2
Operators of Jefferson Lab’s primary particle accelerator are getting a new tool to help them quickly address issues that can prevent it from running smoothly. The machine learning system has passed its first two-week test, correctly identifying glitchy accelerator components and the type of glitches they’re experiencing in near-real-time. An analysis of the results of the first field test of the custom-built machine learning system was recently published.