Predictive maintenance (PdM) is a set of actions and techniques to early detect failures and defects on machines before they occur, and the usage of machine learning and deep learning techniques in predictive maintenance has increased during the last years. Even with this increase of the literature, there is still a gap concerning the application of such techniques for PdM in the industry, as there are no clear guidelines about which information to use for a PdM system, how to process the information, and what machine learning techniques should be used in order to obtain acceptable results. This scoping review is performed in order to observe the current status on the use of Machine Learning and Deep Learning in predictive maintenance in academia and provide answer to the questions related to these guidelines. For this purpose, a literature review of the last five years is carried out, using those articles that cover information about sources of information used for PdM, the treatment given to such data and the machine learning (ML) methods or techniques used. The Web of Science: Core Collection database is used as a source of information, specifically the Science Citation Index Expanded (SCIE). The review shows that there are different information sources used for machine learning and deep learning in PdM, depending on the origin of the data and the availability of it, and as well whether the data sets are private or public. Also, we can observe that data used for both training and making predictions does not only use traditional pre-processing techniques, but that about one fifth of the articles even propose new techniques in this field. Additionally, we compare a wide range of techniques and algorithms which are used in Deep Learning -being ANN the most used- and in Machine Learning, being SVM the most used algorithm, closely followed by Random Forest. Based on the results, we provide indications about how to apply ML for PdM in industry.

(2024). Screening of Machine Learning Techniques on Predictive Maintenance: a Scoping Review [journal article - articolo]. In DYNA. Retrieved from https://hdl.handle.net/10446/258029

Screening of Machine Learning Techniques on Predictive Maintenance: a Scoping Review

Mazzoleni, Mirko;Ferramosca, Antonio;
2024-01-01

Abstract

Predictive maintenance (PdM) is a set of actions and techniques to early detect failures and defects on machines before they occur, and the usage of machine learning and deep learning techniques in predictive maintenance has increased during the last years. Even with this increase of the literature, there is still a gap concerning the application of such techniques for PdM in the industry, as there are no clear guidelines about which information to use for a PdM system, how to process the information, and what machine learning techniques should be used in order to obtain acceptable results. This scoping review is performed in order to observe the current status on the use of Machine Learning and Deep Learning in predictive maintenance in academia and provide answer to the questions related to these guidelines. For this purpose, a literature review of the last five years is carried out, using those articles that cover information about sources of information used for PdM, the treatment given to such data and the machine learning (ML) methods or techniques used. The Web of Science: Core Collection database is used as a source of information, specifically the Science Citation Index Expanded (SCIE). The review shows that there are different information sources used for machine learning and deep learning in PdM, depending on the origin of the data and the availability of it, and as well whether the data sets are private or public. Also, we can observe that data used for both training and making predictions does not only use traditional pre-processing techniques, but that about one fifth of the articles even propose new techniques in this field. Additionally, we compare a wide range of techniques and algorithms which are used in Deep Learning -being ANN the most used- and in Machine Learning, being SVM the most used algorithm, closely followed by Random Forest. Based on the results, we provide indications about how to apply ML for PdM in industry.
articolo
2024
El mantenimiento predictivo (PdM) es un conjunto de acciones y técnicas para detectar tempranamente fallos y defectos en máquinas antes de que ocurran. El uso de técnicas de aprendizaje automático y aprendizaje profundo en el mantenimiento predictivo ha aumentado en los últimos años. A pesar de este aumento en la literatura, todavía existe una brecha en cuanto a la aplicación de tales técnicas en la industria, ya que no existen pautas claras sobre qué información utilizar en un sistema de PdM, cómo procesar la información y qué técnicas de aprendizaje automático se deben usar para obtener resultados aceptables. Esta revisión de alcance se realiza para observar el estado actual del uso del aprendizaje automático y el aprendizaje profundo en el mantenimiento predictivo en la academia y proporcionar respuestas a las preguntas relacionadas con estas pautas. Para este propósito, se lleva a cabo una revisión de la literatura de los últimos cinco años, utilizando los artículos que cubren información sobre las fuentes de información utilizadas para el PdM, el tratamiento dado a dichos datos y los métodos o técnicas de aprendizaje automático (ML) utilizados. La base de datos Web of Science: Core Collection se utiliza como fuente de información, específicamente el Science Citation Index Expanded (SCIE). La revisión muestra que existen diferentes fuentes de información utilizadas para el aprendizaje automático y el aprendizaje profundo en el PdM, dependiendo del origen de los datos y su disponibilidad, así como si los conjuntos de datos son privados o públicos. Además, podemos observar que los datos utilizados tanto para el entrenamiento como para realizar predicciones no solo utilizan técnicas de preprocesamiento tradicionales, sino que aproximadamente una quinta parte de los artículos incluso proponen nuevas técnicas en este campo. Además, comparamos una amplia gama de técnicas y algoritmos que se utilizan en el Aprendizaje Profundo, siendo las Redes Neuronales Artificiales (ANN) las más utilizadas, y en el Aprendizaje Automático, siendo el SVM (Máquinas de Soporte Vectorial) el algoritmo más utilizado, seguido de cerca por Random Forest. Basándonos en los resultados, proporcionamos indicaciones sobre cómo aplicar el Aprendizaje Automático para Mantenimiento Predictivo en la industria.
Campos Olivares, Daniel; Carrasco Muñoz, Alejandro; Mazzoleni, Mirko; Ferramosca, Antonio; Luque Sendra, Amalia
(2024). Screening of Machine Learning Techniques on Predictive Maintenance: a Scoping Review [journal article - articolo]. In DYNA. Retrieved from https://hdl.handle.net/10446/258029
File allegato/i alla scheda:
File Dimensione del file Formato  
10950_english.pdf

Solo gestori di archivio

Versione: publisher's version - versione editoriale
Licenza: Licenza default Aisberg
Dimensione del file 539.11 kB
Formato Adobe PDF
539.11 kB Adobe PDF   Visualizza/Apri
Pubblicazioni consigliate

Aisberg ©2008 Servizi bibliotecari, Università degli studi di Bergamo | Terms of use/Condizioni di utilizzo

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10446/258029
Citazioni
  • Scopus 1
  • ???jsp.display-item.citation.isi??? 1
social impact