a way to monitor which student is playing with the robot, possibly through an app, allows the robot to track an individual child’s progress by recording their time it took to imitate, the number of successful imitations, and duration of play. The data could be displayed on a time series graph for the teachers to analyze and show in Individual Education Program meetings with the students’ parents. The teachers can also better understand the student’s skills that a child may choose to neglect or fail to display due to the discomfort of social interaction. The teachers would have statistical evidence of the child’s progression whether it be an improvement in quick thinking, social imitation, turn-taking, or attention span. The students could then engage in a game that is tailored to their individual abilities, rather than playing with a robot that challenges a non-verbal and deaf student the same way it interacts with a high-functioning student. Introducing data collecting would improve the AuRoRa project by quantifying the social progress of Autistic students.