Gaze-Foot Hybrid Upper Limb Assistive Robotic System: An Innovative Approach to Aid Individuals with Limb Impairments.

Abstract

Individuals with hemiparesis rely on assistive robotic systems to complete grasping tasks. Among the various modalities used for human-robot interaction, the eye-gaze movement has emerged as a promising approach due to its simplicity, intuitiveness, and effectiveness. However, gaze-based assistive robots still face limitations in ensuring effective human-robot interaction and grasping. Our objective is to design an assistive system capable of overcoming these limitations and enabling natural, intuitive control of assistive robots. In this study, we have designed a gaze-foot hybrid upper limb assistive robotic system that integrates gaze-based grasping intention recognition, object detection and localization, and foot pose-based grasping orientation control. This system is capable of recognizing user intentions through gaze and foot signals and executing grasping tasks to assist participants with hemiparesis in performing daily activities. Two typical daily tasks, assistive grasping and cooperative tasks, were identified to establish the workflow of the assistive robot system. Eight healthy individuals and two Individuals with hemiparesis were recruited and the results reveal high success rates between subjects in assistive grasping (95.83%) and cooperative tasks (94.00%), with average task completion times of 30.41 seconds and 43.10 seconds, respectively.The comparative experiments quantitatively demonstrate the advantages of the designed gaze-foot hybrid system. Moreover, assessment through the System Usability Scale indicates the system's user-friendliness and ease of learning. The designed assistive system can help users effectively complete grasping tasks and is intuitive and user-friendly. The designed system offers a novel approach to assistive robotic control and can be applied to other robotic interaction control systems.