Hands-free wheelchair prototype achieves major milestone

In 2020, collaborators from across the University of Illinois campus received a $1.5 million grant from the National Science Foundation (NSF) to develop a robotic wheelchair that provides a novel mode of mobility to individuals with physical disabilities. Now two years in, the project has a working prototype and a patent pending.

The design for wheelchairs has not changed much since the first patent in the 1800s. “Manual wheelchair users, due to this traditional design, have a lot of issues,” said Elizabeth T. Hsiao-Wecksler, PI and Professor in the Department of Mechanical Science and Engineering. “It’s difficult to fit in tight spaces, [the chair] requires both hands to propel, and 70% of manual wheelchair users will experience upper extremity overuse injuries that may lead to inability to use a manual wheelchair. Powered wheelchairs address some of these issues, but they’re generally really heavy, very large, expensive, and can be difficult to maneuver; most people who are able to would use a manual wheelchair.”

The project’s multidisciplinary team includes a multitude of faculty and staff from The Grainger College of Engineering, College of Fine and Applied Arts, Carle Illinois College of Medicine, and the College of Applied Health Sciences: Hsiao-Wecksler, who is also the Interim Director of the Health Care Engineering Systems Center; Industrial Design Professor and co-PI Deana McDonagh from the School of Art+Design and Carle Illinois College of Medicine; Clinical Associate Professor and co-PI Bob Norris from the Department of Industrial and Enterprise Systems Engineering; Assistant Professor João Ramos in MechSE; Adam Bleakney, Head Coach of Men’s and Women’s Wheelchair Track and Field in the Division of Disability Research and Education Services (DRES); Dr. Jeannette Elliott, Head Physical Therapist at DRES; Dr. Pat Malik, retired Director of DRES; and numerous graduate and undergraduate students from across campus.

“Not only does this project represent contributions from diverse colleges, but it is truly interdisciplinary in the expertise the collaborators bring to the table,” McDonagh said.

“The impetus for pulling this team together was to leverage the expertise on campus to increase access to spaces and experiences that daily wheelchair users like me are often marginalized from. Standard ultra-light, manual wheelchairs are no doubt outstanding in many respects and provide an increased level of independence during daily life. At the same time, the act of pushing a manual wheelchair prevents the hands from engaging in other more meaningful activities while moving – such as holding hands with a loved one on a walk, carrying a cup of coffee, holding a toddler while moving around the house. Each of these is possible with PURE,” Bleakney said.

This impressive team was able to create PURE, Personalized Unique Rolling Experience, a hands-free wheelchair that operates similarly to a Segway where the rider leans in a desired direction. The unique aspect of PURE is that it rolls on a ball or “spherical” wheel. It is based on the concept of a dynamically stable ball-based robot (ballbot) and uses an omniwheel system to drive and control the spherical wheel. PURE automatically transitions between three driving behaviors. Steer and Spin are similar to a typical wheelchair, in which the user can steer forward, backward or spin in place. Slide is unique and allows the user to move laterally, like an office chair. To accommodate for limited torso range of motion of some users, PURE uses sensors to estimate leaning and twisting motions and amplifies these signals to control the ballbot’s direction and speed.

“The development of PURE has been guided by our immutables – that it be lightweight and maintain a small footprint. We want to ensure that the current independence of manual wheelchair users would in no way be limited by PURE. If we were to develop a hands-free device that was so heavy that it prevented users from easily transferring it into and out of their vehicle, or if it was so large that it wouldn’t maneuver around typical living spaces, we would have missed the mark. Any device that compromises current levels of independence just won’t be used during daily life,” Bleakney said.

The project started in 2018 with the Toyota Mobility Unlimited Challenge, where the team created a first-generation prototype and flew to London to present their idea. While they did not win the competition, they were undeterred, continuing with the project and ultimately winning the NSF grant in 2020. Their second-generation prototype can support roughly 60 kg (130 lbs), move at up to 2.3 m/s, and brake in 2 seconds from 1.4 m/s. It is only about as wide as the user’s hips and the height of a standard table chair, a significantly smaller footprint than either powered or manual wheelchairs. 

While developing their prototype, the engineering team designed a virtual reality simulation to test controls and train users—also creating a virtual obstacle course to help users practice. The sensitivity of the ballbot’s controls can also be modified for first-time users to help them adjust to the system.

This past summer, ten able-bodied people and ten manual wheelchair users tested the second-generation prototype in the Center for Autonomy’s high-bay testing area, which provided enough space to set up various courses to see how well participants could control PURE, including turning around in a simulated restroom stall and navigating tight corners. Overall, Hsiao-Wecksler said the testing was a success, and the team recently brought the prototype to the opening of the (dis)Ability Design Studio at the Beckman Institute for Advanced Science and Technology.

“[Assistant Dean for Advancement in the College of Applied Health Sciences] Jean Driscoll hopped into the prototype and started driving it right away. We didn’t even have it fully calibrated to her—it was still calibrated to someone who had less range of motion,” said Hsiao-Wecksler, “... but it behaved exactly as we were intending, in that you could just sit in it, feel the organic motion, and drive it with your body.”

Despite reaching an impressive milestone, the team is looking forward to improving the device. Among other things, they are now working on advanced driver assistance using a machine learning vision system for path-keeping, collision avoidance, and autonomous controlled stop. Future PURE prototypes will be able to interpret both user input and obstacle detection to guide the user safely.