A Magic Touch

By Anna Lynn Spitzer

Irvine, Ca, May 15h, 2014 -- Carl-SJR isn’t your typical-looking robot, if, in fact, there is such a thing. It doesn’t have arms and legs, or other human-like features, but it enjoys being touched and rubbed. It’s these traits that may just be what make the robot a winner with its target audience – children with autism and other developmental disorders.

From left: Krichmar, Chou and Bucci get affectionate with Carl-SJR.

“Children with disorders such as autism and ADHD can sometimes be intimidated by people and non-repetitive or sudden movements,” says Jeff Krichmar, UCI cognitive sciences professor and Center for Cognitive Neuroscience and Engineering director. “This is where a lot of newer therapies like trained animals and especially designed, socially assistive robots are stepping in, helping to teach social interaction skills.”

The need is acute; since 2000, the U.S. Centers for Disease Control and Prevention report that the number of children diagnosed with autism has nearly doubled from 1 in 150 to 1 in 88 in 2008, the center’s last published surveillance year.
  
Enter Carl-SJR, short for Cognitive Anteater Robotics Laboratory – Spike Judgment Robot. Developed in Krichmar’s social sciences research lab, the surface of its “body” is covered with 67 track balls, the same as those used in Blackberries and older computer mice. When touched or petted, the balls send a signal to LED lights that can display in a variety of colors. The goal is for the user to get comfortable interacting with an object that responds to his or her actions.

“A lot of socially assistive technologies rely on a secondary sense, like vision, to initiate and operate, and they usually require the use of touch screens and articulating movements with your fingertips to engage with the technology,” says Ting-Shuo Chou, a computer sciences graduate student who works in the lab. “We wanted to create something that responds to touch on a larger scale, more like petting, using something that isn’t vision-based so that the sense of touch could be emphasized.”

Chou worked with Krichmar and Liam Bucci, an engineer on staff in the lab, to design Carl-SJR with this in mind. The track balls, which cover a large surface area, are connected to a complex circuit board and fully self-contained computer unit that records all movements. This will eventually allow the robot to learn and adapt its patterning to the user, says Krichmar. It can also send commands, like a Simon Says color game, to help the user learn to mimic patterns using both touch and vision.
 
The researchers were recently awarded a provisional patent for the technology behind Carl-SJR. And while they are in the process of applying for grant funding to further develop it, they have already begun a pilot study at UCI’s Child Development School, where the robot is being introduced to children as a behavioral intervention during free-time play.

“Many children with developmental disorders have tactile or touch impairments. Carl-SJR addresses this problem by playing fun, engaging games with children where they learn appropriate social skills,” Krichmar says. “Carl-SJR has the ability to learn what the child likes or dislikes and can tailor the game to their needs.”