Details
| Original language | English |
|---|---|
| Pages (from-to) | 116-121 |
| Number of pages | 6 |
| Journal | Procedia CIRP |
| Volume | 118 |
| Early online date | 18 Jul 2023 |
| Publication status | Published - 2023 |
| Event | 16th CIRP Conference on Intelligent Computation in Manufacturing Engineering 2022 - Naples, Italy Duration: 13 Jul 2022 → 15 Jul 2022 |
Abstract
Reliably predicting Remaining Useful Life (RUL) is crucial for reducing asset maintenance costs. Deep learning emerges as a powerful data-driven method capable of predicting RUL based on historical operating data. However, standard deep learning tools typically do not account for the uncertainty inherent in prediction tasks. This paper presents an uncertainty-aware approach that predicts not only the RUL but also the associated confidence interval, capturing both aleatoric and epistemic uncertainty. The proposed approach is evaluated on publicly available datasets of aircraft turbofan engines, showing its ability to estimate accurate RUL and well-calibrated uncertainties that are robust to out-of-distribution data.
Keywords
- Deep learning, Hyper-deep ensemble, PHM systems, Predictive maintenance, RUL prediction, Uncertainty quantification
ASJC Scopus subject areas
- Engineering(all)
- Control and Systems Engineering
- Engineering(all)
- Industrial and Manufacturing Engineering
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Procedia CIRP, Vol. 118, 2023, p. 116-121.
Research output: Contribution to journal › Conference article › Research › peer review
}
TY - JOUR
T1 - Uncertainty-aware remaining useful life prediction for predictive maintenance using deep learning
AU - Xuan, Quy Le
AU - Adhisantoso, Yeremia G.
AU - Munderloh, Marco
AU - Ostermann, Jörn
N1 - Funding Information: This work was supported by the Federal Ministry for Economic Affairs and Climate Action (BMWK), Germany, within the framework of the project IIP-Ecosphere (project number 01MK20006A).
PY - 2023
Y1 - 2023
N2 - Reliably predicting Remaining Useful Life (RUL) is crucial for reducing asset maintenance costs. Deep learning emerges as a powerful data-driven method capable of predicting RUL based on historical operating data. However, standard deep learning tools typically do not account for the uncertainty inherent in prediction tasks. This paper presents an uncertainty-aware approach that predicts not only the RUL but also the associated confidence interval, capturing both aleatoric and epistemic uncertainty. The proposed approach is evaluated on publicly available datasets of aircraft turbofan engines, showing its ability to estimate accurate RUL and well-calibrated uncertainties that are robust to out-of-distribution data.
AB - Reliably predicting Remaining Useful Life (RUL) is crucial for reducing asset maintenance costs. Deep learning emerges as a powerful data-driven method capable of predicting RUL based on historical operating data. However, standard deep learning tools typically do not account for the uncertainty inherent in prediction tasks. This paper presents an uncertainty-aware approach that predicts not only the RUL but also the associated confidence interval, capturing both aleatoric and epistemic uncertainty. The proposed approach is evaluated on publicly available datasets of aircraft turbofan engines, showing its ability to estimate accurate RUL and well-calibrated uncertainties that are robust to out-of-distribution data.
KW - Deep learning
KW - Hyper-deep ensemble
KW - PHM systems
KW - Predictive maintenance
KW - RUL prediction
KW - Uncertainty quantification
UR - http://www.scopus.com/inward/record.url?scp=85173575575&partnerID=8YFLogxK
U2 - 10.1016/j.procir.2023.06.021
DO - 10.1016/j.procir.2023.06.021
M3 - Conference article
AN - SCOPUS:85173575575
VL - 118
SP - 116
EP - 121
JO - Procedia CIRP
JF - Procedia CIRP
SN - 2212-8271
T2 - 16th CIRP Conference on Intelligent Computation in Manufacturing Engineering 2022
Y2 - 13 July 2022 through 15 July 2022
ER -