DOE backs Fermilab-led AI project on accelerator cavity control

The Department of Energy has named Fermilab a recipient of Phase I funding under its Genesis Mission, a national initiative combining AI, supercomputing and quantum systems for scientific discovery. Fermilab leads one project itself, developing AI and machine learning algorithms to control superconducting radio-frequency cavities, the resonators that transfer energy to particle beams and are easily disturbed by small vibrations. Fermilab Director Norbert Holtkamp and chief technology officer Anna Grassellino frame the goal as more autonomous, lower-cost accelerator operation for facilities like Fermilab's own PIP-II linear accelerator. Fermilab also joins eight other DOE-funded projects led by institutions including Purdue University, Duke University and Lawrence Berkeley National Laboratory, spanning neutrino detection, collider data anomaly detection and fusion magnet simulation. The announcement itself specifies no funding amounts or timelines, only that Phase I awards are meant to identify pathways worth scaling later.