Developing a Digital Twin System to Predict and Support Stable Walking for Exoskeleton Users

What this Trial Is About

Researchers are studying the use of digital twin technology to improve exoskeleton control for healthy individuals. This project aims to create a virtual model that runs parallel to the users movements, calculating the torque needed at each joint to maintain stable walking without falling. This approach addresses current limitations in exoskeleton applications that focus mainly on motor torque compensation without considering overall motion stability and fall prevention. Participants will wear sensors such as inertial measurement units IMU and electromyography EMG devices on their lower limbs to non-invasively capture body signals while performing various common actions or transitions between actions. The system will analyze joint angles, angular velocity, and acceleration to predict human intent and adjust exoskeleton support accordingly. This includes warning and reducing auxiliary forces when movements fall outside a stable gait, and recognizing changes in movement modes like stopping or sitting. During the study, participants joint angles, muscle activity, and overall movement intentions will be monitored over three years. The data collected will help refine the machine learning models predicting human intent. Participation involves wearing sensors and performing movements while researchers collect and analyze body signal data. Safety and stability will be assessed continuously, with no invasive procedures involved. The study is led by National Taiwan University Hospital and allows healthy volunteers aged 18 to 80 years.

A Digital Twin for Exoskeleton Pilot