MUMBAI, India, Feb. 6 -- Intellectual Property India has published a patent application (202511134327 A) filed by Maharishi Markandeshwar, Ambala, Haryana, on Dec. 31, 2025, for 'a dynamic data fusion system for agriculture using iot and deep learning.'
Inventor(s) include Er. Ishta; and Dr. Vishal Bharti.
The application for the patent was published on Feb. 6, under issue no. 06/2026.
According to the abstract released by the Intellectual Property India: "The invention discloses a dynamic data fusion system for agriculture integrating IoT devices and deep learning models. Soil sensors, satellite imagery, and weather stations provide heterogeneous inputs synchronized through a novel time stamp mechanism that aggregates data every 15 minutes, eliminating inconsistencies. Edge computing modules filter anomalies and low-quality data, reducing bandwidth and energy consumption. Adaptive weighting prioritizes data streams based on environmental context, ensuring situation-specific predictions. Multimodal deep learning techniques fuse numerical, categorical, and visual data using CNNs, RNNs, and transformers, creating enriched datasets for accurate decision-making. Incremental learning enables continuous model updates without full retraining, while farmer feedback loops refine predictions in real time. The system delivers actionable recommendations for irrigation, fertilizer application, pest control, and yield prediction. Scalable and efficient, the invention provides a robust solution for precision agriculture, enhancing sustainability and productivity in resource-constrained environments."
Disclaimer: Curated by HT Syndication.
