A Comprehensive Analysis of Machine Learning: Water Leakage Prediction

Water scarcity is one of the most pressing issues facing the world today, and highlights the necessity for proper water management systems to be implemented. In urban areas, undetected and unnoticed water leakage in the aging pipeline networks (which is a very common occurrence in cities), is often one of the largest contributors to water loss in a city, and typically occurs as the infrastructure of the city is significantly lagging behind contemporary technological standards. The project focus is to design, using Arduino technology, an automated water leakage detection system. The system will detect water flow at the entrance and exit points using water flow sensors. If the difference between the readings exceeds the set parameters this will indicate a possible leakage. It can then notify the user of the possible leakage situation, and also optionally shut the water flow off using a solenoid valve. In order to enhance the functionality of the system, it has been integrated with the Blynk IoT platform that provides a mobile application for the user to visually observe real-time water flow. This enhancement of the system makes it ideal for use in smart homes and in municipal water systems with the key objective to reduce waste related to infrastructure and build habits of sustainable consumption.