Team of researchers from the University of Southern California have created a personalized learning robot especially for autistic children which uses machine learning in order to understand children’s changing needs.
The “socially assistive robot” known as Kiwi is a two-foot tall green feathered robot and gives each kid a personalized class. The team experimented with the robot in the homes of 17 autistic children.
The study was carried out over a month in the homes of 17 autistic children. The children played space-theme math games on a tablet device. The robot Kiwi was used to provide feedback and instruction, this included acknowledging the children’s success when they gave a correct answer, Kiwi also gave tips for every wrong answer.
As the lessons progressed, the team worked on the algorithms in order to adjust the level of difficulty of the games based on the children’s individual needs and an improved feedback.
The study findings showed that by the end of the month the children showed improved math skills, whereas 92% showed improvement in their social skills.
In order to evaluate every child’s level of engagement, the team studied child’s individual behavior, however experts reveal robotic systems are too rigid for such understanding.
Shomik Jain, lead author and progressive degree mathematics student explained, “If you think of a real learning environment, the teacher is going to learn things about the child, and the child will learn things from them.
“It’s a bidirectional process and that doesn’t happen with current robotic systems. This study aims to make robots smarter by understanding the child’s behavior and responding to it in real-time.”
The new study also determines that the robots are more affordable than human care. In addition the robotic care wouldn’t be limited to locations and times. Critics however warned that such benefits could become a risk.
“AI methods cannot and must not be used as a cheaper substitute for treatment by human doctors,” Alena Buyx, a Professor of Ethics in Medicine and Health Technologies at the Technical University of Munich, said last year.
“Human therapists are crucial, but they may not always be available or affordable for families,” said Kartik Mahajan, an undergraduate student in computer science and the study‘s co-author. “That’s where socially assistive robots like this come in.”