Evaluation of a Hybrid Inverse Kinematics Approach for Wheelchair-Mounted Assistive Robotic Arms in Performing Activities of Daily Living

Powered wheelchair users with upper limb dysfunction rely on assistive aids, such as wheelchair-mounted robots, to support their activities of daily living (ADLs). However, these robots face significant challenges in inverse kinematics (IK), particularly in handling workspace-boundary singularities and positions outside the defined workspace, leading to unpredictable behavior or loss of control. This study evaluates a novel Hybrid Inverse Kinematics (H-IK) approach designed to address these challenges, integrating the Levenberg-Marquardt (LM) method within a BFGS framework to ensure stable and continuous solutions. The H-IK approach was evaluated against traditional IK algorithms, demonstrating superior performance with faster computation times and the ability to handle complex spatial transitions and singular configurations. Experimental results with powered wheelchair users validated H-IK's effectiveness in realworld ADLs. The H-IK method promises to significantly improve the functionality and reliability of assistive robots.