8 June 2026 in SSRN Electronic Journal
Artificial Intelligence (AI)-based health indicators are crucial for reliable condition monitoring of rotating machinery. However, existing unsupervised anomaly detection methods commonly face challenges in accurately modeling the vibration signal distribution and setting robust detection thresholds. To overcome these limitations, this paper proposes a novel gear fault detection approach based on Generative Adversarial Networks (GANs), specifically employing the fast AnoGAN (f-AnoGAN) structure. The method effectively models vibration signals from healthy operating conditions and identifies anomalies based on deviations from this learned distribution. Additionally, a statistically driven threshold-setting procedure using the anomaly scores of healthy data is introduced, enhancing detection robustness and reliability. The proposed methodology is validated using experimental data from a gear accelerated degradation test, demonstrating superior performance in detecting gear pitting at an earlier stage and with greater consistency compared to conventional unsupervised detection methods.
16 January 2024 in International Journal of Prognostics and Health Management
Gear pitting is a common gearbox failure mode that can lead to unplanned machine downtime, inefficient power transmission and a higher risk of sudden catastrophic failure. Consequently, there is strong incentive to create machine learning models that are capable of detecting and quantifying the severity of gearbox pitting faults. The performance of machine learning models is however highly dependent on the availability of training data and since training data for a wide variety of different operating conditions and fault severities is rarely available in practice, machine learning models must be designed to be robust to unseen operating conditions and fault severities. Furthermore, models should be capable of identifying data outside of the training data distribution and adjusting the confidence in a prediction accordingly. This work presents a strategy for pitting severity estimation in gearboxes under unseen operating conditions and fault severities in response to the PHM North America 2023 Conference Data Challenge. The strategy includes the design of dedicated validation sets for quantifying model performance on unseen data, an investigation into the most appropriate preprocessing methods, and a specialized convolutional neural network with an integrated out of distribution detection model for identifying samples from foreign operating conditions and fault severities. The results show that the best models are capable of some generalization to unseen operating conditions, but the generalization to unseen pitting severities is more challenging.
28 September 2026 in Annual Conference of the PHM Society
Gearbox failures are a major contributor to the levelized cost of electricity of wind turbines, owing to their long repair time and large share of total turbine downtime. Condition monitoring enables early fault detection, reducing unplanned downtime and the associated economic impact, while digital twins provide virtual representations of physical assets that support advanced diagnostics. This work couples the two: a phenomenological gear vibration model generates a synthetic dataset spanning healthy operation and several gear pitting severities, and healthy measurements from the real asset are used to estimate a transfer function, through cepstrum liftering, that is applied to the synthetic signals so that they better reproduce the experimental response. By stochastically varying the model parameters within prescribed ranges, a labeled synthetic dataset is produced and used to train a deep learning unsupervised domain adaptation (UDA)networkthatcouples adversarial training with entropy-based sample weighting and an information maximization loss, reducing the mismatch between simulated and measured data and improving diagnostic robustness on real measurements. The framework is validated on vibration data from an in-house back-to-back, single-stage helical gear test rig, in which pitting was artificially introduced on the tooth flank. Against state-of-the-art (SOTA) adversarial methods and a source-only baseline, the proposed approach more accurately distinguishes healthy operation from multiple pitting severity levels, supporting its potential for diagnosis on real applications.
13 January 2026 in PHM Society Asia-Pacific Conference
This paper investigates the diagnostic capabilities of rotary encoders for condition monitoring of commercial gearboxes, focusing on early detection of gear pitting. Within a multi-sensor framework, high-resolution rotary encoder data were collected alongside accelerometer signals, both mounted on a multistage commercial gearbox subjected to gear pitting. Encoder-based health indicators were developed from the square envelope (SE) of the transmission error (TE), synchronously averaged (SA) to isolate the targeted gear component within the gearbox. By automatically capturing images of the targeted gear during an accelerated life test, these indicators were evaluated for their ability to detect the onset of gear pitting and compared against conventional vibration-based metrics. Results show that encoder-based indicators can provide earlier and more consistent detection of gear pitting. These findings highlight the potential of rotary encoders as a complementary or standalone sensing solution in advanced diagnostic frameworks for commercial gearboxes.
19 June 2026 in SSRN Electronic Journal
Accurate remaining useful life (RUL) estimation of rolling element bearings (REBs) is essential for predictive maintenance but remains challenging due to the lack of physically interpretable degradation indicators, such as ones that can be directly connected with the fault size. Conventional approaches typically rely on data-driven trends of vibration-based features, whose relationship with the actual defect evolution is often unclear, limiting interpretability and robustness.This paper proposes a unified framework for online fault-size tracking and RUL prediction of rolling-element bearings from vibration measurements. The method exploits the periodic variation in harmonic amplitudes predicted by a double-impulse model of defect entry/exit impacts, enabling direct estimation of defect size, for both outer-race fault and inner-race fault. A reference-harmonic segmented accumulation strategy is introduced to fuse multi-harmonic information while mitigating phase mismatches and improving measurement reliability. The extracted fault-size growth events serve as observations to an Extended Kalman filter (EKF) governed by a Paris-law-based degradation model, enabling joint estimation of both fault size and model parameters. Finally, Monte Carlo (MC) sampling of the EKF posterior distribution provides probabilistic RUL predictions with physically interpretable uncertainty bounds.Measurement campaigns on the KU Leuven Bearing Prognostics Setup (KUL BPS) include both constant-profile (CP) and varying-profile (VP) campaigns. Both CP and VP cases validate the double-impulse simulation model and confirm the predicted periodic relationship between harmonic amplitudes and defect size. The proposed framework for online fault-size tracking and RUL estimation, where RUL predictions are derived directly from the estimated defect size rather than from abstract statistical features, are further demonstrated at the VP cases. Besides, an inner-race fault bearing damaged on the KU Leuven Gearbox Prognostics \& Diagnostics High-Power 2-Stage setup (KUL GPD-HP-2S) further validates the inner-race fault case. The results show promising accuracy and improved interpretability, as the predicted RUL is explicitly linked to the underlying physical fault evolution.