https://zenodo.org/records/21825601

One of the report’s central conclusions is that a successful EOSC Node does not necessarily require the construction of a new large-scale digital infrastructure. Most of the nodes studied are being developed as coordination and federation layers on top of existing repositories, computing platforms, cloud services, research infrastructures and disciplinary services. Research data generally remain under the control of their original providers, while the node supports discovery, description, controlled access and integration through common mechanisms for authentication, cataloguing, metadata exchange, monitoring and user support. For EOSC Ireland, the report describes this approach as a “lightweight, federation-first” model, in which interoperability and participation in the wider federation take priority over building every component locally.

From a technical perspective, a relatively compact set of core capabilities appears repeatedly across the different nodes. These include authentication and authorisation infrastructure, catalogues of resources and services, metadata aggregation, monitoring, helpdesk functions and service management. MyAccessID is highlighted as an increasingly important mechanism for federated authentication. At the same time, the report stresses that authentication alone is not sufficient. Effective collaborative research also requires mechanisms for managing groups, project membership, roles and permissions. SURF Research Access Management provides an example of this collaboration and authorisation layer, while Life Sciences Connect demonstrates how a mature discipline-specific identity system can be aligned with the broader EOSC Federation environment.

Service and resource catalogues are treated as one of the main user-facing elements of a node. Their purpose is not simply to provide a list of available services, but to support a usable path from discovery to access and actual use. EOSC Finland, EOSC Poland and the EOSC EU Node all rely heavily on metadata aggregation and discovery while leaving the underlying data with the original providers. The report refers to technologies and approaches such as OAI-PMH, the OpenAIRE Knowledge Graph, DCAT and APIs as mechanisms that can support machine-to-machine interaction between repositories, catalogues and services. From the perspective of research communities, the important point is that the value of a node is determined less by the number of new platforms it creates than by its ability to make existing resources visible, interoperable and accessible under clear conditions.

The examples of PaNOSC and Life Sciences Connect are particularly relevant to the development of thematic EOSC Nodes. PaNOSC brings together European photon and neutron research infrastructures and demonstrates a model in which the node acts as a common access point to data, services and scientific workflows rather than as a central repository. Its environment includes repositories, remote data processing services, a service catalogue, community AAI, training resources, PaN-Finder and PaN-Space. Particular attention is given to reproducible workflows, containerised analysis environments and remote processing approaches that allow researchers to work with large scientific datasets without necessarily transferring them to a single central location. Life Sciences Connect provides another model, building on mature infrastructures such as EMBL, ELIXIR, Instruct ERIC and Euro-BioImaging and connecting them through common approaches to AAI, metadata, catalogues, interoperability and trust.

Governance is another major theme of the report. The more mature nodes have a clearly identified coordination function while keeping strategic governance separate from day-to-day operational delivery. The Finnish model combines CSC as a strong national coordinating and service organisation with forums and working groups that provide broader community participation. EOSC-PL uses a consortium structure supported by a wider stakeholder network. Life Sciences Connect builds on governance arrangements that already exist in major European research infrastructures. Across these models, the same lesson emerges: broad participation should not result in fragmented accountability. Sustainable operation requires clear roles, identifiable service owners, transparent decision-making processes and staff who are explicitly responsible for operational tasks.

Funding and sustainability remain among the most difficult issues for almost all of the nodes examined. CSC and SURF benefit from long-term national investment in shared digital research infrastructure and therefore have a substantial operational base on which to build EOSC services. PaNOSC illustrates the limitations of a model that relies heavily on in-kind contributions from participating institutions. Such an approach can support initial collaboration and allow a broad consortium to emerge, but it makes it more difficult to guarantee stable services, formal service-level agreements and long-term development. The report repeatedly emphasises that an EOSC Node requires much more than servers and software. Significant ongoing effort is needed for service management, onboarding, documentation, security, user support, standards development and participation in Federation-level coordination.

The report also examines the challenges associated with sensitive research data. CSC and SURF provide some of the most mature examples of secure environments for controlled storage and analysis. Their experience shows that sensitive-data services depend on more than technical security controls. They also require formal security governance, auditing, incident management, clearly assigned responsibilities and continuous compliance work. For emerging nodes, the report therefore advises against making complex sensitive-data environments part of the initial minimum viable configuration unless the necessary trusted infrastructure and organisational capacity already exist.

Community engagement is treated as an equally important success factor. The report explicitly notes that researchers are not interested in infrastructure for its own sake. They engage with EOSC when it solves a concrete problem, reduces barriers to accessing computing or data, simplifies the discovery of software and services, or makes a research workflow easier to reproduce. Training, documentation, consultation, user forums, pilot use cases and onboarding support are therefore presented as integral node functions rather than secondary communication activities. EOSC Finland places particular emphasis on involving users in service development, while SURF demonstrates the value of environments in which researchers can obtain a ready-to-use tool or virtual research environment without having to understand the underlying infrastructure.

For MATSCI-NASU, the report is particularly valuable as practical evidence supporting the thematic node model. The PaNOSC experience shows that a thematic node can generate substantial added value through discipline-specific workflows, metadata, catalogues, remote processing and access to distributed resources without centralising the underlying data or computing infrastructure. At the same time, the combined experience of the nodes studied shows that technical integration must be supported by clear governance, user support, service management and a realistic sustainability model. The practical value of a thematic EOSC Node therefore comes from combining disciplinary expertise with federating capabilities and reproducible research scenarios that respond to real needs of the scientific community.