Water Quality Index Prediction System Using IoT Enabled Sensor Suite and Machine Learning

There is an increase in the necessity to monitor the water quality of different water resources as nowadays, clean drinking water is not in abundance and water quality needs to be known beforehand to prevent the occurrence of diseases associated with consuming unfit water. This paper presents an approach to determining the water quality index (WQI) of a water sample by measuring the essential water quality parameters such as pH, dissolved oxygen (DO), electrical conductivity (EC), total dissolved solids (TDS), turbidity, and temperature. For calculating and predicting the WQI, the Canadian water quality index (CWQI) calculation is used in this work. An IoT and cloud-based solution module that can be used to access the acquired sensor data and make preliminary conclusions as well as the quality of water is developed and presented. Furthermore, machine learning is used to predict the WQI of the samples. Conclusive results on water quality are decided based on the safe ranges outlined by the World Health Organization (WHO) and CPCB that outline the range for safe water. The training validation and prediction results of the machine learning models such as long short-term memory models (LSTMs) and neural networks are presented and the CWQI calculated for the water samples collected and tested are presented and discussed.