Home▸Robotics & Kinematics▸Assist-as-Needed Rehab Exoskeleton

Assist-as-Needed Rehab Exoskeleton

Interactive rehabilitation-exoskeleton simulator: compare a fixed-assistance robot that always applies the same torque against an assist-as-needed (AAN) controller that scales assistance down as the patient's own effort rises, watching the two curves cross over a training session.

Robotics & Kinematics3DModerate60 FPS

A rehabilitation exoskeleton helps a recovering patient relearn a reaching motion, but how it assists shapes how much the patient actually recovers. This simulator drives a single animated exoskeleton arm through a repeated reaching-training session under two controller strategies: fixed assistance, which always applies the same torque no matter how the patient is doing, and assist-as-needed, which continuously measures the patient's own contribution and hands back only the torque still missing. Run a full session in each mode and watch the live effort-vs-assistance chart and the recovery-progress score reveal why letting the patient struggle a little, on purpose, trains the neuromuscular system faster than doing the work for them.

⚙ Under the hood

Run a repeated reaching-training session on an animated exoskeleton arm and compare two control strategies: fixed assistance, which always applies the same torque no matter how the patient performs, versus assist-as-needed, which measures the patient's own effort and hands back only the torque still missing. A live chart tracks patient effort and robot assistance crossing over as the session progresses, ending in a recovery-progress score comparing self-sufficiency gained under each strategy.

exoskeletonrehabilitation roboticsassist-as-neededadaptive controlmotor learningThree.js

3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install