Internet of FAIR Data & Services

• Internet of FAIR Data and Services is a conceptual infrastructure that applies FAIR principles to enable automated data discovery, integration, and reuse.
• It leverages standardized identifiers, semantic models, and modular digital objects (FDOs) to facilitate cross-domain interoperability and workflow automation.
• IFDS architectures incorporate federated governance, decentralized access, and scalable compute-to-data paradigms, fostering reproducible and secure research.

The Internet of FAIR Data and Services (IFDS) is an emerging paradigm for the global organization, access, exchange, and reuse of digital data and associated computational services in a way that fully realizes the FAIR principles: Findability, Accessibility, Interoperability, and Reusability. IFDS envisions a machine-actionable and semantically robust infrastructure where domain-specific, cross-domain, and computational research resources are governed, linked, and orchestrated through standardized identifiers, protocols, and semantic models. This approach supports not only the technical requirements for data integration and workflow automation, but also the legal, organizational, and cognitive needs of a globally distributed and multidisciplinary research ecosystem.

1. Conceptual Foundation and Motivations for IFDS

The motivation behind IFDS arises from shortcomings in data management across scientific, industrial, and business contexts: fragmentation of data repositories, a lack of scalable and interoperable architectures, and insufficient attention to semantic and operational interoperability. The IFDS paradigm is grounded in the trajectory of the FAIR (Findable, Accessible, Interoperable, Reusable) principles, first articulated to address these challenges in research data stewardship. While the original FAIR principles provided the conceptual foundation, further critical advancements—such as modularity of data into FAIR Digital Objects (FDOs), semantic unit modeling, and federated governance—are required for operationalizing IFDS at a planetary scale (Blumenroehr et al., 27 Nov 2024, Beyvers et al., 28 Apr 2025, Vogt et al., 30 Sep 2025).

Recent research extends these principles to address semantic interoperability at the schema and terminology levels (FAIR 2.0) (Vogt et al., 6 May 2024), introduces cognitive interoperability and human explorability (FAIREr principles) (Vogt, 2023), and formalizes these notions through granular architectures and ontological frameworks (Santos et al., 2023, Vogt et al., 30 Sep 2025).

2. Architectural Components and Formal Models

Central to IFDS is the abstraction and encapsulation of data, metadata, and computational methods as FDOs—modular digital entities with globally unique and persistent identifiers (GUPRIs). These are defined by formal data models specifying kernel information profiles (KIPs), machine-actionable attributes, and operational methods (Blumenroehr et al., 27 Nov 2024, Zoubia et al., 6 Feb 2024). The general structure involves a networked architecture with layers for governance, data, services, and applications (Beyvers et al., 28 Apr 2025):

• Governance/Access Control Layer: Decentralized, attribute-based access control, federated identity management, machine-readable Data Usage Agreements.
• Data Layer: Distributed, peer-to-peer connections of "Data Locations" and "Compute Nodes" using standardized adapters.
• Service Layer: Automated ETL, semantic enrichment, data aggregation, and provenance tracking.
• Application Layer: Web dashboards, APIs, and tools enabling transparent user interaction.

The formal model for FDOs is mathematically specified: each FDO $f$ in the set $F$ is instantiated by a unique KIP $p \in P$ as an information record $R_f$, with key–value pairs for typed attributes, and registered with a persistent identifier $i$ (often using the Handle system) (Blumenroehr et al., 27 Nov 2024):

$$\forall f \in F, \exists p \in P:\ \text{Instantiate}(f, p) = R_f$$

$$\forall f \in F:(\exists p, p' \in P:\ \text{Instantiate}(f, p) = R_f \wedge \text{Instantiate}(f, p') = R_f) \implies p = p'$$

Such rigor in the data model enables strict abstraction, encapsulation, and consistent operations, forming the basis for uniform, machine-actionable decisions across heterogeneous domains.

3. Semantic Units, Interoperability, and Cognitive Layers

Recent advances position modular "semantic units" as the organizing principle for FAIR semantics and IFDS infrastructures (Vogt et al., 30 Sep 2025, Vogt, 2023). A semantic unit is an atomic or compound chunk of information—each with a GUPRI, provenance, and schema declaration—that can be independently referenced, cited, or exchanged. Semantic units may be mapped onto FDOs, facilitating both technical and cognitive interoperability:

• Statement Units: Represent single propositions (e.g., experimental facts, assertions).
• Compound Units: Aggregate statement units into more complex descriptions (e.g., all properties of a specimen).
• Granularity and Modularization: Layered structure similar to biological organization (atom, molecule, cell, tissue, organ) is leveraged as a metaphor—each unit is a "semantic membrane" in the information ecosystem.

Semantic interoperability in IFDS requires both terminological interoperability (linking equivalent terms across vocabularies with mappings such as owl:sameAs) and propositional interoperability (schema crosswalks aligning statement structures), as formalized in the FAIR 2.0 extensions (Vogt et al., 6 May 2024). Cognitive interoperability, or human explorability, is added as a distinct layer to facilitate intuitive navigation of complex, modular knowledge graphs via mind-map views, natural language summaries, and functionality for “overview first, zoom and filter, then details-on-demand” (Vogt, 2023).

4. Federated and Distributed Technical Infrastructures

The realization of IFDS architecture is heavily federated—preserving both domain-specific autonomy ("domain sovereignty") and cross-domain integration (Beyvers et al., 28 Apr 2025, Karsch et al., 2022):

• Horizontal Scalability: Peer-to-peer network topologies support dynamic joining/leaving of nodes, load balancing, and distributed processing. Total capacity $C_\text{total}$ scales linearly with the number of nodes $n$: $C_\text{total} \propto n$.
• Compute-to-Data and Data-to-Compute: Advanced architectures enable either local data processing (minimizing movement) or centralized compute on transferred data; allocation is optimized to minimize a cost function $$C = f(\text{processing time}, \text{data transfer cost}, \text{privacy risk})$$ (Beyvers et al., 28 Apr 2025).
• Persistent Identifiers and Provenance: All digital entities (data, code, workflows) carry persistent IDs (DOIs, Handles, SWHIDs), comprehensive metadata, and linkage to provenance information, supporting traceability and auditability (Santos et al., 2023, Vogt, 2023, Wilkinson et al., 21 May 2025).
• Security and Access Control: Governance frameworks employ ABAC, federated identities, and cryptographic validation (certificates, signed agreements) (Grossman et al., 2022, Beyvers et al., 28 Apr 2025).

5. FAIR Services and Ecosystem Integration

FAIRification in IFDS is achieved not only by packaging data as FDOs but also by orchestrating a family of modular services (Vogt et al., 6 May 2024, Wilkinson et al., 21 May 2025):

• Terminology Service: Registers vocabularies, ontologies, and mappings; ensures terminological harmony across domains.
• Schema Service: Registers and provides crosswalks for data schemata, enabling machines to reconcile structural or logical heterogeneity.
• Operations Service: Declares and serves executable functions (unit conversions, transformations, query templates) linked to schema types.
• Workflow Services: Supports FAIR computational workflows via persistent IDs, machine-actionable metadata, workflow packaging (e.g., RO-Crate), registries, testing, and execution environments (e.g., EOSC-Life Workflow Collaboratory) (Wilkinson et al., 21 May 2025).
• Metadata and Registry Services: Universal catalogues for data, workflows, tools, and services (e.g., WorkflowHub, ResultsDB, the FDO Manager (Zoubia et al., 6 Feb 2024)).

Such services are designed to support automation, provenance, interoperability, and reproducibility, while accommodating diverse technical stacks (RDF, JSON, YAML, REST APIs, etc.).

6. Real-World Implementations and Sectoral Demonstrators

Significant progress has been achieved through pilot projects and domain-specific implementations:

• Astronomy and Physics: Federated hybrid clouds and VO-compliant infrastructures integrating IVOA VOSpace for standardized storage and access, supporting seamless, globally distributed scientific collaboration (Bertocco et al., 2018, Karsch et al., 2022).
• Life Sciences and Health: Decentralized platforms (e.g., GADDS) employ blockchain-based metadata validation, distributed object storage, and version control; clinical and biomedical data exchange achieved under privacy-constrained SAFE environments (Vazquez et al., 2021, Grossman et al., 2022, Glombiewski et al., 6 Dec 2024).
• Simulation and Materials Science: nanoHUB’s Sim2Ls and ResultsDB provide reproducible, API-driven simulation workflows, indexed with DOIs and metadata for full FAIR compliance and educational integration (Mejia et al., 2023).
• Ontological Frameworks and Data Science: Rich BFO-compliant ontologies underpin knowledge graphs for DS and AI research, enhancing interoperability and reusability (Gesese et al., 16 Aug 2024).
• Cross-Sectoral Data Spaces: The FAIR Data Spaces project demonstrates orchestration across health, biodiversity, and engineering, leveraging Gaia-X specifications, microservices, and legal/ethical frameworks for sovereign cloud-native data integration across sectors (Glombiewski et al., 6 Dec 2024).
• FAIR and Federated Data Ecosystems: Layered, decentralized approaches balance domain autonomy with interoperability, using standardized semantic enrichment, adaptive metadata, and robust auditability (Beyvers et al., 28 Apr 2025).

7. Challenges, Limitations, and Future Outlook

Critical challenges for IFDS include managing semantic heterogeneity (no universal ontology or schema), aligning governance norms, balancing domain-specific requirements with cross-domain interoperability, ensuring dynamic provenance, addressing security for sensitive data, and supporting granular, modular citation and usage (Blumenroehr et al., 27 Nov 2024, Vogt et al., 6 May 2024, Beyvers et al., 28 Apr 2025). The trajectory of research suggests further:

• Expansion of community-driven FAIR 2.0 standards (F5–F7), placing greater emphasis on explicit entity/schema mappings and services (Vogt et al., 6 May 2024).
• Development of granular, modular architectures based on semantic units to enable AI-ready infrastructures and citation-granular scholarly communication (Vogt et al., 30 Sep 2025).
• Increased automation of metadata extraction, provenance capture, and FAIRification in computational workflows, leveraging evolving registries and packaging standards (Wilkinson et al., 21 May 2025).
• Accelerated convergence of technical implementations through global standardization forums (e.g., RDA, FDO Forum), rigorous benchmarking, and cross-platform adoption.

Ongoing harmonization and consensus-building are expected to further align sectoral and disciplinary disparities, eventually enabling seamless, machine-actionable, and cognitively accessible data and service ecosystems.

---

The Internet of FAIR Data and Services is a multilayered, modular, and service-centric infrastructure where data, software, workflows, and operational logic are increasingly represented as persistent, machine-actionable, and semantically grounded digital objects. By leveraging standardized IDs, metadata, governance frameworks, and granular semantic architectures, IFDS fosters scalable, transparent, and robust interdisciplinary research, supporting automation, provenance, security, and human cognition at global scale.