AI/ML Data Management Software System for NSRCs

Scientific research facilities are unable to collaborate and fully utilize microscopy image data due to data silos and a lack of common data management standards. Scientists at research facilities are burdened with laborious, complex processes to manage microscopy image data, reducing the impact of scientific datasets and research. To increase the impact of scientific datasets by delivering (1) an extensible software system and (2) microscopy tools which support findability, accessibility, interoperability, and reuse of data in multi- tool, multi-user scientific research facilities. Working directly with scientific research facilities and scientists, an extensible software system with microscopy tools will be developed, improving the impact and reusability of datasets. The proposed software system is a collaboration platform that builds an eco-system of integrated digital tools to support the scientific community. Accompanying models will fall into the two categories of annotation and search/recall, allowing for the automatic searching and indexing of existing data stored in filesystems, the provisioning of suggested links between data and other resources, automated clustering and metadata suggestion, trend discovery, and streamlined data storage. Initially targeting public sector nanoscience scientific research centers for proof-of-concept and minimum viable product user testing, the proposed software system and machine learning models are applicable to a variety of use cases for other multi-tool, multi-user scientific research facilities including 42 national laboratories and over 250 academic research institutions. The solution will enable users to significantly increase the impact of datasets in their facilities by supporting greater use and collaboration, allowing researchers to focus on advancing scientific innovations. The proposed solution will also reduce the substantial costs associated with manual data management and decrease human error, which is proven to corrupt datasets.