RESEARCHarXiv CS.AI·8d ago
Product-Aware Deep Autoencoders for Robust Process Monitoring in Multi-Product Cyber-Physical Systems
Industry 4.0 requires robust anomaly detection in Cyber-Physical Systems (CPS), but product-agnostic models have blind spots for subtle anomalies or targeted attacks. This work demonstrates this vulnerability and proposes a Product-Aware Autoencoder as a mitigation by restricting the learning domain.
27