Robots have been proven to be beneficial in helping children with autism remember how to do things and to learn new tasks. The pilot study that proves robots are as effective as humans if not more effective was developed by Maja Matarić, University of Southern California Viterbi Vice Dean for Research and Chair in Computer Science, Neuroscience and Pediatrics, and colleagues at the University of Southern California.
The development of robot-helpers was based on the concept of graded cueing feedback. Graded cueing feedback is an occupational therapy technique that provides increasingly specific prompts to a person learning a new skill or relearning a skill. The robot is a learning machine that can adapt to the response of an autistic child in learning situations. The robot is small and portable allowing the child to take their humanoid helper friend almost anywhere.
The study analyzed the response of 12 high-functioning autistic children to robot assistance. Two groups of six children each received different types of instruction from their robot. One group was paired with a robot that practiced graded cueing feedback and the other group had a robot assistant that did not present as sophisticated a level of cues. The test involved a comparison of the children’s ability to learn and remember a series of positions in an imitation game. The robot provided the positions to imitate.
The children paired with robots that used graded cueing feedback learned new tasks faster and stuck with the task longer. The robot indicates a correct response with flashing red eyes and repeats the instructions with added information in the event of an incorrect response. The robot can speak and says “good job” when the child performs a task successfully. The idea is to pair every autistic child with a robot helper to allow them to learn and retain the skills they have learned. The robots used are sophisticated enough to be adaptable to an individual’s personality and learning ability.
