FAIR way forward: building an EOSC-aligned interoperability roadmap for the INFRA-ART Spectral Library -

by Ioana Maria Cortea — Last updated on July 14, 2025 — Reading time: 14 min

Highlights

» Interoperability is essential for making research data discoverable, reusable, and machine-actionable across domains.
» FAIR data needs interoperable systems—not just within disciplines, but across institutional and national boundaries.
» EU policies and EOSC frameworks are creating a common foundation for data exchange through clear standards and regulations.
» EOSC IF and CDIF offer structured guidance for aligning metadata, systems, and governance with FAIR and Open Science goals.

The INFRA-ART FAIR Data Management series examines the principles, governance frameworks, and practical approaches that make research data more FAIR and openly accessible. Each article explores a specific aspect of responsible data stewardship—covering standards, tools, workflows, and real-world implementation experiences—drawing on examples from the INFRA-ART Spectral Library’s FAIR journey. By combining practical guidance with policy context and case studies, the series aims to support researchers, data stewards, and infrastructure developers in adopting best practices for effective, responsible, and sustainable data management.

Article sections

» The data interoperability challenge » Achieving interoperability of data, metadata, and applications » Key interoperability requirements and recommendations » Developing an interoperability policy roadmap for INFRA-ART » Further reading and resources

The data interoperability challenge

Digital data serve as both the outputs of research and the foundation for new investigations, studies, and educational initiatives. To enable data to be reused in novel ways—often beyond the original intent of its creators—it must be easily found, understood, and combined across diverse formats and sources. Achieving this depends fundamentally on interoperability—“the ability of two or more systems or components to exchange information and to use the information that has been exchanged”.

To enable seamless and efficient information exchange across organizations and IT systems—and to ensure a shared understanding of data—interoperability requires both semantic alignment (ensuring data refers to the same concepts) and the resolution of structural differences between data models. Data integration is the process of combining heterogeneous data and their structural information to create a unified representation and mapping, enabling seamless access to all available data. More generally, interoperability goes beyond data alone, also relying on compatible hardware, software, and communication protocols to ensure accurate interpretation across systems and institutional boundaries.

According to the Digital Curation Centre (DCC) interoperability can be divided into five different conceptual levels:

No data exchange: Systems operate in isolation without any data sharing.
Unstructured data exchange: Sharing of human-readable, unstructured data (e.g., free text).
Structured data exchange: Exchange of structured data intended for manual and/or automated handling, but still requiring manual processes for compilation and dispatch.
Seamless sharing of data: Automated data sharing within systems based on a common exchange model, facilitating smoother integration.
Seamless sharing of information: Universal interpretation of information through cooperative data processing, enabling full interoperability.

Interoperability offers a wide range of benefits. It improves consistency and usability of data across technological and institutional boundaries. It drives the adoption of standards that enhance data quality and transparency, accelerates data transfer, reduces redundancy and costs, and enables novel data uses—opening the door to unanticipated discoveries and innovation. These benefits are especially critical as research becomes more data-intensive, collaborative, and interdisciplinary.

The European Commission’s Rolling Plan for ICT Standardisation emphasizes the critical role of data interoperability in realizing a unified European data space. This initiative aims to facilitate seamless data sharing and reuse across various sectors and borders, aligning with the FAIR principles. The EU’s data strategy underscores the necessity of data interoperability and quality, particularly for the effective deployment of AI technologies. It identifies significant challenges in combining data from diverse sources, both within and across sectors. To address these, the strategy advocates for the adoption of standardized, compatible formats and protocols to ensure coherent and interoperable data processing. Commission Implementing Regulation (EU) 2023/138 specifies the technical requirements for publishing high-value datasets. It emphasizes the importance of machine-readable formats and APIs to enhance data discoverability and usability, thereby strengthening open data policies across EU member states. These measures collectively aim to overcome existing interoperability barriers, fostering an environment where data can be efficiently shared and utilized across different platforms and sectors.

Achieving interoperability of data, metadata, and applications

In the context of the FAIR guiding principles, interoperability is defined as “the ability of data or tools from non-cooperating resources to integrate or work together with minimal effort”. This definition captures the core challenge faced by researchers, institutions, and infrastructure providers: making diverse data systems speak the same language.

The European Commission’s Turning FAIR into Reality report underscores the role of interoperability frameworks in defining community standards and best practices—covering data formats, metadata standards, tools, and infrastructures. These frameworks must also accommodate cross-disciplinary data exchange, especially in high-priority interdisciplinary research. Developed under the WorldFAIR project, the Cross-Domain Interoperability Framework (CDIF) provides practical, standards-based guidance for implementing FAIR principles—particularly in complex, multi-disciplinary contexts where domain-specific conventions often clash. CDIF is not a new standard, but a set of implementation profiles and best practices that align existing, widely adopted metadata standards and technologies to support five core FAIR data functions: discovery, access, controlled vocabularies, data integration, and universals.

Within the European Open Science Cloud (EOSC), achieving interoperability is essential to allow its federated services to deliver added value to users across disciplines. EOSC services depend on the efficient exchange and interpretation of digital objects, regardless of their origin or scientific domain. For this to work, systems must understand the metadata associated with digital objects, including access restrictions, provenance, and usage requirements. The EOSC Interoperability Framework (EOSC IF) is built to support this vision. It draws on the European Interoperability Framework (EIF)—originally designed for public services—and adapts its four-layer structure to the research domain:

Technical interoperability: Refers to the ability of different IT systems and software applications to communicate, exchange, and process data automatically and effectively without manual intervention. It involves infrastructures and applications that facilitate this interaction, including interface specifications, data integration services, and secure communication protocols.
Semantic interoperability: Ensures that exchanged data carries a shared, unambiguous meaning, allowing systems to interpret and use the information correctly—even when they don’t know each other’s internal logic. It depends on common metadata standards, ontologies, thesauri, and clearly defined vocabularies. Human understanding and alignment are also crucial, especially in establishing shared Service-Level Agreements (SLAs).
Organizational interoperability: Refers to how organizations align their business processes, roles, and expectations to achieve shared, mutually beneficial objectives, particularly those related to Open Science. It includes providing clear documentation, defining who manages interoperability services (like PIDs or service registries), and ensuring services are user-focused.
Legal interoperability: Enables the reuse of data across sources without conflicts in licensing or access rights. It ensures that usage conditions are clearly defined, ideally in machine-readable form, and that combined or derivative datasets do not violate any of the original licenses.

For each interoperability layer, a catalogue of problems, needs, challenges, and high-level recommendations has been identified to guide improvements and inform strategic actions. EOSC IF thus serves as a domain-specific extension of EIF, guiding researchers, institutions, and data services toward interoperable, FAIR-compliant practices that work across disciplines and infrastructures.

Key interoperability requirements and recommendations

The EOSC IF outlines essential requirements and actionable recommendations to enable effective and FAIR-aligned interoperability across the European Open Science Cloud. These are structured along the four interoperability layers—technical, semantic, organizational, and legal—each addressing specific challenges and offering targeted guidance for improvement. These guidelines are intended not only to enhance interoperability within EOSC, but also to enable cross-disciplinary, cross-institutional, and cross-border data exchange, aligning with broader FAIR principles and Open Science goals. The framework is designed to evolve with community input and technological advancement.

Key interoperability challenges and recommendations across EOSC IF layers

Layer	Key issues	Recommendations
Technical interoperability	• Fragmented authentication/ authorization systems • Incompatible data formats • Disparate Persistent Identifier (PID) policies • Difficulty in discovering data at multiple levels of granularity	• Use open specifications for EOSC services • Adopt a unified security and privacy framework (e.g., AAI) • Provide coarse- and fine-grained search tools • Ensure easy access to heterogeneous data formats • Implement a clear and flexible EOSC PID policy • Define easy-to-understand Service-Level Agreements (SLAs)
Semantic interoperability	• Lack of shared definitions and semantic artefacts across disciplines • Poor documentation and discoverability of metadata and ontologies • Inconsistent or absent crosswalks between metadata schemas	• Publish semantic artefacts with open licenses and persistent identifiers • Provide thorough documentation with examples and diagrams • Develop and adopt a Minimum Metadata Model for cross-domain discovery • Maintain repositories and governance for semantic artefacts • Enable extensibility to accommodate discipline-specific metadata • Establish federation protocols for harvesting semantic artefacts
Organizational interoperability	• Absence of a governance framework for interoperability • Lack of standardized terms, use policies, and sustainability planning • Inconsistent service integration across organizations	• Align Rules of Participation with interoperability requirements • Recommend standard formats and metadata for service and data providers • Define permanent identifiers for organizations and their functions • Create certification mechanisms to assess interoperability maturity of services
Legal interoperability	• Incompatible licenses (e.g., CC-BY-SA vs. CC-BY-NC) • Unclear reuse rights and orphan datasets • GDPR constraints on personal and sensitive data • Discrepancies across national and institutional legal frameworks	• Use standardized, machine-readable licenses (preferably CC0 for metadata) • Clearly mark copyright status, license terms, and any reuse restrictions • Ensure GDPR compliance through “privacy by design” principles • Harmonize terms of use across repositories and align national legislation with EOSC • Support mechanisms for license updates and tracking

CDIF builds upon and complements the EOSC IF, providing more detailed, practical guidance for data infrastructures and repositories. By offering domain-neutral metadata profiles, CDIF helps reduce the many-to-many mappings typically required when integrating data from different fields. Instead, it promotes a many-to-one model—allowing diverse systems to map to a shared “lingua franca” for interoperability. It promotes technologies like JSON-LD, FAIR Digital Objects, and signposting, and is designed to evolve with the growing demands of AI, metadata governance, and large-scale federated data systems.

Developing an interoperability policy roadmap for INFRA-ART

FAIR-IMPACT is an EU-funded project advancing FAIR solutions across the European Open Science Cloud (EOSC). Through targeted support actions, FAIR-IMPACT helps research communities assess their current practices, identify gaps, and adopt FAIR-aligned strategies through expert guidance, peer learning, and tailored tools.

Under route 2, support offer #2 (Creating EOSC compliant interoperability policies based on EOSC Interoperability Framework (IF)), we had the opportunity to assess the FAIR maturity of the INFRA-ART Spectral Library data service, identify gaps, and align our data management practices with evolving EOSC interoperability requirements. This was achieved through a blend of expert guidance, peer-to-peer exchange, and hands-on tools tailored to different stakeholder needs. You can read our full FAIR Implementation Story via the FAIR-IMPACT Zenodo Community.

Using a set of questionnaires and checklist templates derived from the FAIRCORE4EOSC Compliance Assessment Toolkit (CAT), we evaluated our service’s alignment with technical and semantic EOSC IF recommendations. This structured self-assessment process revealed several critical gaps that affect the INFRA-ART Spectral Library’s compliance with the EOSC IF:

Absence of API endpoints for automated and programmatic data access.
Lack of persistent identifiers (PIDs) at the dataset (digital object) level.
Missing machine-readable metadata schemas for datasets.
Limited semantic artefact documentation (e.g., no mappings, diagrams, or example usages).
No mechanisms for harvesting of semantic resources by external catalogues or services such as OpenAIRE.

The identified gaps not only hinder the service’s overall interoperability and machine-actionability but also present significant barriers to integration with federated infrastructures. In light of these findings, two key interoperability challenges have been prioritized for the INFRA-ART Spectral Library data service:

Development of FAIR mappings and metadata alignment: Although the service uses structured metadata and applies open licensing (CC0), it currently lacks semantic mappings to external vocabularies or community standards such as DCAT, Schema.org, or DataCite metadata schema at object level. This absence of FAIR mappings restricts metadata interoperability and limits the ability of external services and platforms to interpret, aggregate, or repurpose the data. Without mappings, metadata is not machine-actionable across domains, reducing the visibility and reusability of datasets. Addressing this requires both the identification of relevant community standards (including semantic resources – ontologies/thesauri/terminologies/vocabularies) and the creation of FAIR mappings (crosswalks) between internal metadata elements and recognized schemas.
Integration with OpenAIRE: Currently, the INFRA-ART Spectral Library is not integrated into harvesting or discovery infrastructures such as OpenAIRE, which restricts its discoverability in the European research ecosystem. Integration with OpenAIRE will require the implementation of automated metadata exchange mechanisms and machine-readable dataset descriptions, in alignment with OpenAIRE Guidelines.

To address the technical and semantic interoperability gaps identified, a phased implementation roadmap has been developed. This roadmap acts as a living document, guiding the ongoing evolution of INFRA-ART’s services toward full FAIR and EOSC compliance. It reflects our strategic commitment to enhancing the visibility, machine-actionability, and reusability of the INFRA-ART Spectral Library’s datasets within and beyond EOSC-aligned infrastructures.

Interoperability roadmap for the INFRA-ART Spectral Library data service

Stage description	Estimated timeline	Key actions to be carried out	Persons responsible for implementation
Interoperability Policy Development	Short-term (1-3 months)	• Finalize and adopt the Interoperability Roadmap as an internal guiding document. • Engage with EOSC and domain-specific initiatives to ensure interoperability remains aligned with evolving best practices.	• INFRA-ART project coordinator
Technical Enhancements	Mid to long-term (6-24 months)	• Design and implement a REST API to enable machine-to-machine data access and integration with external infrastructures (e.g., OpenAIRE). • Establish a PID strategy for datasets using a recognized service (e.g., DataCite, Handle). • Begin exposing datasets with persistent identifiers and ensure all metadata is discoverable via the API. • Review and update the current technical documentation to include detailed descriptions of all supported data formats and access workflows.	• INFRA-ART project coordinator • IT team • Data curators
Semantic Artefact Development	Mid to long-term (6-24 months)	• Create and publish mappings (FAIR crosswalks) between internal metadata fields and established community standards. • Express metadata schemas in machine-readable formats (e.g., JSON, DCAT, XML schema). • Document existing semantic artefacts (controlled vocabularies, internal terms, and data models). • Submit semantic artefacts to trusted repositories such as FAIRsharing. • Develop example use cases, relationship diagrams, and usage guidance to support community reuse.	• INFRA-ART project coordinator • IT team • Data curators
Integration with External Infrastructures	Mid to long-term (6-24 months)	• Set up harvesting mechanisms (e.g., via an API endpoint) to enable the federation of metadata records into metadata aggregators and search engines. • Align metadata outputs with the OpenAIRE Guidelines for Data Archives to support future integration.	• INFRA-ART project coordinator • IT team

How to cite this resource

Cortea, I.M. (2025). FAIR way forward: building an EOSC-aligned interoperability roadmap for the INFRA-ART Spectral Library. Zenodo. https://doi.org/10.5281/zenodo.15879511

Associated metadata

Title	FAIR way forward: building an EOSC-aligned interoperability roadmap for the INFRA-ART Spectral Library
Author(s)	Ioana Maria Cortea
Affiliation	National Institute for Research and Development in Optoelectronics – INOE 2000
Resource type	Blog article
Keywords	FAIR data, FAIR implementation, research data management, interoperability framework, technical interoperability, semantic interoperability, EOSC guidelines, FAIR-IMPACT, FAIRCORE4EOSC, interoperability roadmap, research infrastructure, implementation story, case study, INFRA-ART Spectral Library
Target audience	researchers, data curators, data stewards, repository managers
License	CC0 1.0 Universal
Persistent Identifier (PID)	DOI: 10.5281/zenodo.15879511
File download	https://doi.org/10.5281/zenodo.15879511