Water, an essential resource for human survival, must be secured and efficiently used in concert with sustainable development to give a future generation without scarcity. Even though the usage is done efficiently, the stage in which the water reaches mankind from the source experiences some leakages in the system causing the water to get wasted even before reaching the destination. The optimum use of developed technologies and human efforts can cause a decline in the wastage of water. The manual intervention of detecting the leakage is done by analysing the sound of the water leakage area through various devices and replacing the parts with suitable parts. But this traditional way of detecting the leakage causes manual assistance and is comprised of heavy workloads causing high time consumption. Instead, the detection can be done automatically with minimal human assistance reducing time and manual workload. This paper provides a solution to one such extent by enabling the automatic detection of water leakage via suitable microphones as sensor input. Then processing the sound signal for converting the data from the time domain to the frequency domain. The processed data is given to the Machine Learning (ML) model for identifying the water leakage. Three different ML models including KNN (K Nearest Neighbour), SVM (Support Vector Machine), and Random Forest (RF) are employed. Among the three ML models, the best one is identified using the metrics. Finally, the best model is deployed in the mobile app which makes the water leakage detection process simple.

Water leakage, Sensor, Machine Learning, Mobile Application, Accuracy