Physiological and subjective evaluation of a human-robot object hand-over task

Area of Science:

• Robotics
• Human-Robot Interaction
• Ergonomics

Background:

• Safe and compliant robot hardware is insufficient for effective human-robot task sharing.
• Ergonomic robot motions are essential for seamless human-robot object transfer.
• Existing behavioral measures are inadequate for evaluating subtle differences in robot gestures during handover tasks.

Purpose of the Study:

• To develop and validate a Human Aware Manipulation Planner (HAMP) for human-robot object transfer.
• To define precise subjective metrics for assessing robot motion legibility, safety, and physical comfort.
• To obtain quantitative physiological and ocular data to evaluate human-robot interaction during object handover.

Main Methods:

• Developed the Human Aware Manipulation Planner (HAMP) considering legibility, safety, and comfort.
• Designed an experimental setup with a mobile manipulator robot (Jido) and human participants.
• Utilized galvanic skin conductance response, deltoid muscle activity, and ocular activity as physiological measures.
• Collected subjective ratings on a three-dimensional scale after each robot motion.

Main Results:

• Subjective data favored the reference motion, indicating higher perceived legibility, safety, and comfort.
• Distinct physiological and ocular responses were observed for different robot motions.
• The developed metrics showed potential for partially discriminating between robot gestures.

Conclusions:

• The Human Aware Manipulation Planner (HAMP) effectively generates ergonomic robot motions for object transfer.
• Physiological and ocular responses provide valuable quantitative data for evaluating human-robot interaction.
• This approach enhances the safety, comfort, and overall quality of human-robot collaboration.