KIT | KIT-Bibliothek | Impressum | Datenschutz

Context-Aware Performance Benchmarking of a Fleet of Industrial Assets

Murgia, Alessandro; Tsiporkova, Elena; Verbeke, Mathias; Tourwé, Tom

Industrial assets are instrumented with sensors, connected and continuously monitored. The collected data, generally in form of time-series, is used for corrective and preventive maintenance. More advanced exploitation of this data for very diverse purposes, e.g. identifying underperformance, operational optimization or predictive maintenance, is currently an active area of research. The general methods used to analyze the time-series lead to models that are either too simple to be used in complex operational contexts or too difficult to be generalized to the whole fleet due to their asset-specific nature. Therefore, we have conceived an alternative methodology allowing to better characterize the operational context of an asset and quantify the impact on its performance. The proposed methodology allows to benchmark and profile fleet assets in a context-aware fashion, is applicable in multiple domains (even without ground truth). The methodology is evaluated on real-world data coming from a fleet of wind turbines and compared to the standard approach used in the domain. We also illustrate how the asset performance (in terms of energy production) is influenced by the operational context (in terms of environmental conditions). ... mehr

Open Access Logo

Verlagsausgabe §
DOI: 10.5445/KSP/1000087327/17
Veröffentlicht am 15.07.2020
Cover der Publikation
Zugehörige Institution(en) am KIT Institut für Informationswirtschaft und Marketing (IISM)
Publikationstyp Zeitschriftenaufsatz
Publikationsjahr 2018
Sprache Englisch
Identifikator ISSN: 2363-9881
KITopen-ID: 1000121287
Erschienen in Archives of Data Science, Series A (Online First)
Band 5
Heft 1
Seiten A17, 15 S. online
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page