ICALEPCS 2025 - The 20th International Conference on Accelerator and Large Experimental Physics Control Systems

Session

FRAG MC13 Artificial Intelligence and Machine Learning

FRAG

26 Sept 2025, 09:00

Grand Ballroom (Palmer House Hilton Chicago)

Conveners

FRAG MC13 Artificial Intelligence and Machine Learning

• Mirjam Lindberg (MAX IV Laboratory)
• Mike Fedorov (Lawrence Livermore National Laboratory)

166. ML-assisted beamline optimization at LCLS (FRAG001)

Zachary Lentz (Linac Coherent Light Source)

26/09/2025, 09:00

MC13: Artificial Intelligence & Machine Learning

Contributed Oral Presentation

LCLS is currently developing and deploying beamline optimization techniques at our x-ray end-stations. This is an increasingly important topic at LCLS as it fully leverages its new high rep-rate superconducting beam. The increased throughput of LCLS-II shifts the performance bottleneck to on-shift setup time. As part of the Illumine collaboration, LCLS is leveraging Bayesian optimization...

395. Development and experimental validation of a machine learning-based methodology for cyclotron beam control: results from the PSI HIPA facility (FRAG002)

Thierry Neal (TRANSMUTEX)

26/09/2025, 09:15

MC13: Artificial Intelligence & Machine Learning

Contributed Oral Presentation

Transmutex SA is developing an accelerator-driven system (ADS) designed to generate clean energy while reducing the lifetime of radioactive waste. Such a subcritical reactor concept requires high reliability and a high degree of accelerator automation to ensure operational effectiveness.

To address these demands, a machine learning (ML) methodology was developed and experimentally...

238. Adaptive model-based reinforcement learning for orbit feedback control in NSLS-II storage ring (FRAG004)

Zeyu Dong (Stony Brook University)

26/09/2025, 09:30

MC13: Artificial Intelligence & Machine Learning

Contributed Oral Presentation

The National Synchrotron Light Source II (NSLS-II) uses a highly stable electron beam to produce high-quality X-ray beams with high brightness and low-emittance synchrotron radiation. The traditional algorithm to stabilize the beam applies singular value decomposition (SVD) on the orbit response matrix to remove noise and extract actions. Supervised learning has been studied on NSLS-II storage...

414. Agentic Systems in Accelerator Control and Optimization (FRAG005)

Thorsten Hellert (Lawrence Berkeley National Laboratory)

26/09/2025, 09:45

MC13: Artificial Intelligence & Machine Learning

Contributed Oral Presentation

The deployment of agentic AI systems at the Advanced Light Source (ALS) marks a major step toward autonomous, intelligent facility operations. By connecting large language models (LLMs) with diverse data sources, we are developing agents that not only interface with the control system but also provide a natural language interface for operators and scientists. This allows users to interact with...