Robots helping children with autism spectrum disorder to learn about emotions? Yes, these are real and the future is here—and these bots incredibly helpful to the kids who need them. These tech-time teaching tools are already part of therapy, but now researchers at the MIT Media Lab have upped the robotics game with a “deep learning” network.
Emotional recognition, identification and use isn’t always easy for a young child. Add on an ASD diagnosis, and this type of development often becomes more than a challenge to master. But that doesn’t mean a child with autism can’t or won’t learn about emotions.
Photo: elijahssong via Pixabay
Robots are among the many tools that educators, therapists and specialists use to help kids with autism to engage and understand emotions. Recent research, published in Science Robotics, takes the tech aspect of this therapy to a new level by personalizing the robot’s responses to the child’s reactions.
Instead of only exhibiting emotions to help the child learn, these new robots estimate the child’s engagement and tailor the experience to meet a specific child’s needs. This type of personalization is highly important when working with a spectrum disorder. The recently published study tested this idea with 17 children from Japan and 18 from Serbia. The children, ranging in age from 3 to 13 years old, participated in 35-minute-long sessions with the robots.
So what did the researchers find? To start with, the children interacted with the robots more like they were real people than toys. They also found that the robots’ perceptions of the children’s responses matched the human expert’s perceptions with a 60 percent correlation score. While that might not seem like a big deal, consider that even human-to-human correlation scores typically fall between 50 and 55 percent.
Featured Photo: Pixabay via Pexels