Unsupervised Anomaly Detection in Process-Complex Industrial Time Series: A Real-World Case Study

TL;DR

This paper evaluates various anomaly detection models on complex industrial time-series data, highlighting the superior performance of temporal convolutional autoencoders over traditional methods.

Contribution

It provides an empirical comparison of model classes on real-world industrial data, emphasizing the importance of model choice and tuning for complex process variability.

Findings

Isolation Forest underperforms on complex data

Autoencoders outperform classical methods

Temporal convolutional autoencoders are most robust

Abstract

Industrial time-series data from real production environments exhibits substantially higher complexity than commonly used benchmark datasets, primarily due to heterogeneous, multi-stage operational processes. As a result, anomaly detection methods validated under simplified conditions often fail to generalize to industrial settings. This work presents an empirical study on a unique dataset collected from fully operational industrial machinery, explicitly capturing pronounced process-induced variability. We evaluate which model classes are capable of capturing this complexity, starting with a classical Isolation Forest baseline and extending to multiple autoencoder architectures. Experimental results show that Isolation Forest is insufficient for modeling the non-periodic, multi-scale dynamics present in the data, whereas autoencoders consistently perform better. Among them, temporal convolutional autoencoders achieve the most robust performance, while recurrent and variational variants require more careful tuning.