Speaker
Description
The increasing complexity and data demands of modern particle accelerator and large research facilities necessitate a paradigm shift towards unified, intelligent control system architectures. This paper proposes a study to enhance a particle accelerator control framework architecture in order to develop an AI-ready control system, drawing upon the strengths of both open-source and commercial control frameworks, on the basis of the IFMIF-DONES control design experience. By integrating the widely adopted Experimental Physics and Industrial Control System (EPICS) with industrial Supervisory Control and Data Acquisition (SCADA) systems via a key component –an OPC UA server for EPICS pvAccess– a seamless and standardized communication layer is established. This hybrid approach enhances flexibility, scalability, and long-term maintainability while leveraging the benefits of both ecosystems. Furthermore, the proposed architecture explores the potential of unifying traditionally separate networks, such as Timing, Control/Monitoring, and Interlock, through Time-Sensitive Networking (TSN) to simplify infrastructure and improve bandwidth utilization. This paper outlines the architectural design, discusses the advantages of this integrated approach in achieving AI readiness and improved interoperability, and highlights the role of a EPICS - OPC UA Gateway as a cornerstone for future advancements in accelerator control.
