Evaluation of Robot Degradation on Human-Robot Collaborative Performance in Manufacturing.

Area of Science:

• Robotics
• Manufacturing Engineering
• Human-Computer Interaction

Background:

• Human-robot collaborative systems offer adaptability and consistency for smart manufacturing.
• Hesitancy exists due to a lack of metrics on robot degradation's impact on teaming performance.

Purpose of the Study:

• Define teaming performance metrics considering robot degradation.
• Evaluate these metrics in a human-robot collaborative inverse peg-in-hole task.

Main Methods:

• Developed teaming performance metrics sensitive to robot degradation.
• Applied metrics to an inverse peg-in-hole case study with degraded joint encoders and current sensors.
• Compared pure insertion vs. spatial scanning, including manual intervention for failures.

Main Results:

• Pure insertion required manual intervention at 0.04° degradation, while scanning needed it at 0.12°.
• Insertion with scanning demonstrated greater robustness to robot degradation.
• Spatial scanning resulted in a longer insertion time (9.48 s) compared to pure insertion (3.19 s).

Conclusions:

• The defined metrics provide valuable insights into robot degradation's effect on collaborative systems.
• Insertion with scanning is more resilient to robot degradation, albeit slower.
• This research aids in adopting human-robot collaborative systems in manufacturing.