AUTOMATED SYSTEM FOR MONITORING SURFACE QUALITY PARAMETERS OF A PART BASED ON ROBOTIC SYSTEMS

Abstract

This study presents the results of research on the automation of surface quality control of parts in modern manufacturing environments using robotic systems and computer vision technologies. It is shown that traditional quality control methods have limitations related to human factors, low accuracy, and slow data processing. This highlights the need to develop an integrated automated control system capable of self-learning and integration into digital manufacturing environments (MES/ERP). A structural model of a quality control system has been developed based on a UR5 manipulator with six degrees of freedom, a machine vision system using a CMOS camera, and artificial intelligence algorithms. Mathematical relationships have been established between image parameters and quality indicators, allowing for the automatic detection of surface defects with up to 95% accuracy. It is demonstrated that the use of LBP and CNN algorithms reduces the number of false positives to 3% and shortens the inspection time per product by a factor of 1.8. The developed system was tested under laboratory conditions and demonstrated stable performance under varying lighting conditions and object positioning. Consequently, the implementation of robotic quality control systems enhances the efficiency of production process management, ensures objective assessment, and reduces post-control operation costs.

The results obtained can be used to modernize technological processes in the instrument engineering and mechanical engineering industries, as well as in the development of intelligent manufacturing systems within the Industry 4.0 framework.