An expert knowledge based methodology for online detection of signal oscillations

The CERN's accelerator complex and its experiments rely on the proper functioning of a multitude of heterogeneous industrial control systems. Over 600 industrial control systems with more than 40 million sensors, actuators and control objects store more than 100 terabytes of data per year (the volume of generated data is much more). This paper describes a mathematical approach to monitor online a multitude of sensors/actuators and automatically detect signals oscillations. In order to achieve it the presented method combines both expert knowledge and spectrum analysis. Some results, obtained by the application of this analysis to the CERN cryogenics system, are presented showing multiple plant-wide oscillations. Finally the paper briefly describes the deployment of Spark and Hadoop platform into the CERN industrial environment to deal with huge datasets and to spread the computational load of the analysis across multiple hosts.