Search papers, labs, and topics across Lattice.
This study explores the use of liquid neural networks as latent dynamics models for aircraft engine health monitoring, addressing the limitations of traditional multivariate time-series models that fail to reveal coherent degradation processes. By factorizing the latent state into degradation and condition components, the model effectively separates health evolution from operating-condition variations, leading to improved sensor forecasting accuracy on the C-MAPSS benchmark. The results show a significant reduction in RMSE compared to a GRU baseline, particularly in multi-condition scenarios, while also providing a clearer temporal degradation axis for interpretability.
Liquid neural networks can enhance aircraft engine health monitoring by disentangling degradation from operating conditions, improving forecasting accuracy and interpretability.
Multivariate time-series models for prognostics are often evaluated by point prediction accuracy, yet their internal states rarely expose a coherent degradation process. We study liquid neural networks as latent dynamics models for aircraft engine health monitoring on the C-MAPSS benchmark. The proposed model encodes a history window into a latent state, evolves that state with a liquid transition model, and decodes future sensor observations. To separate health evolution from operating-condition variation, the latent state is factorized into degradation and condition components. Remaining useful life, monotonic risk, and latent-consistency losses supervise the degradation component, while condition prediction and decorrelation losses discourage operating-condition leakage. Across FD001--FD004, the full disentangled model improves overall sensor forecasting RMSE from 0.2438 for a GRU baseline to 0.2266, with the largest gains on the multi-condition subsets FD002 and FD004. The learned degradation state also forms a clearer temporal degradation axis, reaching an average state-speed Spearman correlation of 0.5960. Direct remaining-useful-life regression remains stronger for the GRU baseline, indicating that the proposed representation is currently more effective as an interpretable world model for degradation dynamics than as a calibrated lifetime regressor. These results suggest that liquid latent dynamics can bridge predictive maintenance forecasting and inspectable health-state modeling.
