Implementation of YOLOv8 Algorithm on Raspberry Pi for Automated PCB Quality Inspection

This study develops an automated quality inspection system based on machine learning using the You Only Look Once version 8 (YOLOv8) object detection algorithm and deploys it on a Raspberry Pi for the quality inspection of Printed Circuit Boards (PCBs). PCBs are crucial components in modern electronics, such as smartphones, computers, and other devices. This research reports on testing the YOLOv8 algorithm for PCB quality inspection. Experiments applied YOLOv8 to a PCB testing dataset on a Raspberry Pi, demonstrating the system’s efficiency and reliability. The test results show the system's ability to detect defects such as Pad Misalignment, Open Circuit, Solder Bridging, and Misplacement, with varying level of accuracy. Accuracy for Good Assembly (GA), Solder Bridging (SB), Open Circuit (OC), and Misplacement (MP) classes ranged from 70% to 95%, 73% to 84%, 67% to 72%, and 67% to 98%, respectively. These results indicate that this system effectively enhances the efficiency and accuracy of PCB quality inspection when implemented on a Raspberry Pi, potentially improving the overall quality of electronic products.