MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202611090826 A) filed by Dinesh Kumar; and Nandini Tomar on July 26, 2026, for System And Method For Automatic Regulation Of Irrigation Valves Based On Root-Zone Soil Moisture Prediction Using Machine Learning In A Piped Micro-Irrigation Network.
Inventors include Dinesh Kumar; Nandini Tomar; Mukesh Kumar Bhardwaj; Ankita Sharma; and Priyanka Asthana.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: An automatic irrigation regulation system and method are disclosed, comprising one or more solar-powered soil probe modules disposed at the root zone of a crop, each comprising a capacitive soil moisture sensor, humidity and temperature sensors, and a bi-stable solenoid valve, in wireless communication with a compute module. The compute module executes a Long Short-Term Memory (LSTM) machine learning model, pre-trained on multi-year historical soil moisture, humidity, temperature and precipitation data and continually adapted using real-time site-specific sensor data, to predict crop water requirement and to autonomously actuate the solenoid valve(s) of a piped micro-irrigation network without manual intervention. A companion web application enables a user to monitor field conditions, locate deployed hardware, and estimate deployment cost. The system is designed to reduce water wastage, eliminate manual irrigation scheduling, and improve crop yield, while remaining affordable and deployable in remote, unelectrified agricultural fields.
Disclaimer: Curated by HT Syndication.
