SBIR/STTR Award attributes
Neutron scattering makes invaluable contributions to the physical, chemical, and nanostructured materials sciences. Single-crystal diffraction experiments collect volumetric scattering data sets representing the internal structure relations by combining many individual settings at different orientations, times, and sample environment conditions. It is important to be able to access the global picture through simultaneous and interactive visualization and analysis of the data. An online graphical user interface for neutron scattering scientists and visiting researchers will be developed to visualize and post-process multiple instances of wide-ranging data sets. This challenge will be addressed via multiple innovative techniques, including revolutionary server- side hardware acceleration to enable rapid data exploration. Machine learning techniques will be used to automatically extract salient features and crystal data, while removing instrumental noise. In collaboration with a state-of-the-art neutron science center, a working prototype has been developed and deployed. The hardware accelerator runs on a commercial cloud-hosted server; the visualization and graphical user interface are managed on a conventional computing cluster, also in the cloud; and the test user works from a laptop computer. Responsive visualization of remote 3D neutron scattering data has been demonstrated in a way that was not previously possible. A variety of machine learning algorithms were applied to data discovery and analysis, as well as noise reduction, setting the stage for future success. Configurable open-source software will be developed for the single-crystal neutron scattering community and the X-ray beamline community, providing unprecedented artificial intelligence and 3D visualization capabilities. By focusing on 3D visualization and using standard open-source visualization libraries, this new software technology will be compatible with long-range software development plans at the collaborating neutron science center and at several domestic X-ray light source user facilities. Effective management of live data streams from ongoing experiments will be a priority. Enabling the inspection and extraction of data, including automated detection of features and problems, will yield unprecedented advances in experiment management, big data analysis and scientific discovery. Improved capabilities for exploring large 3D data sets will bene?t commercial companies as well as university and lab researchers. The commercialization strategy will include contract R&D and subscription-based sales of our online software. The software developed for this project will be open source, and novel aspects will be published in the scientific literature.
