Speaker
Description
Significant progress has been made in the development of AI-ML tools at the ATLAS heavy-ion linac at Argonne. Although all tools were successfully tested online, they were not accessible to the operators in the control room. To facilitate this, we have developed a dedicated AI-ML interface that allows the user to select the beamline to tune, the task to perform, hit the execute button and watch the results. In addition to presenting the main features and capabilities of the interface and its integration into the ATLAS control system, we will report on the operational experience and feedback from the operators. Another important update is the development of a virtual model for the ATLAS linac. A model that allows offline testing before online deployment, uses the same interface with the same online experience, and most importantly, mimics all aspects of the machine. After presenting a general procedure on how to develop an effective virtual accelerator model, we highlight a few example applications. Finally, moving a step closer towards a digital twin, we plan to add real-time online exchange between the virtual accelerator and the real machine, albeit in one direction at this time.
Funding Agency
This work was supported by the U.S. Department of Energy, under Contract No. DE-AC02-06CH11357. This research used the ATLAS facility, which is a DOE Office of Nuclear Physics User Facility.
| In which format do you inted to submit your paper? | Word |
|---|