MUMBAI, India, July 7 -- Intellectual Property India has published a patent application (202641078248 A) filed by Pinni Srinivasa Varma; K Sowjanya Bharathi; Dr. Boddu Sekhar Babu; and Koneru Lakshmaiah Education Foundation on June 24, 2026, for A Deep Learning-Based Framework For Robust Weather Forecasting And Climate Analysis.
Inventors include K Sowjanya Bharathi; and Dr. Boddu Sekhar Babu.
The application for the patent was published on July 03, 2026, under issue no. 27/2026.
Abstract: A deep learning-based framework for robust weather forecasting and climate analysis is disclosed. The framework collects and synchronises weather data from satellite imagery, IoT sensors and historical climate records; cleans and normalises the data; and processes it through a hybrid model in which a convolutional neural network extracts spatial features and an LSTM/GRU network models temporal dependencies. An attention mechanism fuses the spatial and temporal features, and a prediction engine generates accurate short- and medium-range forecasts together with climate-trend analysis. An adaptive learning mechanism re-trains the model on new data, and an extreme-weather detection module issues early alert. The framework improves accuracy, robustness to noisy or missing data, automation and scalability compared with conventional and single-source methods.
Disclaimer: Curated by HT Syndication.
