This video demonstrates fast walking with real-time classification between four activity modes (sit-stand, walk, stair ascent, stair descent), each using a controller that automatically adjusts to continuous task variations (e.g., walking speed, ground incline, step height, chair height). The classifier uses data from a thigh IMU and ultrasonic range sensor to predict activity transitions, enabling the above-knee amputee pilot to seamlessly transition between activities during continuous locomotion.
This work was supported by the National Institutes of Health.