Machine Learning as a field of computational analysis has been defined for decades, but has only blossomed in the last 5-10 years with the emerging abundance of available big data, connected computing power and standardized algorithms.
Machine Learning is a field of Artificial Intelligence that allows computers to find hidden insights “without being explicitly programmed where to look” – Arthur Samuel, 1959
Machine learning can use known data and learn from it to predict or classify unknown data. This approach can be used in various technical problems from image recognition, to recommendation engines, to detecting hidden insights within unstructured data.
Fracta delivers Likelihood of Failure results rapidly with three step deployment.
No one has clean or organized data. We’ll do the heavy lifting and assess and normalize utility data. Once we have water main pipe data in a geospatial data format and main break history, we’ll examine and process data gaps and deliver a complete data set that can be easily analyzed and visualized.
The cleaned and normalized data set gives an accurate picture of the current state of the asset and is the necessary foundation for the next phase of machine learning analysis.
With cleaned and normalized data, Fracta layers this data set with external geospatial information to form a machine learning model. There are hundreds of variables used in the machine learning model, resulting in a dynamic solution that is tailored to the utility and constantly evolving from new data points it is exposed to.
The output of the machine learning analysis is an accurate prediction for likelihood of water main failure for all the pipe segments of the utility.