MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202611088834 A) filed by Uttaranchal University on July 21, 2026, for An Iot-Based Machine Learning System For Real-Time Flood Prediction And Early Warning..
Inventors include Yashika; Yashita; and Rahul Mahala.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: An IoT-based machine learning system and associated process deploy an integrated edge-cloud hybrid computing architecture to achieve automated, non-invasive, real- time flood risk evaluation and proactive early warning dissemination. The physical architecture comprises a plurality of geographically distributed field sensing nodes, each incorporating an edge processing unit coupled to a multi-parameter hydrological and meteorological sensor array that includes an ultrasonic water level sensor, a hydrostatic pressure cell, a tipping bucket rain gauge, a Doppler flow sensor, and a soil moisture probe. Each field node features a global positioning system (GPS) module for fixed spatial geo-tagging, a real-time clock (RTC) module for precise temporal synchronization, and a backup micro-SD card storage block managed by an automated edge routing script to buffer data frames locally during telecommunication network dropouts. Cleaned and formatted data packets are transmitted via an integrated low-power LoRa transceiver to a central gateway interface, which routes the compiled streams across a secondary Wi-Fi module to a remote cloud computing platform. The cloud platform executes multi-channel pre-processing routines, including interpolation for missing data, feature scaling, trend analysis, and sensor fusion algorithms, to output a unified time-series dataset.
Disclaimer: Curated by HT Syndication.
