A Dual-Stream Physics-Augmented Unsupervised Architecture for Runtime Embedded Vehicle Health Monitoring

TL;DR

This paper introduces a dual-stream, physics-augmented unsupervised learning architecture for vehicle health monitoring that effectively detects both transient anomalies and sustained mechanical loads using low-frequency sensor data on embedded platforms.

Contribution

It presents a novel dual-stream architecture combining unsupervised anomaly detection with physics-based load estimation for improved vehicle health monitoring.

Findings

Effective detection of surface anomalies and steady-state loads.

Low computational overhead suitable for embedded systems.

Validated on RISC-V platform with real-time performance.

Abstract

Runtime quantification of vehicle operational intensity is essential for predictive maintenance and condition monitoring in commercial and heavy-duty fleets. Traditional metrics like mileage fail to capture mechanical burden, while unsupervised deep learning models detect statistical anomalies, typically transient surface shocks, but often conflate statistical stability with mechanical rest. We identify this as a critical blind spot: high-load steady states, such as hill climbing with heavy payloads, appear statistically normal yet impose significant drivetrain fatigue. To resolve this, we propose a Dual-Stream Architecture that fuses unsupervised learning for surface anomaly detection with macroscopic physics proxies for cumulative load estimation. This approach leverages low-frequency sensor data to generate a multi-dimensional health vector, distinguishing between dynamic hazards and sustained mechanical effort. Validated on a RISC-V embedded platform, the architecture demonstrates low computational overhead, enabling comprehensive, edge-based health monitoring on resource-constrained ECUs without the latency or bandwidth costs of cloud-based monitoring.