MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202641111589 A) filed by Rajesh Thonduru; Sneha Valusa; Rajasekhar Reddy Arikatla; Subba Nelakudhiti; and Sankethu Babu Guntaka on September 17, 2026, for A Hierarchical Graph Neural Network-Based Digital Twin System For Critical Infrastructure Management.
Inventors include Rajesh Thonduru; Sneha Valusa; Rajasekhar Reddy Arikatla; Subba Nelakudhiti; and Sankethu Babu Guntaka.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: The present invention provides the system combines a physical network layer that gathers real-time telemetry from various infrastructure domains including transportation, water distribution, power distribution, and communication networks with a specialized twin network layer responsible for overseeing network topologies, node-edge characteristics, security measures, and service mapping. A hierarchical graph neural network analyzes both static structural designs and dynamic operational flows across these interconnected sectors. Utilizing both graph-level and node-level architectures within the graph neural network framework, the system facilitates predictive maintenance, analyzes cascading failures across domains, detects anomalies, and optimizes operations in real time. Additionally, it integrates scalable graph processing and online learning frameworks to effectively handle extensive decentralized datasets while ensuring operational resilience in challenging conditions. FIG.1
Disclaimer: Curated by HT Syndication.
