Robotic prosthetic arms are high tech, big ticket items. Yet their purpose is simply to restore an amputee’s basic human abilities. This gap between need and cost didn’t sit well with Hamayal Choudhry, a mechatronics engineering student at the University of Ontario Institute of Technology. “We have so much cutting edge technology these days,” he says, “yet it’s only being used for commercial or industrial purposes. But there’s this whole other opportunity to improve people’s lives.”
Choudhry had an opportunity to act on his philanthropic engineering
instincts at a hackathon at the university this January. It was there that he ran into a friend from middle school, who he hadn’t seen in years, Samin Khan. Khan was studying computer science at the university. “We shared similar sentiments,” says Choudhry. So they decided to team up. “That’s where we came up with this idea to combine computer vision and machine learning in a way that was very accessible and cheap for everyone to use.”
Within 24 hours they had a prototype of a robotic arm
with a camera in its palm to recognize objects—no mean feat as the hackathon took place over the weekend, a tough time to find spare engineering parts. So they had to make do with what they had. “We pulled together cardboard, straws, and skewers and stuff,” says Choudhry. They 3-D printed the outside of the arm, used various servo motors—calibrated in the code to work similarly—and rubber bands for tension for the fingers. With this day-old prototype the pair managed to win Microsoft’s Sponsor Award at the hackathon.
In the months since then, the SmartARM, as they call it, has gone through some changes. Choudhry and Khan looked at other prosthetics and found that there was little they could borrow from them. Existing robotic prosthetics use complicated neurological connections which are used to determine movements. They’re made with high-grade materials and designed to be used over the long term. “That’s how they drive the cost up to $100,000,” says Choudhry. “The functionality that we had in mind was a lot different than what was out there.”
Instead of relying on signals sensed from the body, Choudhry and Khan wanted to use computer vision to do the processing. The computer would identify the object a person might want to grab and the robotic hand would have the appropriate grip ready.
To keep costs down, they are now using a custom-designed chip—it does exactly what they need and has a small footprint as well. They 3-D print the body of the arm so “someone who wants to embrace the fact that they are an amputee might be interested in having a futuristic sleek robotic prosthetic.” For those who don’t want to stand out, they have a cosmetic glove that goes over the prosthetic to make it look a normal hand. And the rubber bands have disappeared, replaced with low-cost gears and motors.
The SmartARM was originally meant to recognize objects and then form a grip accordingly. But recently Choudhry and Khan have started moving toward 3-D shape recognition. Instead of having to know a bottle is a bottle, it will simply recognize the cylinder. They are also looking into using tactile sensors on the fingers so the hand will better know the strength of the grip to use.
But the technical innovations are only a part of the work that’s gone into making the arm. The pair have been in contact with amputee groups to understand their actual needs. They learned, for instance, that the socket of a prosthetic—the part that touches the amputee’s body, is covered by Canadian insurance. So they didn’t need to make one for their Canadian market. But they did have to mount all their electronics in the palm of the hand instead of the socket. They also learned that in the U.S. only the socket is considered a medical device. The rest is an accessory, so the team will not have to seek FDA approval.
By keeping the price low, people that might not otherwise be able to afford a robotic arm will have a chance to grab and grip things again. This includes children. With a price-tag in the five and six digits, most amputees don’t chose a robotic prosthetic until they’re done growing. With a low cost 3-D printed
arm, a child could get a new one with each growth spurt.
Choudhry and Khan—still just 20-years-old—have done their technical, legal, and sociological research in a matter of months. Their product wowed judges at this year’s Microsoft Imagine Cup. They won first prize and received more than $130,000 in cash and grants. The award speaks not just to their technical abilities, but to the depths of their thinking about real needs.
“We are not just setting how to solve problems with technology,” says Choudhry. “Our overall vision is to combat real-world problems in any way we can.”
Michael Abrams is an independent technology writer based in New Jersey.
Recommended for You: How Engineers Can Give Back