Machine Learning for Monitoring of Complex Detector Systems

The experiments running at the Large Hadron Collider (LHC) at CERN work in challenging conditions and produce an enormous amount of data to be analyzed. The smooth operation of the detectors is vital for data quality. The detector performance monitoring and the certification of the experimental data as usable for physics analysis are of key importance in ensuring the reliability of the obtained results. These human-conducted operations could be largely automated with the use of Machine Learning (ML) techniques. We report the results of a proof-of-concept study addressing this challenge by developing a customized ML tool for detector monitoring and testing its efficiency on a simulated setup of 512 individual detector modules, mimicking detector systems at the ongoing experiments at the LHC accelerator at CERN