Adaptive Statistical Windowing for Anomaly Detection in Industrial Sensor Signals
IEEE Sensors Letters · 2026-06 · 10 (6) · 7003104-7003104
DOI: 10.1109/lsens.2026.3690282 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Sensors Letters, 2026-06. Verfasst von Lincoln Moura de Oliveira und 3 weiteren Autor:innen. Die Arbeit nennt 15 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Sensor signal validation is essential for industrial automation, particularly in safety-critical environments. This letter proposes a training-free statistical framework with a hierarchical sliding-window strategy that sequentially performs sensor validation and anomaly detection. The approach combines signal stationarization with adaptive interquartile-range thresholds updated dynamically to handle nonstationary conditions. The framework is evaluated on the Tennessee Eastman Process (TEP) benchmark and univariate cases from the Skoltech Anomaly Benchmark (SKAB). In TEP, the proposed method effectively identifies abnormal operating regimes and outperforms matrix profile, $k$-Shape, and subsequence-level Isolation Forest baselines, achieving 99.8% precision and an F1-score of 74.3%. In SKAB, it achieves the best overall tradeoff among the compared methods, reaching 84.6% precision, a 66.4% F1-score, and the lowest false positive rate. Owing to its low complexity and sensor-agnostic design, the framework is suitable for large-scale industrial monitoring without model training.
Zitieren
APA
Oliveira, L. M. D., Saavedra, J. A. D. L. C., Kattel, M. B. E., & Antunes, F. L. M. (2026). Adaptive Statistical Windowing for Anomaly Detection in Industrial Sensor Signals. IEEE Sensors Letters, 10(6), 7003104-7003104. https://doi.org/10.1109/lsens.2026.3690282
Vancouver
Oliveira LMD, Saavedra JADLC, Kattel MBE, Antunes FLM. Adaptive Statistical Windowing for Anomaly Detection in Industrial Sensor Signals. IEEE Sensors Letters. 2026;10(6):7003104-7003104. doi:10.1109/lsens.2026.3690282
BibTeX
@article{Oliveira2026,
author = {Oliveira, Lincoln Moura de and Saavedra, Jorge Armando De La Cruz and Kattel, Menaouar Berrehil El and Antunes, Fernando Luiz Marcelo},
title = {Adaptive Statistical Windowing for Anomaly Detection in Industrial Sensor Signals},
journal = {IEEE Sensors Letters},
year = {2026},
volume = {10},
number = {6},
pages = {7003104--7003104},
doi = {10.1109/lsens.2026.3690282}
}
EUVIMED