Florida A&M and FSU researchers build AI-powered robotic unicycles for faster stroke rehab

FAMU-FSU researchers are developing an AI-powered robotic unicycle system to study how humans learn complex motor skills, with the ultimate goal of creating smarter physical therapy tools for stroke survivors and mobility-impaired patients. The $799,000 National Science Foundation project represents a novel approach to rehabilitation robotics by focusing on skill acquisition rather than simply assisting with movements people already know how to perform.

How it works: The research unfolds in three distinct phases, each building toward more sophisticated human-robot interaction.

• Researchers first study how subjects learn to unicycle unassisted, using motion-capture suits to gather data from joint positions and mathematically model movement during the learning process.
• They then use reinforcement learning to train a simulated “robot coach” to guide a simulated human learning to unicycle, with the coach rewarded for accelerating learner skill progress.
• Finally, the team will build a unicycle capable of providing robotic assistance, using learning strategies developed in the simulation phase to teach novice riders and test whether robotic coaches accelerate the learning process.

Why this matters: Traditional walking rehabilitation tools like robotic exoskeletons are typically pre-programmed with movements or require basic patient input, but this research moves toward models that actively interact with patients to accelerate recovery.

• “A lot of assistive robotics research focuses on helping people with movements they already know how to perform,” said Assistant Professor Taylor Higgins, who leads the project. “Our study will focus on how people acquire new motor skills.”
• The approach emphasizes gradual independence, with robotic assistance decreasing as learners progress, helping them quickly acquire and cement new skills.

The technical advantage: Researchers chose unicycles over walking studies because they represent a simpler mathematical model that allows focus on human-robot interaction.

• Walking involves repeatedly making and breaking contact with the ground in what engineers call a hybrid dynamic system, while unicycles remain in contact with the ground as a continuous system that’s easier to model mathematically.
• “It’s really hard to get a robot to walk on two legs at all,” Higgins explained. “The problem gets that much harder when you have to guide a human through the learning process at the same time.”

What they’re saying: The research addresses a critical gap in current rehabilitation technology.

• “We don’t have algorithms that tell us exactly what the robot should do to help you to learn. That’s the missing piece,” Higgins noted.
• “If robots can help healthy people learn a task faster, they can then be adapted to help people undergoing rehabilitation to regain lost skills faster. Faster learning means faster rehab.”

Who else is involved: The project brings together expertise from multiple institutions and disciplines.

• Higgins collaborates with Brady DeCouto, an assistant professor in FSU’s Anne Spencer Daves College of Education, Health, and Human Sciences, who studies human learning processes independently of robotic involvement.
• The team also works with Shreyas Kousik, an assistant professor at Georgia Tech specializing in machine learning for robotics and reinforcement learning in AI systems.

The big picture: This research is supported by the National Science Foundation’s Mind, Machine and Motor Nexus program, which funds fundamental research enabling intelligent engineered systems and humans to interact safely and productively in complex, changing situations.