Artificial intelligence and machine learning (AI/ML) hold great potential to impact the work performed at the APS. Many exciting activities are already underway, and many more are being planned. To address these current and future challenges, APS convened a workshop to discuss current projects and tools, exchange ideas and experiences, and inspire collaboration.  

Researchers from across the laboratory met to address application of AI/ML techniques and approaches to a range of scientific challenges at the APS, including the following: 3D segmentation of tomographic data, self-driving beamline operation, tomographic reconstruction workflows, image acquisition and reconstruction, APS-U accelerator applications, beamline optimization and control, pulse processing and data analysis for superconducting X-ray spectrometers, advanced X-ray spectroscopy, real-time coherent diffraction inversion, problems in microscopy and spectroscopy, cyberinfrastructure for autonomous science, separating components of heterogeneous materials, in situ synthesis, identification and removal of spurious data, and APS linac bunch charge transmission efficiency.

The APS AI/ML Workshop was held over two sessions: Tuesday, January 21, 2020, from 9:00 a.m. to noon, and Friday, January 24, 2020, from 9:00 a.m. to noon in Building 401.