MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202621075915 A) filed by Symbiosis International Deemed University on June 18, 2026, for An Intelligent Early Warning System For Earthquake Detection Using Mems Accelerometers And Ensemble Machine Learning.
Inventors include Dr. Snehlata Wankhade; Gargi Balamwar; Karan Misal; and Anshuman Padole.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: ABSTRACT AN INTELLIGENT EARLY WARNING SYSTEM FOR EARTHQUAKE DETECTION USING MEMS ACCELEROMETERS AND ENSEMBLE MACHINE LEARNING The present invention discloses an intelligent early warning system (100) for earthquake detection comprising a plurality of smart MEMS accelerometer sensor nodes (110) integrated within a 5-layer IoT architecture. The system (100) includes a perception layer with MEMS sensor nodes (110) for ground vibration sensing with on-node digital filtering (114) and STA/LTA trigger logic (116), a network layer (120) with MQTT gateways (122) and time synchronization (124), a processing layer (130) housing a machine learning engine (132) running Random Forest and SVM classifiers in parallel for multi-class alert classification and a multi-sensor voting and validation engine (134) for false alarm rejection, an application layer (140) triggering emergency responses via SCADA automation (142), public notifications (144), and authority alerts (146), and a business layer (150) for system governance. The system achieves 92.69 percent overall accuracy with 98 percent Red Alert recall, providing a scalable alternative to conventional seismographic infrastructure. [
Disclaimer: Curated by HT Syndication.
