MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641058686 A) filed by Agni College Of Technology on May 08, 2026, for Machine Learning Driven Gas Sensor Network For On-Site Agrochemical Hazardous Scent Detection.
Inventors include S. Manoranjitham; Vaheetha Banu S; M. Udith Yashica; and L. Jai.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: The rapid increase in environmental pollution and the need for safety in agricultural and industrial environments have created a demand for efficient gas monitoring systems. This project presents a real-time gas-detection system using the Internet of Things (IoT) and machine-learning techniques to monitor and analyze gas concentrations effectively. The system is built using an MQ135 gas sensor, which detects gas concentration and generates analog signals based on the presence of gases such as ammonia and other harmful pollutants. These signals are processed using the ESP8266 (NodeMCU) microcontroller, which converts the analog values into digital data and transmits them to the ThingSpeak cloud platform through a WiFi connection. This enables continuous data collection, storage, and remote monitoring. A Python-based system is used to retrieve the sensor data from the cloud using API calls. The data is then processed and analyzed using Support Vector Regression (SVR), a machine learning algorithm capable of handling non-linear data and providing accurate predictions. Based on the sensor values and predicted outputs, the system identifies the type of gas condition, such as urea, potassium, or mixed substances, using predefined ranges. The system also provides real-time graphical visualization of gas concentration using Matplotlib, allowing users to observe changes over time. Additionally, an alert mechanism is implemented to notify users through sound signals based on gas intensity levels. This ensures timely detection and response to potentially harmful conditions. The proposed system is cost-effective, easy to implement, and scalable. It improves traditional gas detection methods by integrating real-time monitoring, cloudbased data handling, and intelligent prediction. The system can be effectively used in applications such as environmental monitoring, agricultural safety, and industrial gas detection.
Disclaimer: Curated by HT Syndication.
