The Sat-Net Oracle system is an innovative blend of advanced technologies for supporting Predictive Maintenance Failure analysis through the combination of Bluestaqs Unified Data Library (UDL) and Googles Artificial Intelligence engine TensorFlow.Leveraging the UDLs hybrid data store, (currently consuming over 2.5 million observation and telemetry data points per day), the Sat-Net Oracle would condition a Google AI neural network, training on the data feeds to establish a baseline for normalcy, and then begin to issue alerts to owner/operators that their equipment is beginning to behave anomalously.Anomolous behavior is often an indicator of a potential future failure.Having enhanced insight into the trended life-cycle of hardware, the US government and Commercial industry could potentially save millions through preventive maintenance.The Sat-Net Oracle service will be provided through an intuitive web-portal user interface powered through Kibana graphs for intuitive, visually appealing presentation of data.Previous applications of these techniques applied to Air Force E-3 Sentry aircraft has shown the potential to save up to $80M a year.Extrapolating these benefits to a significantly more expensive space network, Bluestaqs Sat-Net Oracle has the potential save hundreds of millions throughout both the Government, Intelligence Community, and Commercial markets.Predictive Failure,artificial intelligence,Space Operations,Satellite Operations,Network Maintenance
