Water Quality Prediction Implemented a Data Acquisition and Machine Learning System

Drinking water is a public health issue; its availability is threatened by factors such as population growth, pollution, and climate change. Sustainable management of water resources requires protecting bodies of water, specifically those that supply drinking water to the population. The monitoring and analysis of water quality by implementing a data acquisition system for the collection of data that is analyzed through machine learning allows efficient supervision in response to anomalies that may occur. The data is processed with machine learning algorithms to evaluate variables and make informed decisions. In drinking water treatment plants, machine learning models monitor and diagnose in real time, identifying anomalies in the data. The research proposes data collection through a data acquisition system with the use of sensors that measure temperature, turbidity, conductivity, total dissolved solids (TDS) and pH. To process the data, a methodology is implemented where different models are analyzed. Such as linear regression, random forest and k-nearest neighbors (KNN), allowing the data to be analyzed with better precision and efficient decision-making regarding the quality of the data water. The model that showed the best results is Random Forest, it provided the best performance among the evaluated models.