MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641118568 A) filed by Dr. Bhagavant K Deshpande; Ms. Reethiga M; Neerajaveeramachineni; Dr. S K Hiremath; Mr. Kedar Hiremath; Dr. E. Rajesh Kumar; and Dr. C. R. Rene Robin on October 04, 2026, for System For Machine Learning-Based Predictive Maintenance Of Computing Infrastructure.

Inventors include Dr. Bhagavant K Deshpande; Ms. Reethiga M; Neerajaveeramachineni; Dr. S K Hiremath; Mr. Kedar Hiremath; Dr. E. Rajesh Kumar; and Dr. C. R. Rene Robin.

The application for the patent was published on October 09, 2026, under issue no. 41/2026.

Abstract: A system for machine learning-based predictive maintenance of computing infrastructure is disclosed. Heterogeneous telemetry from computing resources is synchronized and normalized while retaining data-quality information. An infrastructure graph represents physical, logical and service dependencies, and a health-state engine generates a topology-aware temporal health representation combining local temporal features, dependency-related features and telemetry-quality attributes. A machine-learning prediction engine determines a predicted failure mode, failure likelihood, predicted failure horizon and uncertainty value. A confidence and risk fusion engine combines the prediction with operational context including service criticality, redundancy, dependency impact and intervention burden. A maintenance orchestrator selects an action under operational policy constraints and may employ staged diagnostic and preventive actions according to prediction uncertainty and the available failure horizon. Post-action telemetry is compared with a pre-action baseline and predicted degradation signature to verify intervention effectiveness. Verified outcomes adapt model weighting, calibration, thresholds or action-selection parameters, thereby providing a closed-loop predictive maintenance architecture resilient to changing workloads and infrastructure configurations. Accompanied Drawing [FIGS. 1-2].

Disclaimer: Curated by HT Syndication.