MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202621066234 A) filed by Dr. Naresh Dhakar; Dr. Ankit Kumar; Dr. Agurla Raju; Kunal Pratap Singh; Dr. Pipas Kumar; and Dr. Aarti Kamboj on May 26, 2026, for Machine Learning Framework For Predicting Pest Outbreaks In Indian Cropping System.
Inventors include Dr. Naresh Dhakar; Dr. Ankit Kumar; Dr. Agurla Raju; Kunal Pratap Singh; Dr. Pipas Kumar; and Dr. Aarti Kamboj.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: The present invention a machine learning framework for predicting pest outbreaks in Indian cropping systems, designated as system (100). System (100) comprises a multi-source data acquisition subsystem (110) integrating an IoT field sensor network (111), agro-meteorological data interface (112), satellite remote sensing platform (113), and historical pest surveillance database interface (114); a data preprocessing and feature engineering module (120) performing quality control, imputation, and temporal feature extraction; a multi-model ensemble prediction engine (130) combining gradient boosting (131), random forest (132), long short-term memory neural network (133), and attention-based temporal models (134) through a ridge regression meta-learner (135) to generate fourteen-day pest outbreak probability scores and population trajectories with quantified uncertainty; a geospatial risk mapping module (140) generating dynamic grid-referenced pest outbreak heatmaps using ordinary kriging interpolation; a real-time early warning and advisory dissemination subsystem (150) delivering personalised, multilingual alerts to farmers and extension services through mobile applications, and SMS.
Disclaimer: Curated by HT Syndication.
