Prosthetics have come a long way from a wooden leg or peg. Engineers are constantly working on ways to reduce challenges for amputees with new and better prosthetics that give them back certain abilities. Robotic limbs revolutionized the prosthetics market in 2015 by combining sensor technology with prosthesis movement.
While robotic limbs are quickly becoming a go-to choice for amputees, they require a lot of practice hours before they are ready for use. Prior to using a robotic limb full-time, an amputee will have to spend hours allowing the limb to adapt to their particular movements. In addition, the limbs may need a variety of tweaks and alterations to make them just right for the wearer.
Helen Huang, a professor in biomedical engineering at North Carolina State University and the University of North Carolina, says, “If you wanted to make this clinically relevant, there are many, many steps that we have to go through before this can happen. So far it’s really just to show it’s possible. By itself that’s very, very exciting.”
In an effort to eliminate this extensive process, researchers have developed an artificial intelligence (AI) system that can adjust a robotic knee, allowing the wearer to begin walking on level ground within minutes. The AI utilizes a trial-and-error algorithm to recognize patterns in sensor data, set safety limits, and learn patterns for stable walking motions.
“Our body does weird things when we have a foreign object on our body,” says Jennie Si, professor of electrical, computer, and energy engineering at Arizona State University and co-author of the paper. “In some sense, our computer reinforcement learning algorithm learns to cooperate with the human body.”
This advancement was no small achievement. There are still limitations to creating a system that is meant to work with a human collaborator on specific movements and gestures. While AI systems can take their time to learn new skills, an amputee may only have a few minutes in intervals before they need to rest. The AI system also can’t learn from mistakes, like falling, because that could injure the amputee. So while it would provide helpful data to further the AI’s learning, it isn’t optimal for the wearer.
The AI system is still in testing and nowhere near ready for implementation. There are many challenges ahead for the researchers including testing for hard terrain, inclines, and stairs. Currently, the AI system is wired to the robotic limb, meaning amputees have to come into the lab. As the researchers continue their work, they hope to produce a wireless version so amputees could enjoy their freedom and only return to the lab for minor tweaks or alterations.