Leveraging welding spot Quality Control using computer vision system

Spot welds are integral to automotive body construction, influencing vehicle performance and durability. Spot welding ensures structural integrity by creating strong bonds between metal sheets, crucial for maintaining vehicle safety and performance. It is highly compatible with automation, allowing for streamlined production processes and increased efficiency in automotive assembly lines. The number and distribution of spot welds directly impact the vehicle's ability to withstand various loads and stresses, including impacts, vibrations, and torsion. Manufacturers adhere to strict quality control standards to ensure the integrity of spot welds in automotive production. Monitoring spot weld count and weld quality during manufacturing processes through advanced inspection techniques such as Image processing by YOLO V8 helps identify the number of spots and quality that could compromise safety Automating quality control processes is paramount, and machine vision offers a promising solution. Leveraging the YOLO v8 model, this research proposes an efficient technique for automatic detection and counting of spot welds on automotive sheets. Through analysis of a comprehensive dataset of annotated images, our approach demonstrates superior accuracy and efficiency in tracking and quantifying spot welds. Quantitative evaluation validates the effectiveness of this vision-based inspection method, highlighting its potential for enhancing car body welding quality control processes. keywords- Machine learning, Weld spot quality, YOLO V8