Transforming Research Visibility with RAiD at Oak Ridge National Laboratory

This blog post was written by Sheila Rabun in collaboration with Audrey Carson and Jason Wohlgemuth at Oak Ridge National Laboratory.

For large federally funded research facilities like the U.S. Department of Energy’s Oak Ridge National Laboratory (ORNL) in Tennessee, the ability to understand, connect and communicate about research activities is just as important as conducting the research itself. ORNL was founded in 1943 as part of the Manhattan Project and has had a strong focus on national security ever since. 

Traditionally, ORNL’s research infrastructure emphasizes publication-tracking elements through staff profiles and records at the Office of Science & Technical Information. While these systems capture important outputs and links between authors and their publications, they often miss the broader research lifecycle, especially for projects that do not result in formal publications, patents or technology transfers. Using the Research Organization Registry (ROR) for organizations, Open Researcher & Contributor Identifier (ORCiD) for researchers, and Digital Object Identifiers (DOIs) for outputs like publications, ORNL can identify relevant organizations, people and outputs, but has not necessarily been able to see the relationships and context in which these entities exist (see Fig. 1).

Figure 1: Persistent Identifiers (PIDs) such as ROR, DOIs and ORCID help to identify organizations, outputs and people, but many of the connections and relationships between entities remain unknown in the larger context of Research Activity Data (RAD). Image from Audrey Carson and Jason Wohlgemuth at ORNL, US RAiD Pilot participants call presentation on January 15, 2026.

To enable a more complete and robust understanding of research activity at ORNL, staff at ORNL have been working on a Research Enablement initiative to develop an iterative, machine-assisted process for collecting, standardizing, analyzing and storing research activity data (RAD) to draw insights that benefit multiple audiences and aims. What makes this effort particularly significant is its focus on projects as the primary unit of research tracking, making Research Activity Identifiers (RAiD) a critical component. With improved and standardized RAD, ORNL can provide current and prospective sponsors with a clearer understanding of ORNL’s capabilities and ongoing efforts in specific scientific domains, helping them identify opportunities for collaboration and investment. For job seekers, ORNL can create a more transparent view of the research opportunities available, allowing individuals to align their skills with relevant technologies and research areas. Policymakers can benefit from increased transparency into how research funding is being allocated, making it easier to assess impact, identify gaps and guide future investments. Internally, ORNL gains a powerful tool for identifying duplicative or overlapping efforts and directing resources more strategically, particularly in support of national security priorities and complex scientific challenges.

“We have a lot of science going on and it’s really hard to capture and quantify…just through DOIs for papers or ORCiDs for researchers…[RAiD] could be the key to unlocking more insights and analysis on our own research, turning that microscope on ourselves and learning how we can be more efficient and do better science” - Audrey Carson, Science Writer, ORNL

As part of their Research Enablement initiative, ORNL has been experimenting with RAiD as a globally unique persistent identifier for research projects, enabling a comprehensive and connected view of research activity beyond individual publications. By assigning a RAiD to each project, ORNL can track research activities from inception through execution and beyond regardless of whether traditional outputs are produced. This shift allows for greater visibility into early-stage work, interdisciplinary collaboration and relationships, and emerging areas of innovation (see Fig. 2).

Figure 2: With Research Activity Identifiers (RAiD), Research Activity Data (RAD) can more clearly represent connections and relationships between facilities, organizations, individual researchers, outputs and research projects as they evolve over time. Image from Audrey Carson and Jason Wohlgemuth at ORNL, US RAiD Pilot participants call presentation on January 15, 2026

A key component of this work centers around researcher-provided project fact sheets and a new open source tool for standardizing, documenting and maintaining project activity metadata called Accessible Content Optimization for Research Needs (ACORN). Researchers at ORNL often had fact sheets created to provide information about their projects, but this content was not standardized, and researchers would often use different vocabulary to describe the same process, workflow or equipment, for example. Additionally, fact sheets may capture ORCIDs and DOIs, but identifiers were missing at the project level. The ACORN team can use compact uniform resource identifiers (CURIEs) to identify projects internally, but they wanted a way to reference a specific project across multiple contexts, capturing external collaboration and provenance of research projects with the ability to share and reference project information across multiple departments and organizations. This is where RAiD came into play (see Fig. 3).

Figure 3: RAiD captures Persistent Identifiers (PIDs) for the people, organizations, inputs and outputs associated with research projects or research activities. Image from Audrey Carson and Jason Wohlgemuth at ORNL, US RAiD Pilot participants call presentation on January 15, 2026.

ORNL staff built ACORN as a command-line “research assistant and metascience multi-tool” designed to capture and transform RAD into “a knowledge graph amenable to automated reasoning and artifact generation” including a corresponding RAiD record for each project (https://acorn.ornl.gov/). They have already modeled RAiD metadata in Rust within the tool, and ACORN is open source so anyone can use it or adapt it. 

The ORNL team used GitLab to create a centralized workflow where researchers can submit their fact sheets for their projects where ACORN then checks and standardizes them using an internal schema for the “perfect ORNL fact sheet” built for both deep data analysis and generating assets such as PDFs, presentation slides and web pages (see Fig. 4, asset examples can be found at https://research.ornl.gov/). The team quickly realized this RAD could be aggregated and analyzed across the organization to answer questions like “What projects are using X equipment?” or “What projects do we have that work with X technology?”

Figure 4: RAD Fact Sheets including PIDs, schemas and scripts for automation can be processed in ACORN to produce analysis-ready research project data, and artifacts like PDFs, slides and web pages, to enable better understanding of research project activities, resources, relationships and capabilities. Image from https://research.ornl.gov/acorn/.

At ORNL, researchers can submit fact sheets that contain information about their projects and related people, organizations, inputs and outputs over time. Data from these fact sheets can be processed through ACORN into standardized information that shows networks, patterns and trends across research projects at ORNL. To make things as easy as possible for researchers, ORNL staff created a template with instructions for using GitLab to enter research project information and then submit a merge request via GitHub. The RAD is then automatically analyzed, misspellings and other issues corrected, and the researcher receives guidance on the best words to accurately convey the meaning of their research. Once the RAD data is submitted, the RAiD gets created based on the RAD (see Fig. 5).