Humanoid Robots Help Children with Autism Learn Interaction

Maja Mataric´(University of Southern California)

29 August 2014. Engineers and computer scientists at University of Southern California in Los Angeles show how commercial humanoid robots can help children with autism spectrum disorder learn basic social behavior. The team from the lab of Maja Mataric´, director of USC’s Robotics and Autonomous Systems Center, presented its findings earlier this week at the IEEE International Symposium on Robot and Human Interactive Communication in Edinburgh, Scotland.

Autism spectrum disorder is a collection of neurodevelopmental conditions, marked by communication difficulties and impaired social interaction, as well as repetitive and stereotyped patterns of behavior. At age 8, some 1 in 88 children have autism spectrum disorder, according to Centers for Disease Control and Prevention. Classic autism is considered the most severe form of the syndrome.

The study devised by Mataric´ and colleagues aimed to find out how children with autism spectrum disorder react to humanoid robots that provide graded cueing, a process shaping behavior that uses increasing specific prompts to improve everyday skills of people in recuperative therapy. Graded cueing is employed for conditions such as brain injury, or to learn new or lost skills, as in the case of children with autism spectrum disorder.

For this pilot study, the USC researchers enrolled 12 children with autism spectrum disorder, who played an imitation game, a form of copycat, with a humanoid robot, the Nao model made by Aldebaran Robotics. The game asked the children to imitate 25 different arm poses.

Children performing the pose correctly received positive feedback from the robot in the form of flashing green eyes, or a nod and the spoken words “good job” in response. Half of the group received graded cueing feedback from the robot partner when the child did not accurately perform the pose. The prompts first began as verbal cues, followed first by more detailed instructions and then demonstrations. For the children in the other (control) group, the robots only repeated the command.

The results show children receiving the graded cueing feedback until the correct pose was performed improve or maintain their performance, while those receiving only a repetition of the command show no improvement or regress. The findings suggest receiving the varied feedback from a socially assistive robot can help children with autism spectrum disorder learn simple interactive skills.

The pilot study is one of a number of projects underway in USC’s Interaction Lab developing adaptive or personalized socially assistive robots to help people with special needs. “There is a vast health care need,” says Mataric´in a university statement, “that can be aided by intelligent machines capable of helping people of all ages to be less lonely, to do rehabilitative exercises, and to learn social behaviors.”

Read more:

• Text Messages Help Young Adults Reduce Binge Drinking
• Online Clinical Trial with Children Tests Autism Therapy
• Mobile App/Sensor Designed to Help Parents Control Stress
• New Jersey Tech, IT Company to Develop Autism Device
• Mobile Low-Power Gesture-Recognition System Developed

*     *     *