A Wearable Human–Machine Interactive Instrument for Controlling a Wheelchair Robotic Arm System

The limitations of traditional manual human–machine interactive instruments (HMIs) prevent individuals with severe motor disabilities from effectively controlling wheelchair robotic arm systems, thereby impacting their independence and quality of life. Addressing this issue, this study developed a wearable hybrid HMI to meet the practical needs of individuals with severe motor disabilities. By recruiting ten healthy participants to complete three experiments—a blink detection test, wheelchair control test, and wheelchair robotic arm system test—the effectiveness of the proposed HMI was validated. In the first experiment, the average accuracy of blink detection in the system reached 97.29%, with an average response time of 1.47 s, and the system generated 0.07 errors per minute in idle state. The second experiment demonstrated an average response time of 1.03 s for wheelchair turning and 1.48 s for wheelchair stopping. In the hybrid control condition, the wheelchair navigated obstacles along the specified route in an average time of 1.14 min, and participants executed a minimum of 25.20 commands. In the third experiment, all participants successfully completed the mobile self-drinking task by controlling the wheelchair robotic arm system, with an average workload reported on the NASA task load index (NASA-TLX) scale of 30.2. The study revealed that the proposed HMI offers a promising solution for nonmanual control in complex rehabilitation assistive systems. It has the potential to assist a wider range of individuals with motor disabilities, improving their daily life experiences.