Artificial intelligence-based data-driven prognostics in industry: A survey

El-Brawany, Mohamed A.; Adel Ibrahim, Dina; Elminir, Hamdy K.; Elattar, Hatem M.; Ramadan, E.A.

doi:10.1016/j.cie.2023.109605

Cited by 17 publications

(7 citation statements)

References 104 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Implementation of transfer learning for system performance insights. [46,48,51,60,73] Lack of online updates Challenges in achieving online RUL prediction and updating models with sequential-arrival data.…”

Section: Discussionmentioning

confidence: 99%

“…The development of smart sensors and IIoT technologies has given rise to an increased highquality data scarcity problem, characterized by noise and uncertainty [60]. The sensor readings involved in this context may encompass incomplete data, misreadings, unlabeled data, signal interference, disconnections, imbalanced classes, and unseen classes, highlighting the complexity of the data landscape [46].…”

Section: Analysis and CM Challengesmentioning

confidence: 99%

“…One of the most important challenges to prognostics is uncertainty management because most of the techniques are based on approximations [60]. Accurate RUL prediction depends on effective uncertainty estimation and management.…”

Section: Prognostics Models Challengesmentioning

confidence: 99%

“…The majority of models are currently offline, and transitioning them to online or onboard introduces additional challenges in constructing the DL model. So, comes the need for developing new algorithms and hardware architectures to cope with WindEurope Annual Event 2024 Journal of Physics: Conference Series 2745 (2024) 012003 continuous online learning of non-stationary sequence data as well as keeping the pre-trained model [60]. Bayesian networks have been used in an attempt to solve online RUL prediction tasks.…”

Section: Prognostics Models Challengesmentioning

confidence: 99%

“…Finally, achieving optimal efficiency and reliability needs not only the standardization of data but also the generalization of performance evaluation metrics [9,46]. To adapt the models to the variable operational conditions, utilizing transfer learning is highlighted as another opportunity in the predictive maintenance of WTs [4,46,51,60,73]. Transfer learning gains insights into the performance of a system and subsequently applies that acquired knowledge to other applications [46], implementing it on different datasets [51].…”

Section: Prognostics Models Challengesmentioning

confidence: 99%

See 4 more Smart Citations

Challenges on prognostics and health management for wind turbine components

Cuesta,

Leturiondo,

Vidal

et al. 2024

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

This paper explores the applicability of prognostics and health management (PHM) for wind turbines (WTs), presenting the PHM approach along with challenges and opportunities in the context of WT components. First, the PHM framework is introduced, consisting of three blocks: observation, analysis, and action. Critical components and failure modes for WTs are identified, and data acquisition strategies using supervisory control and data adquisition (SCADA) and condition monitoring (CM) data are discussed. Prognostics, specifically remaining useful life (RUL) estimation, employs physics model-based, data-driven, and hybrid models. Finally, challenges and opportunities related to data, analysis and CM, and developing RUL prediction models have been found. Data challenges include data standardization, limited public datasets, and data quality issues. Analysis and CM challenges address new sensorless and non-intrusive techniques, as well as the fusion of data sources. Prognostics model challenges involve uncertainty management, interpretability issues, and the need for online updates. Addressing challenges requires incorporating physical knowledge, utilizing transfer learning, and improving online RUL prediction methods.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Analysis and CM Challengesmentioning

confidence: 99%

Section: Prognostics Models Challengesmentioning

confidence: 99%

Section: Prognostics Models Challengesmentioning

confidence: 99%