Liquid Latent State Dynamics for Interpretable Turbofan Degradation Modeling

2026-07-02

AI Summary

This work addresses the limited interpretability of existing degradation modeling approaches for turbofan engines, despite their high prediction accuracy. The authors propose a latent-variable dynamic model based on liquid neural networks that explicitly disentangles degradation-related and operating-condition-related factors by decoupling latent states. A multi-task loss function—incorporating remaining useful life prediction, monotonic risk constraints, operating condition forecasting, and a decorrelation term—is designed to guide the learning process. Evaluated on the C-MAPSS dataset, the method reduces sensor prediction RMSE from 0.2438 to 0.2266 and achieves a Spearman correlation of 0.5960 for temporal consistency of the inferred degradation state, thereby significantly enhancing both model interpretability and dynamic coherence.

Abstract

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.

Problem

• degradation modeling
• interpretable latent state
• turbofan health monitoring
• remaining useful life
• disentangled representation

Innovation

• liquid neural networks
• latent dynamics
• disentangled representation
• degradation modeling
• prognostics