“I can put my foot down and my ankle can roll,” said test patient Derrick Stone. “If I inadvertently step on a rock, my foot rolls over. It’s about the closest thing you can get to having your foot back.”
Photograph courtesy of the Nishikawa Biomechanics Laboratory at Northern Arizona University.
A research team at Northern Arizona University’s (NAU’s) Center for Bioengineering Innovation, under the direction of Kiisa Nishikawa, PhD, said they may have formulated a solution to more seamlessly integrate natural movement with the mechanical support of a prosthetic ankle. An algorithm developed by the center forms the basis of a $225,000 National Science Foundation (NSF) grant awarded June 26. The funding will allow the center and BiOM, Bedford, Massachusetts, to proceed with the commercial refinement of bionic ankles.
“We believe this algorithm is a major breakthrough towards creating a foot/ankle prosthetic for all-terrain walking by the users,” said Nishikawa, Regent’s Professor of Biological Sciences. “We were able to show that our algorithm works not only during level walking at variable speed but also going up and down stairs and walking backward.” The next test, Nishikawa explained, is to demonstrate that it works just as well on rocks or artificial turf.
The challenges of stairs and a variety of surfaces reveal the complications of designing a bionic body part, such as an ankle, that functions even remotely as well as the real thing. Engineers can divide the action of walking into a certain number of stages, then write equations for optimizing each stage and design sensors to detect them, Nishikawa said. At each stage, algorithms are applied to generate the appropriate torque. The result is a highly adaptable machine.
“But our muscles don’t do that,” Nishikawa said. “We sometimes experience unexpected perturbations, and our muscles can respond instantly to changes our environment applies to our bodies. So if we can capture that behavior in a single algorithm, then we don’t need to know the what, when, and how of changes in terrain.”
In robotics, such a do-all algorithm is known as robust control, Nishikawa said. “With minimal information, it can respond in the appropriate way in all terrains and in a variety of circumstances.”
The researchers made their discovery several months ago. After months of testing and modifying their algorithm with a BiOM ankle, a volunteer test subject tried walking up a flight of stairs. It worked the first time. “He could ascend the stairs much faster using our algorithm,” Nishikawa said.
