Machine learning for anomaly detection and process phase classification to improve safety and maintenance activities

• Published in: Journal of manufacturing systems Vol. 56; pp. 117 - 132
• Main Authors: Quatrini, Elena, Costantino, Francesco, Di Gravio, Giulio, Patriarca, Riccardo
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.07.2020
• Subjects: Condition monitoring; Condition-based maintenance; Decision jungles; forests; Machine learning; Random; Condition monitoring; Random; forests; Condition-based maintenance; Decision jungles; Machine learning

•Define a methodology for anomaly detection with real time data from multiphase industrial process.•RFA and DJA have comparable results for the identification of process phases.•RFA has better performance than DJA for the anomaly detection.•DJA underperforms for anomalies close to the thresholds, du...