Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance

Cristian Nuñez; Roberto Moreno; Victor Benitez; Jesus Pacheco

doi:10.1007/978-3-031-19496-2_19

Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance

Cristian Nuñez^*, Roberto Moreno, Victor Benitez, Jesus Pacheco

^*Corresponding author for this work

Departamento de Ingeniería Industrial

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

This paper describes the implementation of a vibration analysis in an industrial servo motor for anomaly detection. For this, a test bench was built with the purpose of simulating an industrial process. The vibration analysis was performed with an accelerometer which took the acceleration data from a running engine. For the detection of anomalies, an Autoencoder was used which was trained with samples of the normal operation of the motor in order to reconstruct a “normal operation” signal. Once the model was trained, the MAE (Mean Absolute Error) is used to see the differences between the analyzed signal and the one reconstructed by the Autoencoder, if the difference is greater than a threshold, the signal is classified as an anomaly. The proposed methodology represents an alternative to perform vibration analysis in rotative machines and can be used to conduct predictive maintenance in several industrial processes.

Original language	English
Title of host publication	Advances in Computational Intelligence - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022, Proceedings
Editors	Obdulia Pichardo Lagunas, Bella Martínez Seis, Juan Martínez-Miranda
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	252-265
Number of pages	14
ISBN (Print)	9783031194955
DOIs	https://doi.org/10.1007/978-3-031-19496-2_19
State	Published - 2022
Event	21st Mexican International Conference on Artificial Intelligence, MICAI 2022 - Monterrey, Mexico Duration: 24 Oct 2022 → 29 Oct 2022

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	13613 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	21st Mexican International Conference on Artificial Intelligence, MICAI 2022
Country/Territory	Mexico
City	Monterrey
Period	24/10/22 → 29/10/22

Bibliographical note

Publisher Copyright:
© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Keywords

Anomaly detection
Artificial intelligence
Autoenconder
Industry 4.0
Machine learning
Neural networks
Predictive maintenance

Access to Document

10.1007/978-3-031-19496-2_19

Cite this

Nuñez, C., Moreno, R., Benitez, V., & Pacheco, J. (2022). Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance. In O. Pichardo Lagunas, B. Martínez Seis, & J. Martínez-Miranda (Eds.), Advances in Computational Intelligence - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022, Proceedings (pp. 252-265). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13613 LNAI). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-19496-2_19

Nuñez, Cristian ; Moreno, Roberto ; Benitez, Victor et al. / Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance. Advances in Computational Intelligence - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022, Proceedings. editor / Obdulia Pichardo Lagunas ; Bella Martínez Seis ; Juan Martínez-Miranda. Springer Science and Business Media Deutschland GmbH, 2022. pp. 252-265 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{2c4da717d87d44a3a17b88cf9f122a9a,

title = "Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance",

abstract = "This paper describes the implementation of a vibration analysis in an industrial servo motor for anomaly detection. For this, a test bench was built with the purpose of simulating an industrial process. The vibration analysis was performed with an accelerometer which took the acceleration data from a running engine. For the detection of anomalies, an Autoencoder was used which was trained with samples of the normal operation of the motor in order to reconstruct a “normal operation” signal. Once the model was trained, the MAE (Mean Absolute Error) is used to see the differences between the analyzed signal and the one reconstructed by the Autoencoder, if the difference is greater than a threshold, the signal is classified as an anomaly. The proposed methodology represents an alternative to perform vibration analysis in rotative machines and can be used to conduct predictive maintenance in several industrial processes.",

keywords = "Anomaly detection, Artificial intelligence, Autoenconder, Industry 4.0, Machine learning, Neural networks, Predictive maintenance",

author = "Cristian Nu{\~n}ez and Roberto Moreno and Victor Benitez and Jesus Pacheco",

note = "Publisher Copyright: {\textcopyright} 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 21st Mexican International Conference on Artificial Intelligence, MICAI 2022 ; Conference date: 24-10-2022 Through 29-10-2022",

year = "2022",

doi = "10.1007/978-3-031-19496-2_19",

language = "Ingl{\'e}s",

isbn = "9783031194955",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "252--265",

editor = "{Pichardo Lagunas}, Obdulia and {Mart{\'i}nez Seis}, Bella and Juan Mart{\'i}nez-Miranda",

booktitle = "Advances in Computational Intelligence - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022, Proceedings",

address = "Alemania",

}

Nuñez, C, Moreno, R, Benitez, V & Pacheco, J 2022, Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance. in O Pichardo Lagunas, B Martínez Seis & J Martínez-Miranda (eds), Advances in Computational Intelligence - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13613 LNAI, Springer Science and Business Media Deutschland GmbH, pp. 252-265, 21st Mexican International Conference on Artificial Intelligence, MICAI 2022, Monterrey, Mexico, 24/10/22. https://doi.org/10.1007/978-3-031-19496-2_19

Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance. / Nuñez, Cristian; Moreno, Roberto; Benitez, Victor et al.
Advances in Computational Intelligence - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022, Proceedings. ed. / Obdulia Pichardo Lagunas; Bella Martínez Seis; Juan Martínez-Miranda. Springer Science and Business Media Deutschland GmbH, 2022. p. 252-265 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13613 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance

AU - Nuñez, Cristian

AU - Moreno, Roberto

AU - Benitez, Victor

AU - Pacheco, Jesus

PY - 2022

Y1 - 2022

N2 - This paper describes the implementation of a vibration analysis in an industrial servo motor for anomaly detection. For this, a test bench was built with the purpose of simulating an industrial process. The vibration analysis was performed with an accelerometer which took the acceleration data from a running engine. For the detection of anomalies, an Autoencoder was used which was trained with samples of the normal operation of the motor in order to reconstruct a “normal operation” signal. Once the model was trained, the MAE (Mean Absolute Error) is used to see the differences between the analyzed signal and the one reconstructed by the Autoencoder, if the difference is greater than a threshold, the signal is classified as an anomaly. The proposed methodology represents an alternative to perform vibration analysis in rotative machines and can be used to conduct predictive maintenance in several industrial processes.

AB - This paper describes the implementation of a vibration analysis in an industrial servo motor for anomaly detection. For this, a test bench was built with the purpose of simulating an industrial process. The vibration analysis was performed with an accelerometer which took the acceleration data from a running engine. For the detection of anomalies, an Autoencoder was used which was trained with samples of the normal operation of the motor in order to reconstruct a “normal operation” signal. Once the model was trained, the MAE (Mean Absolute Error) is used to see the differences between the analyzed signal and the one reconstructed by the Autoencoder, if the difference is greater than a threshold, the signal is classified as an anomaly. The proposed methodology represents an alternative to perform vibration analysis in rotative machines and can be used to conduct predictive maintenance in several industrial processes.

KW - Anomaly detection

KW - Artificial intelligence

KW - Autoenconder

KW - Industry 4.0

KW - Machine learning

KW - Neural networks

KW - Predictive maintenance

UR - http://www.scopus.com/inward/record.url?scp=85142811832&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-19496-2_19

DO - 10.1007/978-3-031-19496-2_19

M3 - Contribución a la conferencia

AN - SCOPUS:85142811832

SN - 9783031194955

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 252

EP - 265

BT - Advances in Computational Intelligence - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022, Proceedings

A2 - Pichardo Lagunas, Obdulia

A2 - Martínez Seis, Bella

A2 - Martínez-Miranda, Juan

PB - Springer Science and Business Media Deutschland GmbH

T2 - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022

Y2 - 24 October 2022 through 29 October 2022

ER -

Nuñez C, Moreno R, Benitez V , Pacheco J. Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance. In Pichardo Lagunas O, Martínez Seis B, Martínez-Miranda J, editors, Advances in Computational Intelligence - 21st Mexican International Conference on Artificial Intelligence, MICAI 2022, Proceedings. Springer Science and Business Media Deutschland GmbH. 2022. p. 252-265. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-19496-2_19

Vibration Analysis of an Industrial Motor with Autoencoder for Predictive Maintenance

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this