Skip to main content Start main content

Visually Guided Robotic Welding System for On-site Application

High-quality welded joints are essential for safety and integrity of steel structures, and adoption of robotic welding can enhance quality and productivity of steel construction. However, welding on construction site is generally location-specific and non-repetitive, rendering an adoption of traditional robotic welding techniques not suitable.

This project integrates existing technologies of artificial intelligence, robotic kinematics with various operating systems, and gas shielded arc welding to establish a visually guided robotic welding system for on-site application. A 9-dimensional Robotic Welding System is developed through advanced programming which has successfully integrated welding operations of two or more robots, depth measurements with RGBD cameras, and location sensing using a high precision laser scanner. Through effective utilisation of computer vision and automation technology, this system identifies and follows complex welding paths autonomously while the welding torch is always maintained in a position perpendicular to the welding path without additional manual operation.

This system has been adopted in fabrication of welded steel sections in a number of research projects on high performance steel to achieve high quality and precision welding. It is readily applicable to on-site welding to achieve precision welding of steel members with consistency and repeatability. In addition, the time-consuming process of aligning structural members with the projected welding tracks will be greatly streamlined. Concurrent welding may be carried out with two or more robots, whenever needed, to minimise thermal expansion and distortion in heavily welded sections.


Your browser is not the latest version. If you continue to browse our website, Some pages may not function properly.

You are recommended to upgrade to a newer version or switch to a different browser. A list of the web browsers that we support can be found here