Speaker
Kathryn Baker
(STFC / ISIS)
Description
Despite progress in the adoption of machine learning to enhance control, automation and decision making in complex systems such as accelerators, there is not yet a standardized methodology for integrating these models with EPICS. In this presentation, we introduce tools and a template for integrating trained models into EPICS environments. It provides a structured approach to deployment and scalability through containerisation. The presentation will include examples of deployments from the accelerators at the ISIS Neutron and Muon Source.
Primary authors
Kathryn Baker
(STFC / ISIS)
Mr
Mateusz Leputa
(STFC / ISIS)
