Extending Asset Lifespan Through Data Augmentation-Assisted Quality Control

Quality control is essential for the life cycle of components and has an impact on the costs and maintenance time of the entire production chain. In modern quality control, AI power could be unleashed if enough information for training AI models is always available. In this work, an approach based on data augmentation has been applied to oil plug quality control to mitigate the lack of images of defective parts, which hinders the training of classification models. The proposed approach has demonstrated a positive impact on quality inspection by maintaining or reducing the percentage of classification errors with respect to a baseline trained with images of real oil plugs.