Details
| Original language | English |
|---|---|
| Title of host publication | Proceedings - 2025 25th International Conference on Software Quality, Reliability and Security Companion, QRS-C 2025 |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 483-491 |
| Number of pages | 9 |
| ISBN (electronic) | 9781665477734 |
| ISBN (print) | 978-1-6654-7774-1 |
| Publication status | Published - 16 Jul 2025 |
| Event | 25th International Conference on Software Quality, Reliability and Security Companion, QRS-C 2025 - Hangzhou, China Duration: 16 Jul 2025 → 20 Jul 2025 |
Publication series
| Name | Proceedings - International Conference on Software Quality, Reliability and Security Companion |
|---|---|
| ISSN (Print) | 2693-938X |
| ISSN (electronic) | 2693-9371 |
Abstract
Fault prognostics is one of the key enablers for the realisation of predictive maintenance. In today's era of digital transformation, deep learning (DL) has proven to be a promising data-driven solution for the task of fault prognostics with the ability to accurately predict the remaining useful life of industrial assets based on their historical condition-monitoring data. However, the deployment of DL-based fault prognostics models in practice still faces a number of critical challenges, especially in application scenarios with dynamic or evolving contexts suffering from data distribution shifts. Jointly training DL models using data from all contexts at once is typically impossible due to practical requirements regarding privacy constraints and resource limitations. Moreover, fine-tuning or training DL models in a classical sequential manner has been observed to typically suffer from catastrophic forgetting where adapting to a new context leads to drastically forgetting what has been learned previously. To address this problem, we proposed a novel lifelong learning framework with Bayesian neural networks for fault prognostics. Our proposed Bayesian lifelong learning method focuses not only on preserving the old knowledge learned so far but also allows to reserve the model flexibility for learning new knowledge from upcoming contexts. Experimental results on the benchmark C-MAPSS dataset of turbofan engine degradation data show the superiority of our proposed framework over other relevant lifelong learning methods. On average, we achieve a performance improvement of 10.2% and 75.3% in terms of final prediction accuracy and forgetting measure, respectively.
Keywords
- Bayesian neural networks, deep learning, lifelong learning, Predictive maintenance, prognostics and health management, RUL prediction
ASJC Scopus subject areas
- Engineering(all)
- Safety, Risk, Reliability and Quality
- Mathematics(all)
- Modelling and Simulation
- Computer Science(all)
- Artificial Intelligence
- Computer Science(all)
- Computer Science Applications
- Computer Science(all)
- Software
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
Proceedings - 2025 25th International Conference on Software Quality, Reliability and Security Companion, QRS-C 2025. Institute of Electrical and Electronics Engineers Inc., 2025. p. 483-491 (Proceedings - International Conference on Software Quality, Reliability and Security Companion).
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Lifelong Learning for Fault Prognostics in Predictive Maintenance with Bayesian Neural Networks
AU - Xuan, Quy Le
AU - Munderloh, Marco
AU - Ostermann, Jörn
N1 - Publisher Copyright: © 2025 IEEE.
PY - 2025/7/16
Y1 - 2025/7/16
N2 - Fault prognostics is one of the key enablers for the realisation of predictive maintenance. In today's era of digital transformation, deep learning (DL) has proven to be a promising data-driven solution for the task of fault prognostics with the ability to accurately predict the remaining useful life of industrial assets based on their historical condition-monitoring data. However, the deployment of DL-based fault prognostics models in practice still faces a number of critical challenges, especially in application scenarios with dynamic or evolving contexts suffering from data distribution shifts. Jointly training DL models using data from all contexts at once is typically impossible due to practical requirements regarding privacy constraints and resource limitations. Moreover, fine-tuning or training DL models in a classical sequential manner has been observed to typically suffer from catastrophic forgetting where adapting to a new context leads to drastically forgetting what has been learned previously. To address this problem, we proposed a novel lifelong learning framework with Bayesian neural networks for fault prognostics. Our proposed Bayesian lifelong learning method focuses not only on preserving the old knowledge learned so far but also allows to reserve the model flexibility for learning new knowledge from upcoming contexts. Experimental results on the benchmark C-MAPSS dataset of turbofan engine degradation data show the superiority of our proposed framework over other relevant lifelong learning methods. On average, we achieve a performance improvement of 10.2% and 75.3% in terms of final prediction accuracy and forgetting measure, respectively.
AB - Fault prognostics is one of the key enablers for the realisation of predictive maintenance. In today's era of digital transformation, deep learning (DL) has proven to be a promising data-driven solution for the task of fault prognostics with the ability to accurately predict the remaining useful life of industrial assets based on their historical condition-monitoring data. However, the deployment of DL-based fault prognostics models in practice still faces a number of critical challenges, especially in application scenarios with dynamic or evolving contexts suffering from data distribution shifts. Jointly training DL models using data from all contexts at once is typically impossible due to practical requirements regarding privacy constraints and resource limitations. Moreover, fine-tuning or training DL models in a classical sequential manner has been observed to typically suffer from catastrophic forgetting where adapting to a new context leads to drastically forgetting what has been learned previously. To address this problem, we proposed a novel lifelong learning framework with Bayesian neural networks for fault prognostics. Our proposed Bayesian lifelong learning method focuses not only on preserving the old knowledge learned so far but also allows to reserve the model flexibility for learning new knowledge from upcoming contexts. Experimental results on the benchmark C-MAPSS dataset of turbofan engine degradation data show the superiority of our proposed framework over other relevant lifelong learning methods. On average, we achieve a performance improvement of 10.2% and 75.3% in terms of final prediction accuracy and forgetting measure, respectively.
KW - Bayesian neural networks
KW - deep learning
KW - lifelong learning
KW - Predictive maintenance
KW - prognostics and health management
KW - RUL prediction
UR - http://www.scopus.com/inward/record.url?scp=105023657829&partnerID=8YFLogxK
U2 - 10.1109/QRS-C65679.2025.00066
DO - 10.1109/QRS-C65679.2025.00066
M3 - Conference contribution
AN - SCOPUS:105023657829
SN - 978-1-6654-7774-1
T3 - Proceedings - International Conference on Software Quality, Reliability and Security Companion
SP - 483
EP - 491
BT - Proceedings - 2025 25th International Conference on Software Quality, Reliability and Security Companion, QRS-C 2025
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 25th International Conference on Software Quality, Reliability and Security Companion, QRS-C 2025
Y2 - 16 July 2025 through 20 July 2025
ER -