Materials Science • ORNL
Neutron scattering: accelerating scientific discovery through adaptive experiments
AI-guided neutron experiments replace fixed measurement plans with adaptive decision making. As experimental data are collected, AI recommends the next measurement while scientists monitor the evolving experiment in real time. The National Science Data Fabric (NSDF) serves as the digital backbone that connects facility instruments, AI services, computing, and live dashboards, making experiments observable, persistent, and reusable without changing facility operations or compromising security.
The Scientific Challenge
Neutron scattering experiments have limited beam time, making every measurement valuable. Instead of exhaustively scanning all possible conditions, scientists need to identify the most informative measurements while an experiment is still running. Achieving this requires AI recommendations, live visualization, and a complete experimental record, all while working within the facility’s existing control, networking, and security infrastructure.
How NSDF Helps
NSDF provides the external digital backbone that enables AI-guided experiments without replacing existing facility software. It captures experimental data as they are produced, delivers synchronized dashboards with measured data and AI recommendations, preserves complete provenance for replay and analysis, and securely shares experiment state across institutional boundaries. Scientists remain in control of every decision while NSDF makes experiments observable, interpretable, and reusable.
Scientific Impact
Using NSDF, scientists observe the evolving experiment, understand why the AI recommends each measurement, and concentrate beam time on the most informative regions of the experiment. The same digital backbone supports multiple neutron instruments with minimal changes, demonstrating a reusable approach for AI-guided experimentation across scientific user facilities while preserving human oversight and trusted facility operations.
Key Outcomes
- Adaptive Measurements — AI recommends the next measurement in real time to focus beam time where it is most informative.
- Live Scientific Insight — Scientists monitor measured data and AI recommendations together through synchronized dashboards.
- Complete Provenance — Every measurement and AI recommendation is preserved for replay, validation, and future analysis.
- Reusable Digital Backbone — The same NSDF infrastructure supports multiple neutron instruments without redesigning the external platform.