Neuromuscular Interfacing for Advancing Kinesthetic and Teleoperated Programming by Demonstration of Collaborative...

Area of Science:

• Robotics
• Human-Robot Interaction
• Neuroscience

Background:

• Programming by Demonstration (PbD) for collaborative robots requires intuitive methods for adding complex programming inputs.
• Existing PbD techniques often struggle to integrate additional programming dimensions without disrupting the demonstration process.

Purpose of the Study:

• To develop and evaluate a wearable human-robot interface for augmenting Programming by Demonstration (PbD).
• To enable intuitive programming of robot compliance and gripper functions simultaneously using neuromuscular signals.

Main Methods:

• Utilized a wearable interface measuring forearm muscle co-contraction levels via surface Electromyography (sEMG).
• Implemented hand stiffening modulation for additional programming inputs, complemented by vibrotactile feedback.
• Tested the framework on a collaborative robot performing an industrial wiring task.

Main Results:

• Achieved a 100% success rate in an industrial wiring task across subjects.
• Vibrotactile feedback reduced programming errors by an average of 33%.
• Subjects demonstrated accurate modulation of muscle co-contraction levels.

Conclusions:

• The proposed neuromuscular interface provides an intuitive method for enhancing PbD in collaborative robotics.
• The framework effectively integrates structured programming logic into teleoperation and kinesthetic teaching.
• This approach offers a significant advancement in human-robot interaction for complex robotic tasks.