MUMBAI, India, Aug. 31 -- Intellectual Property India has published a patent application (202641100352 A) filed by Vignans Foundation For Science, Technology And Research on August 19, 2026, for A Real-Time Ai-Driven Seismic Alert Framework For Earthquake Detection.
Inventors include Dr. K. Santhi Sri; Mr. Irfan Sayyad; Mr. Peddipaga Rahul; and Mr. K. Raghavendra Murali.
The application for the patent was published on August 28, 2026, under issue no. 35/2026.
Abstract: ABSTRACT [0018] Earthquakes strike without warning and can cause extensive loss of life and damage. Conventional early-warning systems rely on threshold rules and dense sensor networks that often incur high cost, latency and false alarms. The present invention discloses a real-time AI-driven seismic alert framework that employs a lightweight Random Forest classifier to distinguish hazardous from non-hazardous seismic events. The model analyses four primary parameters earthquake magnitude, focal depth, seismic frequency and ground acceleration—after synthetic data generation, pre-processing and feature extraction. Once trained, the classifier operates with low computational demand and produces an alert status together with the underlying parameter values. Experimental evaluation yields an accuracy of 95.6 percent, precision of 94.3 percent, recall of 96.7 percent, an F1-score of 95.5 percent and a ROC-AUC of 0.98. Compared with deeper neural architectures the framework achieves competitive detection performance while offering greater interpretability and markedly lower resource requirements, making it suitable for earthquake-prone regions, smart-city platforms and IoT-enabled disaster-management systems.
Disclaimer: Curated by HT Syndication.
