Early Warning System for Water Infrastructures
This IoT use case was developed in collaboration with Saint John Water. Still active today are pipes reaching over 150+ years in age, and a data-driven approach is the most cost-effective method of monitoring and predicting leakage and bursts. Therefore, our attention was given in finding the most cost-effective method for monitoring and predicting leakage and bursts. We have analyzed the trends in the existing real-time water network data gathered by the city and have explored various machine learning algorithms to predict pipe bursts and leaks. The results have pointed out the need for the city of Saint John to retrieve time series data in order to be able to predict pipe bursts and leaks.