MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641051514 A) filed by S. D. Anushna; Sriramakavacham Prudhvi Raj; Hemanthakumar Kappali; Krishnappa H K; and Dr. R. Mary Jeya Jothi, Hyderabad, Telangana, on April 22, for 'privacy-preserving digital twin framework for federated learning-enabled predictive irrigation scheduling via multi-source soil moisture iot networks and satellite remote sensing data fusion.'
Inventor(s) include S. D. Anushna; Sriramakavacham Prudhvi Raj; Hemanthakumar Kappali; Krishnappa H K; and Dr. R. Mary Jeya Jothi.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "The invention provides a novel Privacy-Preserving Digital Twin Framework for Federated Learning-Based Predictive Irrigation Scheduling that intelligently fuses data from Soil Moisture IoT Networks and Satellite Remote Sensing. Traditional irrigation methods suffer from inefficiency, water wastage, and lack of adaptability due to fragmented data sources and privacy barriers in collaborative farming. The proposed system deploys dense IoT sensor networks for real-time, high-resolution soil moisture, temperature, and related parameters. These are fused with satellite remote sensing data (including vegetation indices, evapotranspiration, and broad-area soil moisture estimates) using advanced multi-modal fusion techniques to create accurate, comprehensive field representations. A Federated Learning paradigm allows multiple farms to collaboratively train predictive models on their local data without exposing sensitive raw information, with only model updates shared and aggregated centrally. This global knowledge enhances local predictions while preserving privacy. The centerpiece is the Digital Twin-a dynamic virtual model of the physical farmland that mirrors real-time conditions, simulates future soil moisture states and crop responses using hybrid physics-ML models, and optimizes irrigation schedules through predictive analytics and scenario testing. Recommendations are delivered to automated actuators for precise, zone-specific water application. The framework addresses key challenges in precision agriculture by enabling scalable, privacy-aware collaboration, reducing water consumption significantly (demonstrated 25-40% savings in case studies), improving crop yields, and supporting sustainable practices amid climate variability. It offers robustness, edge deployment for low-latency decisions, and extensibility to other agronomic parameters. This integrated cyber-physical solution represents a significant advancement over existing centralized or non-collaborative smart irrigation technologies."
Disclaimer: Curated by HT Syndication.
