Condition monitoring and anomaly detection in cyber-physical systems

TL;DR

This paper compares machine learning methods for condition monitoring and anomaly detection in cyber-physical systems, highlighting the superior performance of supervised learning, especially tree-based algorithms, in accurately localizing faults.

Contribution

It provides a comprehensive analysis of recent ML approaches for anomaly detection and localization in industrial cyber-physical systems, emphasizing the effectiveness of supervised methods.

Findings

Supervised learning outperforms unsupervised algorithms in accuracy.

Tree-based supervised algorithms achieve 98% accuracy.

Unsupervised methods reach only 63% accuracy.

Abstract

The modern industrial environment is equipping myriads of smart manufacturing machines where the state of each device can be monitored continuously. Such monitoring can help identify possible future failures and develop a cost-effective maintenance plan. However, it is a daunting task to perform early detection with low false positives and negatives from the huge volume of collected data. This requires developing a holistic machine learning framework to address the issues in condition monitoring of high-priority components and develop efficient techniques to detect anomalies that can detect and possibly localize the faulty components. This paper presents a comparative analysis of recent machine learning approaches for robust, cost-effective anomaly detection in cyber-physical systems. While detection has been extensively studied, very few researchers have analyzed the localization of the anomalies. We show that supervised learning outperforms unsupervised algorithms. For supervised cases, we achieve near-perfect accuracy of 98 percent (specifically for tree-based algorithms). In contrast, the best-case accuracy in the unsupervised cases was 63 percent :the area under the receiver operating characteristic curve (AUC) exhibits similar outcomes as an additional metric.