MA-VAE: Multi-head Attention-based Variational Autoencoder Approach for Anomaly Detection in Multivariate Time-series Applied to Automotive Endurance Powertrain Testing

TL;DR

This paper introduces MA-VAE, a novel variational autoencoder with multi-head attention, designed for effective anomaly detection in complex multivariate time-series data from automotive testing, reducing false positives and addressing bypass issues.

Contribution

The paper presents a new MA-VAE model that improves anomaly detection accuracy and robustness in multivariate time-series data, with innovative methods for bypass prevention and time series remapping.

Findings

Detects 67% of anomalies with 9% false positive rate

Reduces bypass phenomenon in anomaly detection

Effective with limited training data

Abstract

A clear need for automatic anomaly detection applied to automotive testing has emerged as more and more attention is paid to the data recorded and manual evaluation by humans reaches its capacity. Such real-world data is massive, diverse, multivariate and temporal in nature, therefore requiring modelling of the testee behaviour. We propose a variational autoencoder with multi-head attention (MA-VAE), which, when trained on unlabelled data, not only provides very few false positives but also manages to detect the majority of the anomalies presented. In addition to that, the approach offers a novel way to avoid the bypass phenomenon, an undesirable behaviour investigated in literature. Lastly, the approach also introduces a new method to remap individual windows to a continuous time series. The results are presented in the context of a real-world industrial data set and several experiments are undertaken to further investigate certain aspects of the proposed model. When configured properly, it is 9% of the time wrong when an anomaly is flagged and discovers 67% of the anomalies present. Also, MA-VAE has the potential to perform well with only a fraction of the training and validation subset, however, to extract it, a more sophisticated threshold estimation method is required.