Fig. 1: A schematic for adaptively driven XRD with autonomous phase identification.
From: Adaptively driven X-ray diffraction guided by machine learning for autonomous phase identification
After performing a fast initial, the resulting XRD patten is fed to a pre-trained ML model which proposes likely phases. If the confidence associated with any of these phases low (< 50%), the diffractometer is instructed to perform selective rescans around peaks that distinguish the suspected phases. If necessary, the scan range is also expanded to detect additional peaks and boost the prediction confidence of the ML model.
