Thesis
REAL-TIME TOOL-PATH ADAPTATION OF A COBOT FOR IMPROVED SAFETY IN HUMAN-ROBOT INTERACTION
Washington State University
Master of Science (MS), Washington State University
01/2022
DOI:
https://doi.org/10.7273/000004463
Handle:
https://hdl.handle.net/2376/123170
Abstract
Collaborative robots (cobots) are becoming increasingly popular in industrial environments because they enable close collaboration between a human and a robot. Unlike traditional robots, cobot installations do not require a safety cage around the robot. This is because the cobot can come to a stop if it collides with a human due to additional sensors within the actuators for torque feedback. Although the cobot can stop, it is not a pleasant or safe working environment for the human to be hit by the cobot possibly multiple times a day. To improve working conditions and reduce or eliminate possible injuries, as well as improve downtime, collision avoidance is needed. This research introduces a real-time collision avoidance algorithm for a collaborative robot in a work cell with a human operator. The new algorithm uses inexpensive proximity sensors on the cobot to detect an object, to trigger a modified bug algorithm, and to stop motion of the cobot, if needed. First, a static human operator was used as an obstacle within the experiments. The operator’s location was fixed while the cobot approached a table to pick up a part. If a collision occurred or a successful pickup ensued, the work cell would reset, the operator’s position would change, and then another trial would begin. Second, the same experiment was performed, but the orientation of the operator was altered. The operator was rotated relative to the table. Third, a yaw angle was introduced which would adjust the target orientation for the cobot. This would allow the UR10’s arm to maneuver and warp around the operator after a solution was found. Finally, dynamic experiments were considered.
The modified bug algorithm could significantly reduce or eliminate collisions in many instances and increased the number of part pick-ups. The use of the algorithm alone with no yaw angle showed a decrease in collisions by 82%. With the adjusted yaw angle, the number of collisions were reduced by 100%, and the number of successful pickups were increased by 58%. The new collision avoidance algorithm is a promising improvement to the cobot’s capabilities.
Metrics
7 File views/ downloads
25 Record Views
Details
- Title
- REAL-TIME TOOL-PATH ADAPTATION OF A COBOT FOR IMPROVED SAFETY IN HUMAN-ROBOT INTERACTION
- Creators
- Cody Cichosz
- Contributors
- Hakan Gurocak (Advisor)Jong-Hoon Kim (Committee Member)Xiaolin Chen (Committee Member)
- Awarding Institution
- Washington State University
- Academic Unit
- Engineering and Computer Science (VANC), School of
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University
- Number of pages
- 57
- Identifiers
- 99900881631401842
- Language
- English
- Resource Type
- Thesis