EMG controlled mobile robot equipped with gripper mechanism for fine motor skills training in rehabilitation

• This work proposes a novel approach by integrating Electromyography (EMG) signals with a mobile robot, allowing muscle activity associated with fine motor movements to control the robot. • A key innovation of this system is a gripper mechanism, designed using 3D modelling techniques, which operates in response to specific hand gestures detected via EMG signals. • This enables the execution of interactive and task-oriented rehabilitation exercises. The hand gestures are classified using a Normal and Complementary Gaussian Mixture Network (NACGMN), that is capable of differentiating between learned and unlearned gestures. • The network utilized, enables seamless operation of the proposed system by exhibiting an accuracy of 95.81%, F1 score of 95.84%, precision of 95.93 % and recall of 95.80%. • By harnessing the capabilities of mobile robotics and EMG signals, this research aims make an impactful contribution to advancing rehabilitation technology and improving outcomes for individuals undergoing fine motor skills rehabilitation. According to the World Health Organization (WHO), approximately 15 million individuals suffer from stroke annually, with a significant portion requiring rehabilitation to restore fine motor skills necessary for daily activities. Traditional rehabilitation methods often encounter limitations in delivering personalized and engaging training experiences, resulting in suboptimal recovery outcomes. This work introduces an EMG-controlled mobile robot, where muscle activity from fine motor movements controls the robot and operates a 3D-modeled gripper mechanism. This system responds to specific hand gestures detected through EMG signals for interactive rehabilitation. The hand gestures are classified using a Normal and Complementary Gaussian Mixture Network (NACGMN), that can differentiate between learned and unlearned gestures. The network enables seamless operation of the proposed system by exhibiting an accuracy of 95.81%, F1 score of 95.84%, precision of 95.93 % and recall of 95.80%. By harnessing the capabilities of mobile robotics and EMG signals, this research aims make an impactful contribution to advancing rehabilitation technology and improving outcomes for individuals undergoing fine motor skills rehabilitation.