For years the persistent-identifier ecosystem had a conspicuous hole in the middle. We could identify outputs with DOIs, people with ORCID iDs, and organisations with ROR IDs — but the thing that connects them all, the research project itself, had no standard identifier. The Research Activity Identifier (RAiD) fills that hole. It identifies the research activity, not just its outputs, and it does so across the activity’s whole lifecycle. This article explains what RAiD is and why identifying the project changes what the research record can do. It draws on the research-lifecycle domain.
What RAiD is
A RAiD is a persistent identifier for a research project or activity. It is governed by an international standard, ISO 23527:2022, which gives it the same kind of formal grounding that underpins the DOI and ORCID systems. RAiD is in production: the Australian Research Data Commons (ARDC) operates as a registration authority, with services also running in the European context through the FAIRCORE4EOSC project and adoption progressing in the United States. It is not a proposal or a pilot; it is operational infrastructure.
A RAiD record is a small, structured container for the project’s identity. It holds the project’s name, its start and end dates, its current status, and — crucially — typed links to the other entities involved: the people (by ORCID iD), the institutions (by ROR ID), the funding (by grant identifier), and the outputs (by DOI). The RAiD does not duplicate all that information; it references it. The record is the hub; the spokes point out to the identifiers that already exist.
Why identifying the project, not just the outputs, matters
The conventional research record is output-centric. We record papers, datasets, and software, each with its own identifier, each citing its funding and listing its authors. From that scatter of outputs, anyone wanting to understand the project — what it set out to do, who was involved, what it produced in total, how it evolved — has to reconstruct it by inference, gathering up outputs that happen to share an acknowledgement or a grant number. The project exists only as a pattern in the outputs, never as a thing in itself.
RAiD inverts this. The project becomes a first-class entity with its own identifier, and the outputs become things the project produced. This sounds like a small change of emphasis; it is not. It means a funder can point at a project and see everything it generated, without grepping acknowledgements. It means a researcher moving institutions can carry the project’s identity with them. It means the connections between people, funding, and outputs are asserted in one authoritative place rather than re-inferred by every system that needs them.
The lifecycle dimension
The word “activity” in Research Activity Identifier is doing real work. An output is a fixed artefact: once a paper is published, its DOI points at something that does not change. A research activity is a process that moves through stages — and RAiD is designed to identify that process as it evolves, not just to label its endpoint.
A research lifecycle typically runs from idea, through planning, into execution, then dissemination, evaluation, and archiving. A project’s RAiD record persists across all of these. Its status changes — from proposed, to active, to completed — while its identifier stays constant. New outputs are linked as they appear. New collaborators are added as they join. The RAiD is the stable handle on a moving target, which is exactly what an output identifier cannot be, because an output does not move.
This lifecycle persistence is why RAiD connects naturally to the other stage-aware identifiers. The data-management plan has a DMP ID; the funding has a grant ID; the outputs have DOIs; the people have ORCID iDs. The RAiD is the spine that threads through the whole lifecycle, linking the plan made at the start to the outputs produced at the end and the evaluation that follows. Without it, those stage-specific records are islands; with it, they are a connected account of a single activity over time.
What RAiD is not
It is worth marking the boundaries to avoid over-claiming. A RAiD is not a project-management tool: it does not run your Gantt chart, track your milestones in detail, or manage your budget. It is an identifier with enough structured metadata to locate the project in the wider research graph, not a system for executing the project. Nor does RAiD replace the institution’s CRIS; rather, the CRIS is one of the systems that can mint and reference RAiDs, using the RAiD as the shared, portable identity for a project that the CRIS otherwise holds in local form. And a RAiD is not automatically a public disclosure of everything about a project — its metadata can be managed with appropriate visibility, which matters for projects that are commercially sensitive or in early stages.
Why this is infrastructure worth adopting
The case for RAiD is the same case that justifies the whole persistent-identifier ecosystem: it lets a fact be asserted once, authoritatively, and read everywhere. The project’s membership — who, what funding, which outputs — is exactly the kind of fact that is currently re-derived independently by funders, institutions, and aggregators, each from incomplete evidence, each slightly differently. A standard project identifier with a structured record is the mechanism for asserting it once. As RAiD adoption grows, the research graph gains the connective tissue it has been missing, and the project stops being a ghost inferred from its outputs and becomes a thing the record can actually point at.
What to do now
For institutions and CRIS owners: begin minting and referencing RAiDs for projects, using them as the portable identity that links your local project records to the wider graph. For funders: reference RAiDs in award and reporting workflows so that a project’s outputs aggregate against its identifier automatically. For researchers: where your institution supports it, ensure your projects carry RAiDs and that your ORCID iD is linked, so your project work travels with you.
Leave a Reply