September 2021 Unsupervised streaming anomaly detection for instrumented infrastructure
Henrique Hoeltgebaum, Niall Adams, F. Din-Houn Lau
Author Affiliations +
Ann. Appl. Stat. 15(3): 1101-1125 (September 2021). DOI: 10.1214/20-AOAS1424


Structural health monitoring (SHM) often involves instrumenting structures with distributed sensor networks. These networks typically provide high frequency data describing the spatiotemporal behaviour of the assets. A main objective of SHM is to reason about changes in structures’ behaviour using sensor data. We construct a streaming anomaly detection method for data from a railway bridge instrumented with a fibre-optic sensor network. The data exhibits trend over time, which may be partially attributable to environmental factors, calling for temporally adaptive estimation. Exploiting a latent structure present in the data motivates a quantity of interest for anomaly detection. This quantity is estimated, sequentially and adaptively, using a new formulation of streaming principal component analysis. Anomaly detection for this quantity is then provided using conformal prediction. Like all streaming methods, the proposed method has free control parameters which are set using simulations based on bridge data. Experiments demonstrate that this method can operate at the sampling frequency of the data while providing accurate tracking of the target quantity. Further, the anomaly detection is able to detect train passage events. Finally, the method reveals a previously unreported cyclic structure present in the data.


Download Citation

Henrique Hoeltgebaum. Niall Adams. F. Din-Houn Lau. "Unsupervised streaming anomaly detection for instrumented infrastructure." Ann. Appl. Stat. 15 (3) 1101 - 1125, September 2021.


Received: 1 February 2020; Revised: 1 November 2020; Published: September 2021
First available in Project Euclid: 23 September 2021

MathSciNet: MR4316644
zbMATH: 1478.62328
Digital Object Identifier: 10.1214/20-AOAS1424

Keywords: adaptive estimation , conformal prediction , Stochastic gradient descent , streaming PCA , Structural health monitoring

Rights: Copyright © 2021 Institute of Mathematical Statistics


This article is only available to subscribers.
It is not available for individual sale.

Vol.15 • No. 3 • September 2021
Back to Top