Task 3.7: L1 scouting data compression for efficient data acquisition and anomaly detection
Task leads: Jennifer Ngadiuba (Fermilab)
L1 scouting provides the possibility for unbiased HL-LHC data acquisition and storage for future analysis, but the resulting datasets would be prohibitively large, order 100-1000 PB per data-taking year depending on the kind of information saved. In this task, we propose to apply cutting-edge compression techniques, including nonlinear lossy compression with AI algorithms (e.g., autoencoders) to reduce the L1-scouting dataset size. Autoencoders are also a promising algorithm for anomaly detection, and so they will also be explored for that purpose, that can be already applicable to Run 3 data. For this task, we would also work on optimizing the hardware design of the algorithm (resource consumption and latency), to potentially run it as part of the main L1 trigger system to add scouting, both for Run3 and HL-LHC.
Publications and other resources
- Jannicke Pearkes, Presentation at ML4JET2023, “Realtime Anomaly Detection in the CMS Experiment Global Trigger Test Crate”
- Jennifer Ngadiuba, Presentation at ICHEP 2024, 18/07/2024, “Transforming the LHC Physics Program with AI”
- Noah Zipper, Presentation at FastML24, 15/10/2024, “Real-time Anomaly Detection in the CMS Experiment“
- Melissa Quinnan, Presentation at CHEP2024, 24/10/2024, “Anomaly Detection in the CMS L1 Trigger”
- Jennifer Ngadiuba, Presentation at ML4JETS, 04/11/2024, “Experimental highlights: Edge AI for particle physics”
- Jannicke Pearkes, Presentation at Openlab Technical WP, 04/03/2025 “CMS Experience with Anomaly Detection in the L1 Trigger”
- Jannicke Pearkes, Presentation at LLP2025 05/06/2025, “CMS: Anomaly Detection Triggers”
- Melissa Quinnan, Presentation at AD4HEP2025, 16/06/2025, “Anomaly Detection in the CMS L1 Trigger with AXOL1TL”
- Jannicke Pearkes, Presentation at ML4JETS2025, 21/08/2025 “AI at the extreme edge”
- Jannicke Pearkes, Presentation at Lepton-Photon 2025, 26/08/2025 “Anomaly Triggers for Exotics Searches”
- Sabrina Giorgetti, Presentation at FastML, 03/09/2025, “Real-Time Anomaly Detection in the CMS Level-1 Trigger with AXOL1TL”
- Eric Moreno, Presentation at FastML, 03/09/2025, “COLLIDE-2V – 750 Million Dual-View LHC Event Dataset for Low-Latency ML”
- Maciej Glowacki, Presentation at ACAT 08/09/2025, MLOps Pipeline for Continuous Deployment of Machine Learning Algorithms in the CMS Level-1 Trigger”
- Maciej Glowacki, Poster at ACAT 08/09/2025, “Real-Time Unsupervised Anomaly Detection in the CMS Level-1 Trigger”