June W. Lau (Fed)

While AI's impact is undeniable, the reach of AI into the disciplinary depths of scientific domains and subdomains has been haltingly slow and incremental. For the most part, scientific institutions are not enjoying the pan-laboratory boom that is more commonplace, for instance, in our social spheres. The reasons for the unyielding nature of research are many-fold, but perhaps the most salient reason is the lack of a common standard, often even within a single scientific field.

A reasonable step towards machine-actionable scientific research data is described by the FAIR (findable, accessible, interoperable, and reusable) framework. Reducing this framework to practice has been, and continues to be, a long and challenging road. While at NIST, my research team and I built the first laboratory information management system (LIMS) to automatically manage, curate, and tag electron microscopy datasets from a research microscopy facility. Doing so has enabled data reuse from precious tax-supported resources. Following this success, we are broadening this vision of machine-actionable scientific research data to all of NIST.

Recent events had given us incredible opportunities to showcase our innovations in measurement science that will directly benefit our nation's manufacturing engine. Our research workflows are baked into the data curation processes where a secure and auditable pipeline takes research data from the point of creation to the end publications and data products. We are AI-ready!

Where this could lead is a new paradigm for R&D. If we succeed in de-siloing discipline-specific research, we will enhance research partnerships and help researchers achieve results far faster because the shared knowledge can be made persistent and discoverable. My work motivates me because each day delayed is another day lost for scientific breakthroughs that might solve some of humanity's most pressing challenges.