EOSC SRIA 2.0: turning shared infrastructure into research capability
Published in September 2026, the European Open Science Cloud Strategic Research and Innovation Agenda 2.0 sets priorities for 2026–2030. Its central concern is making the EOSC Federation a sustainable environment for research. For institutions and competence centres, the agenda connects infrastructure development with a practical question: can researchers reliably combine data, software and services from different providers?
The document strengthens the operational focus of EOSC. Fragmented funding, inconsistent access rules, uneven implementation of standards and limited user uptake continue to constrain collaboration. The response combines technical integration with governance, professional support and predictable funding. The five priorities below organise this approach; they should be understood as a consolidation and development of earlier work, rather than five entirely new subjects.
| Strategic priority | Challenge | Strategic response |
|---|---|---|
| Operationalisation of the EOSC Federation | Fragmented services and short funding cycles undermine continuity. | Establish sustainable funding, operational accountability and working connections between Nodes. |
| Sovereign, secure and trusted research data | Uncertain rights, variable quality and preservation gaps weaken trust. | Support controlled sharing, quality assessment, provenance and long-term retention. |
| AI and EOSC | Data readiness and AI transparency remain uneven. | Prepare data for AI and govern AI-assisted research and curation. |
| User-centricity | Skills gaps and complex procedures discourage adoption. | Develop competence centres, practical training and continuous user feedback. |
| Strategic alignment and global collaboration | Disconnected initiatives restrict cross-border reuse. | Align policies and interfaces with other data spaces and research infrastructures. |
A particularly important distinction concerns interoperability. Common standards provide a basis for cooperation, but their existence does not demonstrate that resources work together. SRIA 2.0 calls for implementation across Nodes, including compatible metadata, access mechanisms and service interfaces. Practical use cases must show that a researcher can find a resource, obtain authorised access and use it within a workflow involving another provider.
The agenda also separates FAIRness from scientific quality. Persistent identifiers, metadata and licences support discovery and reuse, but cannot establish whether data are accurate, complete or appropriate for a particular analysis. Assessment must extend beyond repository metadata to validation, provenance and traceability. Likewise, FAIR data are not automatically ready for machine learning: harmonisation, preprocessing and assessment of bias or representativeness may still be required.
Long-term preservation is treated as a research lifecycle responsibility. Reproducibility depends on retaining software, workflows, containers and documentation alongside datasets. Institutions need transparent criteria for selecting outputs for preservation, allocating resources and maintaining their interpretability. Expanding storage alone cannot provide this assurance. Data stewards, curators and research software engineers are therefore essential to maintaining the scientific value of digital outputs.
SRIA 2.0 contains an implementation agenda, but should not be read as a fully scheduled and costed delivery plan. Section 3.2 presents recommended actions and intended outcomes for each priority, while progress is linked to the existing EOSC Partnership Monitoring Framework and its performance indicators. Detailed responsibilities, budgets and milestones still require operational decisions. The following institutional steps are an analytical interpretation of these recommendations, not a timetable prescribed by SRIA.
| Institutional step | Practical action |
|---|---|
| Establish the baseline. | Identify datasets, software, services, users and support gaps. |
| Test collaboration. | Demonstrate a research workflow that uses resources from several institutions. |
| Secure continuity. | Assign operational responsibilities, preservation arrangements and funding. |
| Assess benefits. | Track reuse, usability and scientific gains alongside technical performance. |
The evidence behind the agenda is also useful for institutions planning their own consultations. Annex B reports input from 80 of 186 EOSC Winter School registrants and a public consultation with 83 respondents, conducted from 18 December 2025 to 31 January 2026. The questionnaire combined open questions with 45 statements rated on a five-point relevance scale. Its results identify sustainable funding, sovereignty and trust, service interoperability, infrastructure coordination, adoption barriers and secure data exchange as the six strongest consensus priorities.
These findings need careful interpretation. The participants were predominantly connected with EOSC and based in Western and Northern Europe, so the sample does not represent all researchers. Because ratings clustered towards the upper end, the analysis used the proportion selecting the highest score to distinguish priorities. Annex B includes the questionnaire itself, offering a reusable reference for institutional surveys, provided its language and scope are adapted to local research practices.
For the NASU Research Data Management Competence Center, the agenda supports a shift towards assistance embedded in real research tasks. Training can help teams prepare documented datasets, publish licensed software and package reproducible workflows. Support should also cover controlled access where unrestricted publication is inappropriate. For prospective NASU thematic Nodes, the practical test is whether shared resources support a research task across institutions. Progress should be demonstrated through successful reuse, functioning services and stronger research collaboration.
Source: Strategic Research and Innovation Agenda of the European Open Science Cloud, version 2.0, September 2026, particularly Sections 3.1–3.2 and Annex B.