Human CNS-inspired AI Framework for Personalized Wearable Robot Assistance

Human CNS-inspired AI Framework for Personalized Wearable Robot Assistance

Wearable robots have great potential, but standard assistance often fails to meet the unique gait pattern of every individual. Our Neuromorphic Locomotion Control (NLC) framework addresses this by mimicking the hierarchical and distributed structure of the human nervous system. The NLC framework consists of five specialized neural modules: a body schema generator, locomotion planner, speed predictor, rhythm generator, and torque pattern generator.

By applying this framework to the Angel SUIT H10 (a commercial hip-assistive robot by Angel Robotics), we have created a more responsive way to assist walking. The core innovation of NLC lies in its efficiency; it selectively personalizes only the specific neural modules that require tuning for a new user. This allows the system to find the optimal assistance pattern much faster than traditional end-to-end AI methods.

At our AI for Good live demonstration, visitors can experience this technology firsthand. Participants can wear the H10 robot and witness real-time, online personalization. As you walk, the system analyzes your unique gait and instantly adjusts its supportive force to fit you. This showcase highlights how AI can evolve into a responsive partner, enhancing mobility and independence for everyone.

Share this speaker
Share This Speaker
  • Organization
    KAIST Exoskeleton Laboratory
  • Profession
    AI-driven exoskeleton

Are you sure you want to remove this speaker?