Movia service robots tackle autism


Chris is standing on top of a desk. He’s short, so this way he can see when his friend walks through the door.

“How are you?” asks Chris, waving.

Waving back, his friend responds, ready for whatever games they’ll play. There will probably be some breathing and stretching. They might work on eye contact. Drumming will definitely be involved.

Half an hour later, his friend goes back to his first-grade class. Chris stays behind until his programmer shuts him down.

Chris, you see, is a robot.

“Chris” is a product of Movia Robotics, a collaborative robotics software company that works with clients to create service robots. In West Hartford, Conn., the system is improving the social and academic skills of special needs children, particularly those with Autism Spectrum Disorder (ASD).

Tim Gifford, Movia’s founder, was researching artificial intelligence at University of Connecticut in the early 2000s when his wife inspired the idea that would become his company.

“My wife is a teacher and she was telling me about more and more kids presenting at school with autism,” he explains. “There weren’t many good tools for working with them.”

The U.S. Centers for Disease Control (CDC) noted the increase as well: From 2008 to 2010, the number of children presenting with autism jumped from one in 88 to one in 68.

The conversation spurred Gifford to build on his research, creating something that could meet the needs of his wife and her students. In 2009, he started with a National Institutes of Health (NIH) research grant, co-investigated with Anjana Bhat. One year later, Movia Robotics was officially launched and licensed, and in 2012, after several beta trials, the company sold its first product as a consulting service to the West Hartford school district in Connecticut.

“We said we weren’t ready,” says Gifford, chuckling as he recalls the company’s first days. “They still said they’d take it.

Four years later, Movia still works with young students in West Hartford, and the success of the program has thrilled Janet Gregorian-Michaelsen, the school social worker.

“To work in a district that does this, where they took this risk, has been really satisfying for me,” she says.

Currently, Gregorian-Michaelsen works directly with Movia to plan the half-hour sessions her students have with Chris. The resulting programming is tailored to each child’s individual education plan, setting the baseline for what Chris will ask, say and do.

A Movia programmer will then be present at the school for the session, in case the student asks an unexpected question. They are ready to quickly code so that Chris can respond naturally.

This collaboration creates an environment where Chris exists somewhere between human and object–his actions are human-enough for students to interact with, but object-enough not to be stressful. It also creates a situation where the students can take their time.

“What Chris brings to the table is no judgment,” says Gregorian-Michaelsen, who has a child of her own with autism. “Children with autism can have slow processing, so he’s not impatient, there’s no body language saying ‘come on!’”

“Now you’ve got a kid practicing language, getting self-esteem and taking risks,” she continues.

The activities these children do with Chris are translating beyond the sessions–overall, the students working with Chris have improved their engagement in class. One parent even noted seeing more engagement and communication at home.

For both Gifford and Gregorian-Michaelsen, this is the definition of success–the ability for a child to learn a skill with a robot and apply it to their daily life.

“For me as a social worker, when you see this, that means everything,” says Gregorian-Michaelsen. “That’s true organic change.”

Still, the service is not perfect. Only a select number of students can achieve that success through Movia. Current economics and availability limit access–it is not cheap to build a robot or have it in your class, and programmers must be available to run the sessions.

This also means Chris is only used with students below the third grade, unless they are at the low end of the autism spectrum. Older students will understand there is someone else typing in the room.

For Gifford, these are not limitations, but challenges to meet. His next goals for Movia are to find partners to bring an economical model of his product to target markets, thus reaching more kids, and to develop a next-generation system that does not require special education or computer training to use.

“We want our software to be able to run on multiple robots, across different platforms,” says Gifford. “We want to create content that’s not only useful for special needs kids, but for the general public.”

Movia is still a small company, funded entirely from small investors and bootstrapping. To expand to the next level, Gifford took Movia through two Boston-based MassChallenge accelerators, in 2015 and 2016. The accelerator works with companies that demonstrate impact, providing them with mentors and resources as they compete for a prize of $1.5 million in zero-equity.

Though Movia Robotics never came away with the money, Gifford saw the experience as essential to the company’s next phase.

“Being in the group, working with mentors, that was really for us the goal,” he says. “It was more about learning and refining our work, especially with the scope of what we’re trying to do.”

His dream for his company and what it can do for kids is ambitious, but Gifford says he’ll persevere.

“The robots are coming,” he says.