IoT-Based Real-Time Water Quality Assessment with Predictive Analytics

Over a billion people worldwide lack access to clean water, and water scarcity continues to pose a growing global challenge. In recent decades, water quality has been affected due to contamination and pollution, creating serious risks to public health and ecosystems. Water Quality Indices (WQIs) help condense complex water data into a single score for easier evaluation. In this study, we used a USGS dataset containing water temperature, pH, turbidity, and dissolved oxygen to compute WQI using AI/ML algorithms. The dataset was preprocessed, and missing values were imputed using the K-Nearest Neighbors (KNN) method. Classification was performed using the derived WQI to determine Water Quality Classification (WQC) levels. Prediction results were validated using Brown’s method to calculate the Weighted Arithmetic Water Quality Index (WAWQI). The proposed AI/ML-based model supports real-time water quality assessment and enhances decision-making in water resource management.