Proposal for Concept Drift Detection in Factory Automation

This research article by KAWANOUE Shinsuke proposes a novel method for detecting concept drift in anomaly detection models within the Factory Automation domain. It addresses challenges posed by 4M changes (Man, Machine, Material, Method) that can lead to false detections and increased product defects. The proposed method integrates multiple models considering causal relationships, enhancing real-time detection capabilities. Verification results from an injection molding machine demonstrate its e