APPLICATION OF A 3-AXIS ROBOT ARM COMBINED WITH A CAMERA SUPPORTS AUTOMATIC WELD DATA COLLECTION

Abstract

The aim of this study is to design a three-axis translational robotic arm model combined with a camera to automate the data collection of welding joints. The robotic arm model includes three degrees of freedom designed with an inverted structure and can translate along the X, Y, and Z axes. The end of the robotic arm is equipped with a servo motor combined with an industrial camera. The camera can rotate 180 degrees thanks to the servo motor. With this mechanism, the robotic arm model can adjust its translation and rotation to collect image data from various angles. Experimental results show that the system has proven its ability to determine coordinates with 99.67% accuracy and the angle of the welding joint, ensuring consistency in the collected dataset. Furthermore, in addition to the ability to collect highly consistent image data regarding viewing angles and positions, the model also ensures randomness such as lighting conditions during sampling. These initial research results not only open up the possibility of inspecting welding quality but also integrate machine learning technology into industrial production processes that are conducted quickly and accurately.