Attribution-Gated Joint Estimation of Intention and Capability for Assist-as-Needed Upper-Limb Robots
The research developed an attribution-gated controller to improve Assist-as-Needed (AAN) control for upper-limb robots by distinguishing between changes in intended motion and changes in voluntary torque capability. By simultaneously estimating both quantities and using a gating mechanism, the system achieved significantly lower tracking errors across various load and height conditions compared to existing methods.