MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641025990 A) filed by Srinivasan P, Nagercoil, Tamil Nadu, on March 5, for 'smart gas leakage monitoring for mining safety using machine learning assisted optimization.'
Inventor(s) include Srinivasan P; and T. Keerthi Nethra.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "Gas leakage is a problematic safety aspect in both residential and industrial settings, hence the gas leakage detection must occur with high accuracy in real time. The important aspect of this research is to develop a smart gas leakage detection system by incorporating gas detection with machine learning hybrid with different optimization algorithm. The acquired gas concentration data from sensors are processed through an Arduino UNO microcontroller and analyzed on an optimized decision ML model. The GWO - PSO algorithm with hybrid to random forest and XGBoost dynamically optimizes the thresholds of detection and system parameters, hence improving the accuracy and reducing the false alarm probabilities. Such Hybrid model plays well in the improvement of accuracy as well as the model. This is also not just based on normal threshold, but on decision making after training the model perfectly with the dataset and feeding it all the patterns. Upon detecting abnom1al levels of gas concentration, the activated mechanisms include a local alert and sending notifications via wireless communication modules for remote monitoring. Experimental results prove that the response time, accuracy of detection, and reliability of the proposed system are better than conventional approaches that use only fixed thresholds, and, therefore, this work suits well for real-time applications intended for safety purposes."
Disclaimer: Curated by HT Syndication.
