Intellectual Property India Publishes Patent Application for An Explainable Federated Ensemble Learning System And Method For Remaining Useful Life Prediction Of Aircraft Engines

MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202611089105 A) filed by Manipal University Jaipur on July 22, 2026, for An Explainable Federated Ensemble Learning System And Method For Remaining Useful Life Prediction Of Aircraft Engines.

Inventor includes Dr. Bhawana Sharma.

The application for the patent was published on September 18, 2026, under issue no. 38/2026.

Abstract: The present invention relates to an explainable federated ensemble learning system and method for remaining useful life prediction of aircraft engines. The system comprises a dataset collection module configured to collect aircraft engine operational and multivariate sensor data from distributed data sources; a data processing module configured to normalize sensor data, remove redundant features, calculate RUL values, and generate time-series sequences; a client-side deep learning module configured to train multiple prediction models; a local ensemble module configured to combine model predictions; a federated learning module configured to aggregate model parameters at a central server; and an explainable ai module configured to interpret RUL predictions using SHAP-based feature-importance analysis. The results are then ensembled and transmitted to server where the global model shows the high performance and Explainable AI concept for providing reliable prediction SHAP method explains each feature effecting the RUL of turbofan engine and provides the important features to build trust in the AI model.

Disclaimer: Curated by HT Syndication.