EPICS and Open Source Data Analytics Platforms

Abstract

SKA scale distributed control and monitoring systems present challenges in hardware sensor monitoring, archiving, hardware fault detection and fault prediction. The size and scale of hardware involved and telescope high availability requirements suggest the machine learning and other automated methods will be required for fault finding and fault prediction of hardware components. Modern tools are needed leveraging open source time series database & data analytic platforms. We describe DiaMoniCA for The Australian SKA Pathfinder Radio Telescope which integrates EPICS, our own monitoring archiver MoniCA, with an open source time series database and web based data visualisation and analytic platform.