Prediction model for drinking water network failures

Prediction model for drinking water network failures The large amount of data nowadays available for the companies creates wide business opportunities. Data Science meets application in different domains. Data science processes leverage internal or external resources, such as big data, from which useful insights and predictive models can be obtained, to create value for the company through the significant support that can be provided to the decision making through these data analytics tools. This research traces the foundations of the Data Science approach, Business Intelligence and Big Data. In this context, is placed the PME Project, a Data Science project that proposes a solution for the management of pipeline failures in the drinking water networks. In particular, this project is based on the search for a predictive model of failures in a drinking water network, a study that will allow a better understanding of the complexity of the network management. However, the research focuses on predicting and its purpose is to support decision making in the context of renovations and repairs. In this way, the aim is to reduce the uncertainty about network management, reduce costs related to emergency purchases, avoid fines and improve six-monthly renewal planning.