Effective Multi-Mode Grasping Assistance Control of a Soft Hand Exoskeleton Using Force Myography

Area of Science:

• Robotics and Human-Computer Interaction
• Biomedical Engineering
• Wearable Technology

Background:

• Accurate human intention detection is crucial for effective human-robot collaboration.
• Existing methods for hand motion detection often lack precision or require invasive sensors.
• Controlling assistive robotic devices necessitates reliable real-time feedback on user intent.

Purpose of the Study:

• To develop and validate a novel system for detecting hand motion intentions (rest, open, close, grasp) and estimating grasping force.
• To utilize force myography (FMG) for non-invasive measurement of muscle activity related to hand movements.
• To integrate the developed system for controlling a soft hand exoskeleton (SEM Glove).

Main Methods:

• Two sensor bands with force sensing resistor (FSR) sensors were developed and placed on the arm.
• Normal forces from muscle contraction/relaxation were measured using the FSR sensors.
• A threshold-based classification method was employed to process sensor data and identify hand motion states.

Main Results:

• The developed FMG-based method accurately detected various hand motion intentions in human subjects.
• The system demonstrated reliable grasping force estimation.
• The system successfully controlled a soft hand exoskeleton (SEM Glove) for object-grasping tasks.

Conclusions:

• The proposed force myography approach offers a viable and accurate solution for human intention detection in the context of controlling assistive robotic devices.
• The system's ability to detect hand motion and estimate force enables personalized and task-specific robotic assistance.
• This technology has the potential to significantly enhance the capabilities of soft hand exoskeletons and other human-interactive robotic systems.