MUMBAI, India, Feb. 27 -- Intellectual Property India has published a patent application (202641017454 A) filed by Shri Vishnu Engineering College For Women, Bhimavaram, Andhra Pradesh, on Feb. 17, for 'an iot-enabled real-time water quality prediction and early warning system using multi-model machine learning framework.'
Inventor(s) include Ms. R Sarada; Ms. P Archana; P. Suvarchala; and P. Srinidhi.
The application for the patent was published on Feb. 27, under issue no. 09/2026.
According to the abstract released by the Intellectual Property India: "Water quality is a fundamental aspect of environmental health, directly affecting ecosystems, human health, and economic development. Traditional methods of water quality assessment often rely on periodic sampling and laboratory analysis, which are time-consuming, expensive, and may not provide real-time insights. This research investigates the use of machine learning (ML) to predict water quality using real-time data collected from IoT sensors. In this project, we aimed to develop a predictive model for assessing water potability using machine learning techniques. The dataset utilized for this analysis contains various physicochemical properties of water samples, along with a binary target variable indicating potability. Initially, we performed data preprocessing, which included loading the dataset, handling missing values, and splitting the data into features and target variables. A Random Forest Classifier was employed to identify the most significant features influencing water potability, leading to the selection of the top four features based on their importance scores. These features were then visualized to understand their distribution concerning the target variable. To enhance the interpretability of the dataset, we created binary features based on predefined threshold values for critical water quality parameters, such as pH, hardness, chloramines, and sulfate levels. This transformation allowed us to simplify the classification task. Subsequently, we trained and evaluated multiple classification models, including Logistic Regression, Gaussian Naive Bayes, Support Vector Classifier, and K-Nearest Neighbors. Each model's performance was rigorously assessed using metrics such as accuracy, precision, recall, specificity, and F1 score. The results indicated varying levels of effectiveness among the models, with Logistic Regression achieving the highest accuracy. The findings from this project not only highlight the importance of specific water quality parameters in determining potability but also demonstrate the potential of machine learning techniques in environmental monitoring and public health. The models developed can serve as a valuable tool for water quality assessment, aiding in the identification of unsafe drinking water sources and promoting better water management practices."
Disclaimer: Curated by HT Syndication.
