Predictive Maintenance Planning Using Renewal Reward Processes and Probabilistic RUL Prognostics: Analyzing the Influence of Accuracy and Sharpness of Prognostics

M.A. Mitici; I.I. de Pater; Zhiguo Zeng; Anne Barros

doi:10.3850/978-981-18-8071-1_P064-cd

Predictive Maintenance Planning Using Renewal Reward Processes and Probabilistic RUL Prognostics: Analyzing the Influence of Accuracy and Sharpness of Prognostics

M.A. Mitici, I.I. de Pater, Zhiguo Zeng, Anne Barros

Air Transport & Operations

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

2 Downloads (Pure)

Abstract

We pose the maintenance planning for systems using probabilistic Remaining Useful Life (RUL) prognostics as a renewal reward process. Data-driven probabilistic RUL prognostics are obtained using a Convolutional Neural Network with Monte Carlo dropout. The maintenance planning model is illustrated for aircraft turbofan engines. The results show that in the initial monitoring phase, the accuracy and sharpness of the RUL prognostics is relatively small. The maintenance of the engines is therefore scheduled far in the future. As the usage of the engine increases, the accuracy of the prognostics improves, while the sharpness remains relatively small. As soon as the estimated probability of the RUL is skewed towards 0, the maintenance planning model consistently indicates it is optimal to replace the engines immediately, i.e., "now". This shows that probabilistic RUL prognostics support an effective maintenance planning of the engines, despite being imperfect with respect to accuracy and sharpness.

Original language	English
Title of host publication	Proceedings of the 33rd European Safety and Reliability Conference (ESREL 2023)
Editors	mario brito, Terje Aven, Piero Baraldi, Marko Cepin, Enrico Zio
Publisher	Research Publishing
Pages	1034-1041
ISBN (Electronic)	978-981-18-8071-1
DOIs	https://doi.org/10.3850/978-981-18-8071-1_P064-cd
Publication status	Published - 2023
Event	The 33rd European Safety and Reliability Conference (ESREL 2023): The Future od Safety in a Reconnected World - University of Southampton, Southampton, United Kingdom Duration: 3 Sept 2023 → 7 Sept 2023 https://www.esrel2023.com/

Conference

Conference	The 33rd European Safety and Reliability Conference (ESREL 2023)
Abbreviated title	ESREL 2023
Country/Territory	United Kingdom
City	Southampton
Period	3/09/23 → 7/09/23
Internet address	https://www.esrel2023.com/

Bibliographical note

Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care
Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

Keywords

Predictive maintenance planning
Probabilistic RUL
prognostics
Aircraft engines
Renewal processes
Convolutional neural network
Monte Carlo dropout

Access to Document

10.3850/978-981-18-8071-1_P064-cd

P064Final published version, 359 KB

Cite this

Mitici, M. A., de Pater, I. I., Zeng, Z., & Barros, A. (2023). Predictive Maintenance Planning Using Renewal Reward Processes and Probabilistic RUL Prognostics: Analyzing the Influence of Accuracy and Sharpness of Prognostics. In M. brito, T. Aven, P. Baraldi, M. Cepin, & E. Zio (Eds.), Proceedings of the 33rd European Safety and Reliability Conference (ESREL 2023) (pp. 1034-1041). Research Publishing. https://doi.org/10.3850/978-981-18-8071-1_P064-cd

Mitici, M.A. ; de Pater, I.I. ; Zeng, Zhiguo et al. / Predictive Maintenance Planning Using Renewal Reward Processes and Probabilistic RUL Prognostics : Analyzing the Influence of Accuracy and Sharpness of Prognostics. Proceedings of the 33rd European Safety and Reliability Conference (ESREL 2023). editor / mario brito ; Terje Aven ; Piero Baraldi ; Marko Cepin ; Enrico Zio. Research Publishing, 2023. pp. 1034-1041

@inproceedings{0c77360aa629466db04c5e6e05f51694,

title = "Predictive Maintenance Planning Using Renewal Reward Processes and Probabilistic RUL Prognostics: Analyzing the Influence of Accuracy and Sharpness of Prognostics",

abstract = "We pose the maintenance planning for systems using probabilistic Remaining Useful Life (RUL) prognostics as a renewal reward process. Data-driven probabilistic RUL prognostics are obtained using a Convolutional Neural Network with Monte Carlo dropout. The maintenance planning model is illustrated for aircraft turbofan engines. The results show that in the initial monitoring phase, the accuracy and sharpness of the RUL prognostics is relatively small. The maintenance of the engines is therefore scheduled far in the future. As the usage of the engine increases, the accuracy of the prognostics improves, while the sharpness remains relatively small. As soon as the estimated probability of the RUL is skewed towards 0, the maintenance planning model consistently indicates it is optimal to replace the engines immediately, i.e., {"}now{"}. This shows that probabilistic RUL prognostics support an effective maintenance planning of the engines, despite being imperfect with respect to accuracy and sharpness. ",

keywords = "Predictive maintenance planning, Probabilistic RUL, prognostics, Aircraft engines, Renewal processes, Convolutional neural network, Monte Carlo dropout",

author = "M.A. Mitici and {de Pater}, I.I. and Zhiguo Zeng and Anne Barros",

note = "Green Open Access added to TU Delft Institutional Repository {\textquoteleft}You share, we take care!{\textquoteright} – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.; The 33rd European Safety and Reliability Conference (ESREL 2023) : The Future od Safety in a Reconnected World, ESREL 2023 ; Conference date: 03-09-2023 Through 07-09-2023",

year = "2023",

doi = "10.3850/978-981-18-8071-1_P064-cd",

language = "English",

pages = "1034--1041",

editor = "mario brito and Terje Aven and Piero Baraldi and Marko Cepin and Enrico Zio",

booktitle = "Proceedings of the 33rd European Safety and Reliability Conference (ESREL 2023)",

publisher = "Research Publishing",

address = "Singapore",

url = "https://www.esrel2023.com/",

}

Mitici, MA, de Pater, II, Zeng, Z & Barros, A 2023, Predictive Maintenance Planning Using Renewal Reward Processes and Probabilistic RUL Prognostics: Analyzing the Influence of Accuracy and Sharpness of Prognostics. in M brito, T Aven, P Baraldi, M Cepin & E Zio (eds), Proceedings of the 33rd European Safety and Reliability Conference (ESREL 2023). Research Publishing, pp. 1034-1041, The 33rd European Safety and Reliability Conference (ESREL 2023), Southampton, United Kingdom, 3/09/23. https://doi.org/10.3850/978-981-18-8071-1_P064-cd

Predictive Maintenance Planning Using Renewal Reward Processes and Probabilistic RUL Prognostics: Analyzing the Influence of Accuracy and Sharpness of Prognostics. / Mitici, M.A.; de Pater, I.I.; Zeng, Zhiguo et al.
Proceedings of the 33rd European Safety and Reliability Conference (ESREL 2023). ed. / mario brito; Terje Aven; Piero Baraldi; Marko Cepin; Enrico Zio. Research Publishing, 2023. p. 1034-1041.

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Predictive Maintenance Planning Using Renewal Reward Processes and Probabilistic RUL Prognostics

T2 - The 33rd European Safety and Reliability Conference (ESREL 2023)

AU - Mitici, M.A.

AU - de Pater, I.I.

AU - Zeng, Zhiguo

AU - Barros, Anne

N1 - Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

PY - 2023

Y1 - 2023

N2 - We pose the maintenance planning for systems using probabilistic Remaining Useful Life (RUL) prognostics as a renewal reward process. Data-driven probabilistic RUL prognostics are obtained using a Convolutional Neural Network with Monte Carlo dropout. The maintenance planning model is illustrated for aircraft turbofan engines. The results show that in the initial monitoring phase, the accuracy and sharpness of the RUL prognostics is relatively small. The maintenance of the engines is therefore scheduled far in the future. As the usage of the engine increases, the accuracy of the prognostics improves, while the sharpness remains relatively small. As soon as the estimated probability of the RUL is skewed towards 0, the maintenance planning model consistently indicates it is optimal to replace the engines immediately, i.e., "now". This shows that probabilistic RUL prognostics support an effective maintenance planning of the engines, despite being imperfect with respect to accuracy and sharpness.

AB - We pose the maintenance planning for systems using probabilistic Remaining Useful Life (RUL) prognostics as a renewal reward process. Data-driven probabilistic RUL prognostics are obtained using a Convolutional Neural Network with Monte Carlo dropout. The maintenance planning model is illustrated for aircraft turbofan engines. The results show that in the initial monitoring phase, the accuracy and sharpness of the RUL prognostics is relatively small. The maintenance of the engines is therefore scheduled far in the future. As the usage of the engine increases, the accuracy of the prognostics improves, while the sharpness remains relatively small. As soon as the estimated probability of the RUL is skewed towards 0, the maintenance planning model consistently indicates it is optimal to replace the engines immediately, i.e., "now". This shows that probabilistic RUL prognostics support an effective maintenance planning of the engines, despite being imperfect with respect to accuracy and sharpness.

KW - Predictive maintenance planning

KW - Probabilistic RUL

KW - prognostics

KW - Aircraft engines

KW - Renewal processes

KW - Convolutional neural network

KW - Monte Carlo dropout

U2 - 10.3850/978-981-18-8071-1_P064-cd

DO - 10.3850/978-981-18-8071-1_P064-cd

M3 - Conference contribution

SP - 1034

EP - 1041

BT - Proceedings of the 33rd European Safety and Reliability Conference (ESREL 2023)

A2 - brito, mario

A2 - Aven, Terje

A2 - Baraldi, Piero

A2 - Cepin, Marko

A2 - Zio, Enrico

PB - Research Publishing

Y2 - 3 September 2023 through 7 September 2023

ER -

Mitici MA, de Pater II, Zeng Z, Barros A. Predictive Maintenance Planning Using Renewal Reward Processes and Probabilistic RUL Prognostics: Analyzing the Influence of Accuracy and Sharpness of Prognostics. In brito M, Aven T, Baraldi P, Cepin M, Zio E, editors, Proceedings of the 33rd European Safety and Reliability Conference (ESREL 2023). Research Publishing. 2023. p. 1034-1041 doi: 10.3850/978-981-18-8071-1_P064-cd

Predictive Maintenance Planning Using Renewal Reward Processes and Probabilistic RUL Prognostics: Analyzing the Influence of Accuracy and Sharpness of Prognostics

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Fingerprint

Cite this