Big Data and AI

A considerable amount of data is created by security, network, and endpoint devices.  As an example, a typical oil platform has over thirty thousand sensors, which produces petabytes of data per day.  Forty percent of all data is never stored; and sixty percent of the data is stored locally off shore.  Only one percent of the data can be streamed for daily use onshore.  Data management becomes a problem, in that real time data cannot be analyzed.  The reporting is limited to a few metrics, which is monitored after the fact and maintenance is still only done on the equipment during recommended intervals by the manufacturer.  

Why is this a problem? It is a problem because real time data could potentially have an economic impact of three hundred billion dollars in savings with just equipment maintenance alone, and a further financial impact regarding the employee could result in thirty billion dollars in health and safety issues.