MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202641094166 A) filed by Vardhaman College Of Engineering on August 04, 2026, for Multimodal Disaster Prediction System Using Satellite, Social Media, And Iot Data Fusion.
Inventors include Ms. S Uma Maheswari; Mr. Rajashekhar Panjala; Ms. Chinta Anusha; Ms. Sri Ranga Lakshmi M; Ms. Farhana Begam; and Dr. Kathirisetty Nikhila.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: Multimodal Disaster Prediction System Using Satellite, social media, and IoT Data Fusion are the proposed invention. The present invention provides a Multimodal Disaster Prediction System based on a Multimodal Graph Foundation Model (MGFM) combined with Spatio-Temporal Graph Neural Networks (ST-GNNs) and Retrieval-Augmented Reasoning (RAR) for real-time and accurate disaster prediction. The system collects heterogeneous data from satellite imagery, IoT sensors, meteorological data and geo-tagged social media platforms to build a dynamic disaster knowledge graph that reflects spatial, temporal and contextual relationships. The ST-GNN models the interactions of environment conditions, infrastructure, sensor networks, and affected regions to extract disaster propagation patterns, while the RAR module retrieves similar historical disaster events to improve the reliability of prediction and reduce false alarms. A transparent risk assessment engine estimates disaster probability, severity, impact zones, and infrastructure vulnerability with confidence scores to enable explainable decision making. An adaptive early warning module gives location-specific alerts, evacuation instructions and suggestions for emergency resources. The proposed invention enhances prediction accuracy, situational awareness, response efficiency, scalability and resilience for smart cities, disaster management agencies and critical infrastructure protection systems.
Disclaimer: Curated by HT Syndication.
