Deep neural networks have become a cornerstone of modern artificial intelligence applications, yet their decision-making processes often remainopaque. In this publication, the integration of explainable AI (XAI) techniquesinto the manufacturing processes of the optical and glass-processing industry isexplored. The work addresses the correlation between sensor-derived process data and the resulting quality of manufactured components using both classical and deep learning models. The need for transparency and interpretabilityis highlighted, especially in industrial contexts where human operators must understand and trust the system’s output to make informed decisions. The pro-posed approach allows for proactive identification of influencing error factors, paving the way for optimized process control and quality assurance.
Berteit et al. (Mon,) studied this question.