Automated quality control of vacuum insulated glazing by convolutional neural network image classification

Vacuum insulated glazing (VIG) is a highly thermally insulating window technology, which boasts an extremely thin profile and lower weight as compared to gas-filled units of equivalent performance. The VIG double-pane configuration with submillimeter vacuum gap maintained by small pillars positioned in between the panes, can damage glass during manufacturing, transportation installation. For purpose automatically classifying damage, we have developed, trained, tested deep learning model using convolutional neural networks. We employ state-of-the-art methods Grad-CAM Score-CAM explainable Artificial Intelligence (XAI) provide understanding internal mechanisms were able show that our classifier outperforms ResNet50V2 for identification crack locations geometry. Further analysis model's predictive capabilities demonstrates its superiority over ResNet models terms convergence speed, accuracy, precision at 100% recall AUC ROC.