Please visit Jefferson Lab Event Policies and Guidance before planning your next event:

AI4DQM workshop


Thomas Britton (JLab), Torri Jeske (JLAB)

There is a growing desire to use AI/ML to monitor data quality as evident by the increasing number of presentations on the topic coupled with the demands of autonomous systems employed during data taking (e.g. Streaming Read-Out and triggering systems). The field at large would benefit from an exchanging of ideas, tools, and techniques that have been developed for AI/ML based data quality monitoring. The AI4DQM workshop seeks to provide a forum for the discussion of data quality monitoring both in technical solutions as well as sociological challenges with adoption in a fairly informal, virtual, setting. This one day workshop will consist of a series of quick talks with room for discussions covering 4 main topics:

 1. Current workflows and methods of DQM


 2. ML techniques for anomaly detection/time series anomaly detection etc

 3. Current applications of AI4DQM


 4. Human interfacing, Trustworthiness, Interpretability, Utility

AI4DQM workshop registration
    • 1
      Speaker: Thomas Britton (JLab)
    • 2
      ATLAS status and future plans
      Speaker: Florian Haslbeck (University of Oxford / CERN)
    • 3
      Traditional DQM at JLAB
      Speaker: Torri Jeske (JLAB)
    • Requirements/Challenges with traditional methods Discussion
    • 11:00 AM
    • 4
      Siamese twin models
      Speaker: Kishansingh Rajput (Thomas Jefferson National Accelerator Facility)
    • 5
      Data-Driven Detection, Identification, and Prediction of Accelerator Cavity Faults
      Speaker: Christopher Tennant (Jefferson Lab)
    • 6
      Machine Learning based Anomaly Detection for Online Data Quality Monitoring of the CMS Electromagnetic Calorimeter

      Online Data Quality Monitoring (DQM) of the CMS electromagnetic calorimeter (ECAL) is a vital operational tool that allows ECAL experts to quickly identify, localize, and diagnose a broad range of detector issues that could affect the quality of physics data. Although the ECAL DQM system has been in operation since the start of the LHC and continuously updated to respond to new problems, it is challenging to anticipate anomalies in different shapes and sizes that had not been observed before. With the need for a more robust anomaly detection system, a real-time semi-supervised machine learning based method is developed using an autoencoder model. After accounting for spatial and time-dependent deviations in the ECAL response, the autoencoder based online DQM system is able to detect and localize anomalies with an estimated false discovery rate of 10^{-2} to 10^{-4} at 99% anomaly detection rate. We present anomaly detection results from the ECAL Barrel and Endcap regions, including the deployment results with early LHC Run3 collision data.

      Speaker: Kyungmin Park (Carnegie Mellon University / CERN)
    • 7
      Hydra: Computer Vision for Data Quality Monitoring
      Speakers: Thomas Britton (JLab), Torri Jeske (JLAB)
    • Applications Discussions