Semi-Autonomous Robotic Arm Reaching With Hybrid Gaze-Brain Machine Interface.

Area of Science:

• Robotics
• Human-Computer Interaction
• Neuroscience

Background:

• Non-muscular human-robot interfaces (HRI) and shared control are promising for assistive robotic arms.
• Current discrete-valued commands in non-muscular HRIs lead to non-intuitive control and reduced user agency.
• Switching between user and robot control in shared strategies causes frustration.

Purpose of the Study:

• To propose an intuitive hybrid HRI combining Brain-Machine Interface (BMI) and gaze-tracking.
• To develop a shared control paradigm with dynamic regulation of user input and robot autonomy.
• To validate the hybrid gaze-BMI and shared control for robotic arm tasks.

Main Methods:

• Developed a hybrid gaze-BMI for simultaneous control of movement speed and direction.
• Implemented a shared control paradigm dynamically combining user commands and robot assistance.
• Validated the system with healthy subjects performing a robotic arm reaching task.

Main Results:

• Users successfully controlled the robotic arm to reach targets and avoid obstacles.
• The shared control paradigm balanced user control with necessary robotic assistance.
• The system generated smooth, collision-free trajectories, reducing task completion time and effort.

Conclusions:

• The hybrid gaze-BMI offers intuitive and effective control for assistive robotic arms.
• The proposed shared control paradigm enhances user experience and task efficiency.
• This semi-autonomous system significantly improves robotic arm control for users with severe impairments.