A New Human-Likeness and Comfort Index for Robot Movements Along Prescribed Paths

Area of Science:

• Robotics
• Human-Robot Interaction (HRI)
• Biomechanics

Background:

• Robot acceptance is crucial as human-robot interaction (HRI) expands.
• Visual human-likeness alone does not guarantee acceptance; perceived comfort and ergonomics are key.
• Comfort is directly tied to the perceived quality of robot movement during physical interaction.

Purpose of the Study:

• To investigate the relationship between robot movement similarity to human movement and perceived comfort.
• To develop a quantitative index for evaluating the human-likeness of robot trajectories.
• To enable a priori assessment of trajectory generation algorithms for human-like movements.

Main Methods:

• Kinematic characterization of human movement, focusing on time laws.
• Application of the lognormality principle to model human motion.
• Definition and utilization of a "human-likeness index" for trajectory analysis.
• Experimental validation with 68 subjects rating comfort during physical robot interaction.

Main Results:

• A consistent trend was observed between subjects' comfort preferences and the proposed human-likeness index.
• The human-likeness index effectively predicts perceived comfort in physical HRI scenarios.
• Trajectory generation algorithms can be evaluated using this index before physical execution.

Conclusions:

• Robot movement human-likeness is a significant factor influencing acceptance and comfort in HRI.
• The defined human-likeness index provides a valuable tool for designing more acceptable and comfortable robots.
• This approach facilitates the development of robots that move in ways humans perceive as comfortable and natural.