Högskolan i Skövde

In pursuing Zero Defect Manufacturing (ZDM), this study explores an innovative quality inspection method that combines computer vision with mixed reality for real-time error detection. Aligned with Industry 5.0, the proposed solution not only enhances operational efficiency in manufacturing but also promotes worker well-being by simplifying and automating the inspection and error detection process—a task that is usually demanding both mentally and physically for human operators. This approach enables operators to view and interact with accurate inspection data directly overlaid on real-world objects, improving their ability to spot and correct defects immediately. A case study in electrical terminal assembly demonstrates how deep learning-powered object detection integrated with MR improves inspection accuracy and efficiency. This work represents a significant step forward in automated quality control, supporting ZDM’s goals for sustainable, high-precision, and human-centered manufacturing.