This study proposes a deep learning-based image anomaly detection system to identify irregular defects in metallic DC fuses. In the manufacturing industry, product quality control is directly linked to productivity, and appearance-related defects, such as surface imperfections, impact competitiveness and reliability. However, conventional visual inspection and simple image-processing methods have limitations in inspection speed and precision, often struggling to accurately detect complex and diverse irregular defects. To address these challenges, this study employs Dinomaly, an unsupervised reconstructionbased Transformer model that can be trained using only normal images. The proposed method takes images of DC fuses as input, automatically detects anomalous regions, and visualizes the location and morphology of irregular defects through an anomaly map, facilitating intuitive interpretation. Additionally, various data preprocessing techniques, including adjustments for diverse illumination conditions and brightness, are applied to augment the dataset, reflecting real-world environmental variations and ensuring robust performance. Experimental results demonstrate that the proposed approach effectively detects irregular defects in DC fuse data. This study is expected to contribute to the automation of DC fuse quality inspection processes, enhance the reliability of automotive electronic components, and advance quality management in smart manufacturing.

Deep learning-based fault diagnosis systems for prognostics and health management of mechanical systems is an active research topic. Notably, the absence and class imbalance of fault data (insufficient fault data compared to normal data) have been shown to cause many challenges in developing fault diagnosis systems for the manufacturing fields. Therefore, this paper presents case studies using deep learning algorithms in the absence or class imbalance of fault data. Auto-encoder-based anomaly detection method, which can be used when fault data is absent, was applied to diagnose faults in a robotic spot welding process. The anomaly detection threshold was set based on the reconstruction error of trained normal data and the confidence level of the distribution of normal data. The anomaly detection performance of the auto-encoder was verified using non-trained normal data and three sets of fault data through the threshold. As a case study for insufficient fault data, synthetic data was generated based on cGAN and applied to diagnose fault of bearing. Using the imbalanced dataset to generate synthetic fault data and to reduce the imbalance ratio, it was confirmed that the accuracy of the synthetic data generation-based 2DCNN fault diagnosis model was improved.