Name: 5th ICFA Beam Dynamics Mini-Workshop on Machine Learning for Particle Accelerators
Start: 2025-04-08T08:00:00+02:00
End: 2025-04-11T19:30:00+02:00
Location: CERN

5th ICFA Beam Dynamics Mini-Workshop on Machine Learning for Particle Accelerators

from Tuesday 8 April 2025 (08:00) to Friday 11 April 2025 (19:30)

Monday 7 April 2025
Tuesday 8 April 2025

08:30 Registration
Registration
08:30 - 09:00
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
09:00 Introduction and welcome - Verena Kain (CERN) Tia Miceli
Introduction and welcome
- Verena Kain (CERN)
- Tia Miceli
09:00 - 09:20
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
09:20 Highlights from RL4AA'25 - Annika Eichler (DESY) Andrea Santamaria Garcia (University of Liverpool)
Highlights from RL4AA'25
- Annika Eichler (DESY)
- Andrea Santamaria Garcia (University of Liverpool)
09:20 - 09:50
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
09:50 Introduction to Control Theory - Annika Eichler (DESY)
Introduction to Control Theory
- Annika Eichler (DESY)
09:50 - 10:45
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor Control theory is a pivotal field of study that focuses on the behavior of dynamical systems and the development of strategies to influence these systems towards desired outcomes. The principle of control theory find its application in plenty of disciplines including engineering, economics, biology and beyond. It were control concepts like the Kalman filter that has flew the Apollo to the moon. Also for the operation of particle accelerator there are thousands of control loops running, ranging from low-level hardware-focused systems, as magnet powering, synchronization or RF control, over fast transverse feedback systems to beam steering. In this tutorial, we begin by exploring the basic concepts of control, emphasizing the importance of stability and robustness. The tutorial then introduces the mathematical modeling of systems using differential equations and the representation of these systems in state-space form. Starting from classical control strategies, we will go towards modern optimal control techniques and will finally see the link to optimization and ML approaches like reinforcement learning.
10:45 Coffee Break
Coffee Break
10:45 - 11:15
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
11:15 Multi-objective RL - Kishan Rajput (Jefferson Lab)
Multi-objective RL
- Kishan Rajput (Jefferson Lab)
11:15 - 12:15
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor Multi-objective reinforcement learning (MORL) extends traditional reinforcement learning (RL) by addressing environments where multiple conflicting objectives must be optimized simultaneously. In real-world applications, such as autonomous systems, particle accelerator optimization and control, agents often face trade-offs between competing goals. This lecture provides an overview of the key concepts, techniques, and challenges in MORL. We will explore different approaches to handling multiple objectives, including scalarization methods, Pareto dominance, and reward decomposition. Additionally, we will discuss the concept of the Pareto front and how it can guide decision-making in complex environments. The lecture will also highlight the role of exploration-exploitation trade-offs in multi-objective settings and the development of algorithms that can balance these objectives effectively. By the end of the session, attendees will gain a deeper understanding of how to design and implement multi-objective RL systems capable of navigating the complexities of real-world decision-making tasks.
12:15 Lunch
Lunch
12:15 - 14:00
Room: Restaurant 1
14:00 Differentiable simulations - Juan Pablo Gonzalez Aguilera (University of Chicago) Andrea Santamaria Garcia (University of Liverpool) Chenran Xu Jan Kaiser
Differentiable simulations
- Juan Pablo Gonzalez Aguilera (University of Chicago)
- Andrea Santamaria Garcia (University of Liverpool)
- Chenran Xu
- Jan Kaiser
14:00 - 15:00
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor Many accelerator physics problems, such as beamline design, beam dynamics model calibration, online tuning and phase space measurements rely on solving high-dimensional optimisation problems over beam dynamics simulations. Numerical optimisers have successfully been applied to such tasks, but they struggle as the dimensionality and complexity of the objective function increase. In machine learning, gradient-based optimisation algorithms are successfully used to optimise billions of model parameters over complex loss functions when training large neural network models. This is made possible by reverse-mode automatic differentiation, which enables the fast computation of gradients of complex functions. In this tutorial, you will learn to use novel beam dynamics simulations with support for automatic differentiation to your advantage and harness the power of gradient-based optimisation in accelerator physics. Multiple hands-on examples using the Cheetah beam dynamics code will allow you to try these methods for yourself. While we will present multiple example applications of gradient-based optimisation on differentiable beam dynamics simulators, the space of potential applications here is vast, and we believe that participants will go on to discover numerous novel applications for differentiable beam dynamics simulations that were intractable to solve with existing methods.
15:00 Coffe break
Coffe break
15:00 - 15:30
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
15:30
15:30 - 17:10
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
Contributions

15:30 Serval Applications of Machine Learning at HEPS- 15'+5' - wei bao

15:50 Photon systems automation activities at EuXFEL 15'+5' - Sarlota Birnsteinova (European XFEL GmbH)

16:10 ML optimization methods for APS-U commissioning- 15'+5' - Nikita Kuklev (Fermilab)

16:30 Data-Driven Feedback Optimization for Particle Accelerator Control- 15'+5' - Christian Hespe (DESY)

16:50 Machine Learning for Online Control of Particle Accelerators- 15'+5' - Xiaolong Chen (Institute of modern physics)
17:10 MALAPA Community Resources
MALAPA Community Resources
17:10 - 18:10
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor

18:45 Cocktails with Jazz and finger food
Cocktails with Jazz and finger food
18:45 - 20:45
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
Wednesday 9 April 2025
08:30
08:30 - 10:30
Room: 503/1-001 - Council Chamber
Contributions

08:30 Physics-informed Bayesian inference and optimization of the closed orbit in synchrotrons -15'+5' - Victoria Isensee

08:50 Optimising Injection Efficiency at Diamond Light Source using Gaussian Processes with Non-Gaussian Likelihoods - 15'+5' - Shaun Preston (John Adams Institute, University of Oxford)

09:10 Eliminating mains noise with Machine Learning- 15'+5' - Verena Kain (CERN)

09:30 Beam halo losses reduction with simulation constrained Bayesian Optimization- 15'+5' - Andrea De Franco (National Institutes for Quantum Science and Technology (QST))

09:50 Efficient Dynamic and Momentum Aperture Optimization for Lattice Design Using Multipoint Bayesian Algorithmic Execution- 15'+5' - Daniel Ratner (SLAC) Zhe Zhang

10:10 Progress on Automating Experiments at the Argonne Wakefield Accelerator Using Advanced Bayesian Optimization Algorithms - 15'+5' - Ryan Roussel
10:30 Coffee break
Coffee break
10:30 - 11:00
Room: 61/1-201 - Pas perdus - Not a meeting room -
11:00
11:00 - 12:00
Room: 503/1-001 - Council Chamber
Contributions

11:00 Harnessing the Power of Gradient-Based Simulations for Multi-Objective Optimization in Particle Accelerators- 15'+5' - Kishansingh Rajput (Jefferson Lab)

11:20 Virtual to Physical: Reinforcement Learning to Optimize SNS Particle Accelerator Controls- 15'+5' - Armen Kasparian (Jefferson Lab)

11:40 Explainable physics-based constraints on reinforcement learning for accelerator controls- 15'+5' - Jonathan Colen (Old Dominion University)
12:00
12:00 - 12:45
Room: 503/1-001 - Council Chamber
Contributions

12:00 Machine learning and computer vision in robotic interventions at CERN boosting maintainability- 10'+5' - Eloise Matheson (CERN)

12:15 Application of boosted decision trees in beam anomaly detection with high throughput data streaming system - 10'+5' - Maciej Mleczko (National Synchrotron Radiation Centre)

12:30 An Integrated Research Infrastructure framework for digital twins of laser-plasma acceleration experiments- 10'+5' - Remi Lehe (LBNL)
12:45 Lunch
Lunch
12:45 - 14:00
Room: Restaurant 1
14:00
14:00 - 15:00
Room: 503/1-001 - Council Chamber
Contributions

14:00 Lume Deployment - standardising model deployment- 15'+5' - Mateusz Leputa

14:20 Software Infrastructure Plans for End-to-End and Bottom-Up AI/ML Capabilities at the Electron-Ion Collider (EIC)- 15'+5' - Linh Nguyen (Brookhaven National Laboratory)

14:40 Machine Learning for the SNS Accelerator and Target- 15'+5' - Willem Blokland (ORNL)
15:00 Coffee break
Coffee break
15:00 - 15:30
Room: 61/1-201 - Pas perdus - Not a meeting room -

15:30 MLOps topic
MLOps topic
15:30 - 16:00
Room: 503/1-001 - Council Chamber

16:00 Posters
Posters
16:00 - 18:00
Room: 61/1-201 - Pas perdus - Not a meeting room -

19:00 Dinner
Dinner
19:00 - 21:00
Room: Restaurant 1
Thursday 10 April 2025
08:30
08:30 - 10:30
Room: 503/1-001 - Council Chamber
Contributions

08:30 A path to efficient machine learning-based beam diagnostics: complete six-dimensional generative phase space reconstruction without RF deflecting cavity- 15'+5' - Seongyeol Kim

08:50 Measurement of CSR effects using generative phase space reconstruction- 15'+5' - Juan Pablo Gonzalez-Aguilera (University of Chicago)

09:10 Semi-supervised detection of optics errors in beamlines- 15'+5' - Jonathan Edelen

09:30 Anomaly Forecasting and Adaptive Learning in Fast Kicker Magnet Systems- 15'+5' - Malik Marco Algelly (Universite de Geneve (CH))

09:50 Enhancing Quench Detection in SRF Cavities at the European XFEL: Machine Learning Approaches and Practical Challenges- 15'+5' - Lynda Boukela (DESY) Burak Dursun (DESY)

10:10 Improving Coincident Learning for Beam-based RF Station Fault Identification Using Phase Information at the SLAC Linac Coherent Light Source- 15'+5' - Jason Liang (Stanford University)
10:30 Coffee Break
Coffee Break
10:30 - 11:00
Room: 61/1-201 - Pas perdus - Not a meeting room -
11:00
11:00 - 12:20
Room: 503/1-001 - Council Chamber
Contributions

11:00 Graph Learning for Explainable Operation of Particle Accelerators- 15'+5' - Chris Tennant

11:20 Compensating hysteresis for the CERN SPS main magnets with transformers- 15'+5' - Anton Lu (Technische Universität Wien (AT))

11:40 Improving fast beam transport simulations using transfer learning- 15'+5' - Paul Stanik III (University of Nevada--Las Vegas)

12:00 Optimizing High-Energy Beam Transport with AI: Advances in IFMIF-DONES Design- 15'+5' - Galo Gallardo Romero (HI Iberia) Guillermo Rodríguez Llorente (Artificial Intelligence Engineer at HI Iberia, Department of Mathematics at UC3M, Gregorio Millán Barbany Institute)
12:20 Lunch
Lunch
12:20 - 14:00
Room: Restaurant 1

14:00 Framework for Digital Twins
Framework for Digital Twins
14:00 - 15:00
Room: 503/1-001 - Council Chamber

15:00 Coffee break
Coffee break
15:00 - 15:30
Room: 61/1-201 - Pas perdus - Not a meeting room -
15:30
15:30 - 16:50
Room: 503/1-001 - Council Chamber
Contributions

15:30 End-to-end differentiable digital twin for the IOTA/FAST facility- 15'+5' - Nikita Kuklev (Fermilab)

15:50 An Online Virtual Model for the ATLAS Ion Linac at Argonne- 15'+5' - Adwaith Ravichandran (Argonne National Laboratory)

16:10 ML-Based Multi-Fidelity Model Calibration Toward Precision Control of Electron Beams- 15'+5' - Eric Cropp

16:30 Developing Radiation-Tolerant Transverse Beam Imaging Using Synthetic Data and Multimode Fiber- 15'+5' - Qiyuan Xu
16:50 AI roadmap topic
AI roadmap topic
16:50 - 17:50
Room: 503/1-001 - Council Chamber

18:15 Bus transportation to Geneva city center for the Gala dinner
Bus transportation to Geneva city center for the Gala dinner
18:15 - 19:30
Room: 39

19:30 Dinner at Beau Rivage Hotel
Dinner at Beau Rivage Hotel
19:30 - 22:30
Room: Hotel Beau Rivage Geneve

22:30 Bus transportation to Cern B39
Bus transportation to Cern B39
22:30 - 23:00
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
Friday 11 April 2025
08:30 Large Language Model Chatbots for Enhanced Documentation Access - Florian Rehm (CERN)
Large Language Model Chatbots for Enhanced Documentation Access
- Florian Rehm (CERN)
08:30 - 09:30
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor This tutorial applies Retrieval Augmented Generation (RAG) as a method to improve documentation retrieval in accelerator physics. Participants will learn how combining information retrieval with generative AI models can provide precise, context-aware answers from vast technical resources. The session includes a hands-on demonstration of implementing RAG in combination with Large Language Models (LLMs) and retrieving information from your documentation.
09:30 sulz
sulz
09:30 - 09:50
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
Contributions

09:30 Multi-modal LLMs at EuXFEL: Knowledge Assistants, Data Exploration, and Policy- 15'+5' - Florian Sohn (European XFEL GmbH)
09:50 Group photo and Coffee break
Group photo and Coffee break
09:50 - 10:20
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
10:20
10:20 - 11:00
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
Contributions

10:20 The Journey of Developing Specialized Text Embedding Models- 15'+5' - Thorsten Hellert

10:40 Towards Agentic AI on Particle Accelerators- 15'+5 - Raimund Kammering
11:00 summary, preview, final words - Remi Lehe Jan Kaiser Thorsten Hellert Verena Kain (CERN) Daniel Ratner (SLAC) Hirokazu Maesaka Ryan Roussel Georg Hoffstaetter Auralee Edelen Kishan Rajput (Jefferson Lab)
summary, preview, final words
- Remi Lehe
- Jan Kaiser
- Thorsten Hellert
- Verena Kain (CERN)
- Daniel Ratner (SLAC)
- Hirokazu Maesaka
- Ryan Roussel
- Georg Hoffstaetter
- Auralee Edelen
- Kishan Rajput (Jefferson Lab)
11:00 - 12:00
Room: 80/1-001 - Globe of Science and Innovation - 1st Floor
12:00 Lunch
Lunch
12:00 - 13:30
Room: Restaurant 1

13:30 CERN visits
CERN visits
13:30 - 16:30