MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202641071972 A) filed by Mr. Vaishak Kuchinad; Ms. Bindu M N; Dr. Jyoti Metan; Dr. Naveen Kumar B; Ms. Yogitha R; Dr. Ramesh Nuthakki; Ms. Sonia S B; Ms. R Malini; Dr. Vinoth Kumar S; and Ms. Asha M L on June 10, 2026, for An Iot Based Air Quality Prediction And Pollutant Source Identification Using Ai And Neural Networks.
Inventors include Mr. Vaishak Kuchinad; Ms. Bindu M N; Dr. Jyoti Metan; Dr. Naveen Kumar B; Ms. Yogitha R; Dr. Ramesh Nuthakki; Ms. Sonia S B; Ms. R Malini; Dr. Vinoth Kumar S; and Ms. Asha M L.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: AN IOT BASED AIR QUALITY PREDICTION AND POLLUTANT SOURCE IDENTIFICATION USING Al AND NEURAL NETWORKS ABSTRACT The present invention proposes an intelligent IoT-Based Air Quality Prediction and Pollutant Source Identification System Using Artificial Intelligence and Neural Networks for real-time environmental monitoring and analysis. The system integrates a network of IoT-enabled sensors to continuously collect atmospheric parameters, including particulate matter (PM2.5 and PM10), carbon monoxide (CO), nitrogen dioxide (NO2), sulfur dioxide (SO2), ozone (O3), temperature, humidity, and wind characteristics from multiple geographic locations. The acquired data are transmitted through wireless communication protocols to a cloud-based processing platform for storage and analysis. Advanced Artificial Intelligence (Al) algorithms and deep neural network models are employed to preprocess sensor data, remove noise, extract significant features, and predict future air quality levels with high accuracy. The proposed neural network architecture learns temporal and spatial pollution patterns to forecast the Air Quality Index (AQI) over shortterm and long-term periods. Additionally, a pollutant source identification module utilizes machine learning classification techniques and pattern recognition mechanisms to determine the probable origins o f pollution, such as vehicular emissions, industrial activities, construction sites, biomass burning, and natural sources. The system further generates automated alerts and recommendations when predicted pollution levels exceed predefined thresholds, enabling authorities and citizens to take preventive actions. By combining IoT sensing, cloud computing, Al-driven prediction, and neural network-based source attribution, the invention enhances environmental surveillance, supports smart city initiatives, improves public health protection, and facilitates data-driven air quality management. The proposed framework offers scalable deployment, real-time operation, high prediction accuracy, and efficient pollutant source tracking for sustainable urban and industrial environments.
Disclaimer: Curated by HT Syndication.
