A PDA-Based Human-Robot Interaction for Disabled Persons using Electromyography

The paper discusses a novel human-robot communication system for people with disability using electromyography (EMG) signals via a personal digital assistant (PDA). This paper is a continual research from our previous work with the objective of extending our approach to portable device with adaptive user interface capability. It presents an innovative architecture of human-robot interaction where the disabled persons could command the robot through an EMG-based Morse code-inspired communication paradigm on PDAs. The disabled persons may issue high-level commands to the robot via their muscle's EMG signal. The system will decompose these EMG commands into low-level subtasks. Then, the robot dynamically constructs an action selection network to form the action plan. The architecture is developed on a PDA that offers small and lightweight mobile interaction devices. Utilizing the PDA with EMG-based modified Morse code could provide flexibility, mobility, and durability for handicapped people to control the robot-aid system. Additionally, the system is designed as an agent-based user interface (UI) for mixed-initiative interaction for the user. The proposed paradigm has been implemented to develop a prototypical system, and the experiments also illustrate the conventionality, and the robustness of such a human-robot framework