Developing an AI-enabled framework to support multiple exosuit activities
1
For people who need assistance with walking, running, and heavy lifting, hip exosuits have emerged as a way to reduce metabolic cost and muscle fatigue. These wearable robotic devices use fabric-based interfaces, motors, and cable systems to help with hip flexion and extension.
But while many mobility challenges occur during transitions, such as sit-to-stand, current hip exosuits are optimized for walking.
To address this issue, MIE Associate Professor Myunghee Kim is developing an intelligent, activity-specific personalization framework, which is a structured, strategic, and technical approach used to deliver tailored experiences to improve engagement and efficiency, for the hip exosuits.
“What’s new here is that we extend personalization beyond steady walking or squatting to reflect the multi-activity nature of daily mobility,” said Kim, director of the Rehabilitation Robotics Laboratory at UIC. “Depending on an activity – walking, standing, and sit-to-stand – we will develop a smart, AI-enabled personalization framework that switches to an activity-specific assistance strategy and continuously refines assistance within each activity, rather than relying on a single fixed setting across all tasks.”
The project, titled “Multi-Activity Personalized Assistance for a Hip Exosuit,” is sponsored by the Chung-Ang University Industry-Academic Cooperation Foundation. The research builds directly on Kim’s prior work on personalized assistance for wearable robots/exosuits, in which she tuned assistance parameters based on measurable user responses, including biomechanical measures, effort-related metrics, and comfort proxies.
The framework will benefit users of hip exosuits, which include older adults and individuals with reduced mobility due to neurological conditions, such as a stroke, or general weakness/deconditioning. Clinicians and therapists are also important users, since a smart personalization approach can reduce manual trial-and-error during setup and make assistance more consistent across activities.
“This will support reliable assistance during home and community mobility and can also streamline clinical workflows by making tuning faster and more repeatable across tasks,” Kim said. “By developing an AI-enabled, multi-activity personalization approach, we aim to improve effectiveness, comfort, and overall usability in daily life.”