The fourth edition of the Reinforcement Learning for Autonomous Accelerators (RL4AA’26) workshop was held in Liverpool from 30th March to 1st April 2026 as a LIV.INNO workshop.
The workshop also received sponsorship from the Cockcroft Institute and Radiasoft. Organised by Dr Andrea Santamaria Garcia, the event brought together over 60 international researchers at the intersection of reinforcement learning and accelerator physics.
The workshop focused on the practical application of reinforcement learning to complex, real-world systems, with particular attention to the challenges of moving from proof-of-concept studies to reliable deployment in large-scale scientific facilities. Discussions reflected both the strong interest in these methods and the current technical barriers, including infrastructure, safety, and integration into existing control systems.
Feedback from participants was very positive, with an average rating of 4.38/5 and 96.9% of survey participants
The programme combined expert keynote contributions, invited and contributed talks, and introductory material aimed at supporting researchers entering the field. A dedicated advanced coding challenge, based on the CLARA accelerator at Daresbury Laboratory, provided a hands-on component and attracted strong participation, highlighting the level of technical engagement across the community.
The event also included a poster session, prize awards, and a visit to the Cockcroft Institute (CI), alongside a programme of social activities that supported interaction and collaboration throughout the workshop.
Feedback from participants was very positive, with an average rating of 4.38/5 and 96.9% of survey participants indicating they would attend a future edition.
The workshop was delivered with the support of the organising committee and sponsors, including LIV.INNO, the Cockcroft Institute, and Radiasoft, whose contributions were essential to the success of the event.
RL4AA has established itself as a focused meeting point for researchers working on reinforcement learning in scientific facilities, contributing to the growth of this emerging area and strengthening Liverpool’s role within this community.
Workshop materials: https://indico.ph.liv.ac.uk/event/2025/timetable/#all.detailed
Talk recordings: https://youtube.com/playlist?list=PLq4dcK85wL43Unzo_xgf2GHPm7eLyrCFh&si=vPDghNfapBGGcU-K
Coding challenge: https://github.com/RL4AA/rl4aa26-challenge