A Temporal Anomaly Detection System for Vehicles utilizing Functional Working Groups and Sensor Channels

TL;DR

This work addresses anomaly detection in vehicle sensor data organized by Functional Working Groups (FWGs) using a multi-phase approach built on Temporal Convolution Networks (TCN). A Seq2Seq forecasting model predicts future multi-channel sensor observations from a current window, and anomalies are detected by comparing predictions to actual values via a Mahalanobis-distance-based threshold derived from ROC analysis. Leveraging the VePRO dataset, the study demonstrates high performance, achieving ROC-AUC of $0.982$, 91% true-anomaly detection, and 96% overall detection accuracy, with improved results when integrating channels across FWGs. The approach enables precise localization of anomalies to specific FWGs and channels, offering practical potential for predictive maintenance in modern vehicles. Future work includes comparing with LSTM/ConvLSTM variants and exploring hybrid or digital-twin-inspired anomaly-detection frameworks.

Abstract

A modern vehicle fitted with sensors, actuators, and Electronic Control Units (ECUs) can be divided into several operational subsystems called Functional Working Groups (FWGs). Examples of these FWGs include the engine system, transmission, fuel system, brakes, etc. Each FWG has associated sensor-channels that gauge vehicular operating conditions. This data rich environment is conducive to the development of Predictive Maintenance (PdM) technologies. Undercutting various PdM technologies is the need for robust anomaly detection models that can identify events or observations which deviate significantly from the majority of the data and do not conform to a well defined notion of normal vehicular operational behavior. In this paper, we introduce the Vehicle Performance, Reliability, and Operations (VePRO) dataset and use it to create a multi-phased approach to anomaly detection. Utilizing Temporal Convolution Networks (TCN), our anomaly detection system can achieve 96% detection accuracy and accurately predicts 91% of true anomalies. The performance of our anomaly detection system improves when sensor channels from multiple FWGs are utilized.

Figures (8)

• Figure 1: A causal dilated network with dilated factors $d$ = 1,2, 4 and a kernel size of 3. The first hidden layer has a dilation factor of 2, and the second hidden layer has a dilation factor of 4. The receptive field can accommodate all input sequence values bai2018empirical.
• Figure 2: The figure on the left shows a TCN residual block with a dilation factor of $d$ and a filter of $k$. When the dimensions of the residual input and output differ, a $1 \times 1$ convolution is added. The figure on the right is an example of residual connections in a TCN with $d$ = 1 and $k$ = 3. In this figure, the green lines show identity mappings, while the blue lines show filters in the residual function bai2018empirical.
• Figure 3: A graphical representation of the various FWG in a vehicle, including the engine, transmission, fuel, and brake, as well as the sensor channels for each of these FWG CoverHMMWVTM.
• Figure 4: Snapshot of VePro Dataset raw observations for a period of 25 minutes on 2014-Jan-30.
• Figure 5: An overview of the proposed three-phased system architecture. In the initial phase, raw sensor data is cleansed, transformed, and scaled. The second phase predicts the sequence of outputs. In the third phase, anomalies are identified by comparing error values to a predefined threshold.