Related Experiment Video
Updated: Aug 14, 2026
05:47
Simulation of a Scaled Assembly Process with Collaboration of a Robotic Arm and Monitoring through a Vision System for Quality Control
Published on: August 29, 2025
Stereo Vision-Based Human-Robot Interaction for Weld Seam Detection in Robotic Welding
Pushkar Kadam1, Gu Fang1, Farshid Amirabdollahian2
1Centre for Advanced Manufacturing Technology, School of Engineering, Western Sydney University, Locked Bag 1797, Penrith, NSW 2751, Australia.
Sensors (Basel, Switzerland)
|August 13, 2026
Summary
This study introduces an intuitive robotic welding method allowing experts to guide robots using hand gestures. This human-robot interaction (HRI) approach simplifies programming for small-batch manufacturing, improving efficiency.
Area of Science:
- Robotics
- Manufacturing Engineering
- Computer Vision
Background:
- Robotic welding enhances manufacturing efficiency but is often impractical for small batches due to programming complexity.
- Non-experts require intuitive methods to program robots for tasks like welding.
Purpose of the Study:
- To develop a human-robot interaction (HRI)-based weld seam detection system for intuitive robotic welding instruction.
- To enable non-robotics experts to demonstrate welding paths via hand gestures without programming.
Main Methods:
- Implemented a vision-based hand detection and tracking system for path demonstration.
- Developed algorithms to isolate weld seam lines and identify welding paths.
- Projected detected seam paths from image space to the robot's coordinate system.
Main Results:
- Achieved accurate weld seam detection and path identification.
- Demonstrated successful application on a UR10e robot.
- Evaluated accuracy to be 1 pixel in the image plane, equivalent to 1 mm in physical space.
Conclusions:
- The HRI-based weld seam detection system offers an intuitive and efficient solution for robotic welding in small-batch manufacturing.
- This approach significantly reduces the barrier to entry for using robotic welding by eliminating the need for programming.
- The system's accuracy supports practical implementation in real-world manufacturing scenarios.
