MUMBAI, India, May 29 -- Intellectual Property India has published a patent application (202621036015 A) filed by Bhise Suhas Balkrishna; Bhosale Sanskruti Vinod; Biradar Vedant Venkatesh; Bundhe Prajwal Dipak; and Butala Manthan Shashank, Pune, Maharashtra, on March 25, for 'forecasting future water requirements and assessing storage capacities in reservoirs.'
Inventor(s) include Bhise Suhas Balkrishna; Bhosale Sanskruti Vinod; Biradar Vedant Venkatesh; Bundhe Prajwal Dipak; and Butala Manthan Shashank.
The application for the patent was published on May 29, under issue no. 22/2026.
According to the abstract released by the Intellectual Property India: "The present invention discloses a multi-stage, hybrid artificial intelligence computing system for the real-time forecasting and analysis of hydrological reservoir storage. The system overcomes the limitations of single-model paradigms and "black-box" deep learning by integrating a parallel processing architecture. Multimodal environmental data is simultaneously evaluated by a SARIMAX statistical engine, an XGBoost machine learning model, an LSTM recurrent network, and a Multi-Head Attention Transformer. A specialized Hybrid Aggregator temporally aligns and synthesizes these distinct outputs to generate a highly robust ensemble forecast. To ensure operational transparency for water resource management, the system features an integrated Explainable AI (XAI) layer that utilizes SHAP values to quantify and visually map the exact physical drivers behind each prediction. This creates a forecasting tool that is simultaneously highly accurate, resilient to non-linear climate shocks, and physically interpretable."
Disclaimer: Curated by HT Syndication.
