Water Quality Index Based Prediction of Ground Water Properties for Safe Consumption

Abstract

Water as one crucial element for the survival of human beings is necessary to be handled with care. With a 59% of the Sri Lankan population depending themselves on this indispensable element, the authorities and the people must heed the importance of safe consumption to avoid severe consequences like Chronic Kidney Disease (CKD). In an attempt to address this social predicament, a smart device was fashioned as an initiative to predict safe consumption of groundwater and, to oblige and upskill the users to identify the quality of a groundwater sample in real-time. With the inclusion of machine learning techniques, the implementation was done by predicting the Water Quality Index (WQI) which is a single numeric index that mirrors the overall quality of any water sample, with an accuracy of 97.82 % . In addition, two more serviceable functionalities to predict possibilities of CKD outbreak and forecasting water quality parameters were also implemented with accuracies of 76.99% and 92% respectively. The sole of this research relies on the hardware device that embeds a set of sensors which accompanies the individual functionalities. The readings and outputs will be displayed through the mobile application which is real-time and of high performance with a friendly user-interface.