Sequence-to-sequence modeling for failure prognosis: Predicting the complete operating sequences of turbofan engines and their end-of-life

Increasing demands on modern machinery, such as turbofan engines in passenger aircraft, necessitate proper predictive maintenance strategies to ensure reliability. Traditional failure prognosis approaches primarily predict Remaining Useful Life (RUL), which can limit the understanding of engine degradation progression. This article proposes a sequence-to-sequence (seq2seq) machine learning model to predict complete degradation trajectories of turbofan engines, enabling implicit end-of-life (EoL) estimation. Complete degradation profiles provide in-depth insights into progression of engine health, and help identify critical points, such as rapid deterioration, or turning points in the degradation trend. The proposed model combines Long Short-Term Memory (LSTM) and Autoencoders (AE) to process multivariate time-series data and generate future degradation sequences in an autoregressive manner. The model is developed and validated on the FD001 subset of the C-MAPSS dataset, which contains engines operating under a single condition. By recursively generating future cycles, the model tracks degradation progression across sensors, and identifies the cycle where each sensor reaches its extreme value, marking the engines’ EoL.

Authors

Institutions

Publication Details

Journal
Proceedings of the Institution of Mechanical Engineers Part O Journal of Risk and Reliability
Published
2026-11-19
DOI
https://doi.org/10.1177/1748006x261507210
Primary Topic
Machine Fault Diagnosis Techniques
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Sequence-to-sequence modeling for failure prognosis: Predicting the complete operating sequences of turbofan engines and their end-of-life

Harun Taha Hayvaci, Halil Ibrahim Turan, Mohand Djeziri, Evangéline Bènevent et al.
Proceedings of the Institution of Mechanical Engineers Part O Journal of Risk and Reliability
Machine Fault Diagnosis Techniques
article

Sequence-to-sequence modeling for failure prognosis: Predicting the complete operating sequences of turbofan engines and their end-of-life

Harun Taha Hayvaci, Halil Ibrahim Turan, Mohand Djeziri, Evangéline Bènevent, Marc Bendahan
article en

Abstract

Increasing demands on modern machinery, such as turbofan engines in passenger aircraft, necessitate proper predictive maintenance strategies to ensure reliability. Traditional failure prognosis approaches primarily predict Remaining Useful Life (RUL), which can limit the understanding of engine degradation progression. This article proposes a sequence-to-sequence (seq2seq) machine learning model to predict complete degradation trajectories of turbofan engines, enabling implicit end-of-life (EoL) estimation. Complete degradation profiles provide in-depth insights into progression of engine health, and help identify critical points, such as rapid deterioration, or turning points in the degradation trend. The proposed model combines Long Short-Term Memory (LSTM) and Autoencoders (AE) to process multivariate time-series data and generate future degradation sequences in an autoregressive manner. The model is developed and validated on the FD001 subset of the C-MAPSS dataset, which contains engines operating under a single condition. By recursively generating future cycles, the model tracks degradation progression across sensors, and identifies the cycle where each sensor reaches its extreme value, marking the engines’ EoL.

Proceedings of the Institution of Mechanical Engineers Part O Journal of Risk and Reliability
Centre National de la Recherche Scientifique (FR), Aix-Marseille Université (FR), American University of the Middle East (KW), Institut des Matériaux, de Microélectronique et des Nanosciences de Provence (FR)
Machine Fault Diagnosis Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon