Intellectual Property India Publishes Patent Application for A System And Method For Blockchain-Verified Federated Graph Neural Network Fault Prediction And Lyapunov-Optimized Self-Healing Resource Allocation In Industrial Internet-Of-Things Networks

MUMBAI, India, Oct. 12 -- Intellectual Property India has published a patent application (202641117760 A) filed by Siddharth Institute Of Engineering & Technology on October 01, 2026, for A System And Method For Blockchain-Verified Federated Graph Neural Network Fault Prediction And Lyapunov-Optimized Self-Healing Resource Allocation In Industrial Internet-Of-Things Networks.

Inventors include Dr. E. Kartheek; Dr. K. K. Srimitra; Soundharya V; Rpvg Ashok Reddy; Avilala Rajasekhar Yadav; Dr. N. Chinna Alluraiah; Kunde Santhosh Kumar; M Kezia Rani; Dr. P. Karthikeyan; and Dr. S. Kannappan.

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

Abstract: A system for predictive fault detection and self-healing resource allocation in an industrial Internet-of-Things network comprises a plurality of edge sensor nodes, each sensing physical operating parameters of a respective industrial asset, constructing a local sensor-topology graph, and training a local graph neural network without transmitting raw sensed data. Hashed local model updates are submitted to a permissioned blockchain ledger, which performs consensus-based validation and records validated updates. A federated aggregator combines validated updates into a global graph neural network model, broadcast to the edge nodes, and used to compute a fault-prediction confidence score for each asset. When a confidence score exceeds a preset threshold, a smart contract automatically triggers a resource-allocation module, which computes a self-healing resource-reallocation decision using a Lyapunov drift-plus-penalty optimisation framework that bounds a virtual queue backlog representing accumulated risk-weighted load, and applies the decision to redistribute operational load away from the at-risk asset. The reallocation is logged on the blockchain ledger and displayed on a monitoring dashboard, providing a mathematically provable, closed-loop integration of electronic sensing, communication, distributed computing, artificial intelligence, and optimisation.

Disclaimer: Curated by HT Syndication.