Measured acceleration · Materials science
CAMEO closes the loop for active-learning materials discovery
Closed-loop active learning at a synchrotron beamline for materials phase mapping and optimization.
Summary
The Nature Communications paper presents CAMEO, a closed-loop active-learning system for materials exploration and optimization at a synchrotron beamline. It combines autonomous phase mapping, property optimization and human-in-the-loop input, and reports discovery of an epitaxial phase-change memory nanocomposite.
AI role
Selected experiments and optimized materials exploration while incorporating human input.
Narrative role
CAMEO shows AI embedded inside scientific instrumentation, a route to shortening experiment-selection loops in materials science.
Caveat
It is a specialized workflow at a beamline, so generality depends on instrument and problem transfer.