ANN ARBOR, Mich. – The University of Michigan in Ann Arbor has partnered with Shirley Ryan AbilityLab to create an open-sourced artificially intelligent prosthetic leg.
On top of hoping to improve the quality of life for patients needing prosthetics, the open-source nature of the design allows members of the scientific community to further innovate and advance their research in bionics.
According to Elliott Rouse, lead designer and faculty at the University of Michigan’s Robotics Institute, having an open-source design that is downloadable, along with a website that walks researchers through the design and programming process, makes the work being done by the U-M and Shirley Ryan AbilityLab more accessible as well as duplicable.
Despite its state-of-the-art technology and programming, the leg’s design has been kept fairly simple. Two flat, electrical motors, one in the knee joint and one in the ankle joint, have been incorporated into a modular design with an onboard power supply and control electronics. The design allows the leg to perform finer, more controlled movements by the user.
While open-sourced prosthetics are not new to the field, most other open-source projects focus on upper extremities such as the hand. The U-M’s leg is one of the first to focus solely on low body parts.
According to Rouse, many researchers working with prosthetics and with people who have disabilities are required to build their own systems. By having the new leg open-sourced on a common platform, other researchers can build upon the work pioneered by the U-M and Shirley Ryan AbilityLab and better support their patients. Companies and labs can build their own prototypes and versions of the leg while discussing modifications with researchers throughout the scientific community.
✉ Like what you're reading? Sign up for our email newsletter here!
“This represents the future of research -- rapid prototyping of open-source robotic hardware and embedded systems with shared code,” Rouse said.
The AI portion of the leg, led by Levi Hargrove, director of the Center for Bionic Medicine at the Shirley Ryan AbilityLab in Chicago, has been designed to adapt to different activities and movements. Muscle contractions and sensor data help the AI to predict which movements are needed.
Through its website, opensourceleg.com, those interested can find details on purchasing parts, and the legs' free coding. While the coding and details are free online, the parts and computers used in the U-M’s leg cost up to $28,500 when combined into a full leg.
In early June, Rouse and Hargrove unveiled the artificially intelligent leg at the inaugural Amazon re:MARS conference in Las Vegas. The conference specifically focused on artificial intelligence, robotics, machine learning, automation and space.
All About Ann Arbor is powered by ClickOnDetroit/WDIV.