A Benchmark for Real-Time Anomaly Detection Algorithms Applied in Industry 4.0

Autor(en): Stahmann, Philip
Rieger, Bodo 
Herausgeber: Nicosia, G.
Giuffrida, G.
Ojha, V.
La Malfa, E.
La Malfa, G.
Pardalos, P.
Di Fatta, G.
Umeton, R.
Stichwörter: Algorithm benchmark; Anomaly detection; Anomaly recognition; Anomaly-detection algorithms; Benchmarking; Chemical detection; Data stream; Industry 4.0; Petroleum reservoir evaluation; Production-machines; Real time analysis; Real-time analysis; Real-time anomaly detections; Self-diagnose; Signal detection; Streaming analytic; Streaming analytics
Erscheinungsdatum: 2023
Herausgeber: Springer Science and Business Media Deutschland GmbH
Journal: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen: 13810 LNCS
Startseite: 20 – 34
Industry 4.0 describes flexibly combinable production machines enabling efficient fulfillment of individual requirements. Timely and automated anomaly recognition by means of machine self-diagnosis might support efficiency. Various algorithms have been developed in recent years to detect anomalies in data streams. Due to their diverse functionality, the application of different real-time anomaly detection algorithms to the same data stream may lead to different results. Existing algorithms as well as mechanisms for their evaluation and selection are context-independent and not suited to industry 4.0 settings. In this research paper, an industry 4.0 specific benchmark for real-time anomaly detection algorithms is developed on the basis of six design principles in the categories timeliness, threshold setting and qualitative classification. Given context-specific input parameters, the benchmark ranks algorithms according to their suitability for real-time anomaly detection in production datasets. The application of the benchmark is demonstrated and evaluated on the basis of two case studies. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
Cited by: 0; Conference name: 8th International Conference on Machine Learning, Optimization, and Data Science, LOD 2022, held in conjunction with the 2nd Advanced Course and Symposium on Artificial Intelligence and Neuroscience, ACAIN 2022; Conference date: 18 September 2022 through 22 September 2022; Conference code: 291259
ISBN: 9783031255984
ISSN: 0302-9743
DOI: 10.1007/978-3-031-25599-1_3
Externe URL: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85151059619&doi=10.1007%2f978-3-031-25599-1_3&partnerID=40&md5=4a0333dd84784575ad385a8a84461596

Show full item record

Page view(s)

Last Week
Last month
checked on May 27, 2024

Google ScholarTM