Integrating Surface Electromyography with Assistive Robots for Elderly Care: A Novel Human-Robot Interface for Grasping Tasks

This paper presents a novel human-robot interface (HRI) for elderly care, integrating Surface Electromyography (sEMG) signals with an assistive robot to enhance grasping tasks. We introduce innovations in sEMG signal processing and classification methods optimized for older adults, along with efficient mapping of sEMG signals into precise control commands for assistive robots. Using a custom-designed armband device, we collected a unique dataset of four essential grasping gestures from older adults. Our machine learning-based classification algorithm demonstrates improved accuracy in gesture recognition for elderly users compared to existing methods. Experiments conducted with elderly participants showed shorter training times and higher success rates for grasping-related tasks compared to conventional interfaces. This research not only enhances the flexibility of robotic systems for elderly care but also provides a foundation for personalized assistance in daily activities. Our findings suggest that this sEMG-based robotic control system could improve the independence and quality of life for older adults. The paper concludes by discussing the advantages of our approach in the context of elderly care and outlines future directions for sEMG-based assistive robotics in gerontechnology.