Horizon Europe · AIDAVA · CC0 Open Standard

Turning fragmented health data
into a shared understanding.

Across healthcare, research, and AI, we collect vast amounts of data — but systems still can't reliably understand each other. SULO is an open, minimal standard that gives data from any source a common machine-readable meaning.

See EU Policy Relevance Join the Community Read the Paper
3+
Health data standards bridged
(FHIR · OMOP · SPHN)
CC0
Fully open license
No restrictions
OWL 2
W3C standard
Machine-readable
2025
Published at FOIS/JOWO
Peer reviewed

"Without a shared semantic layer, we don't have a data problem — we have a meaning problem. Systems exchange bytes, but don't agree on what those bytes mean."

The Problem

Data doesn't speak for itself.

Health data is abundant. Shared meaning is not. Three root causes drive the interoperability crisis.

🗂️
Everyone models the world differently
One hospital says "hasDiagnosis." Another says "diagnosis_code." A third encodes it as a spreadsheet column. These mean the same thing — but machines can't tell. Every institution reinvents the wheel.
⚖️
Schemas are too weak; ontologies too heavy
Simple formats (JSON, SQL) lack formal semantics — machines can't infer meaning. Full upper ontologies (e.g. BFO, DOLCE) are powerful but complex, leading to low adoption and modeling errors.
🔧
Integration becomes bespoke and fragile
Every data partnership requires custom mappings, hand-crafted adapters, and one-off fixes. This doesn't scale. It breaks when systems change. It costs institutions enormous time and money.
🤖
AI systems inherit the confusion
If the underlying data is inconsistently structured, AI models learn inconsistent patterns. The result: poor generalization across hospitals, unreliable outputs, and limited trust — especially in cross-border settings.
WITHOUT SULO
FHIR
MedicationAdministration.subject
OMOP
drug_exposure(person_id=123)
SPHN
sphn:hasAdministeredDrug
Three vocabularies. Same fact. Machines can't align them automatically.
WITH SULO
:DrugAdmin a sulo:Process ;
  sulo:hasParticipant :InputRole ;
  sulo:hasParticipant :OutputRole .

# PRO pattern — role bearers
:InputRole a sulo:Role .
:Patient sulo:hasFeature :InputRole .

:PrescriptionNote a sulo:InformationObject ;
  sulo:hasFeature :OutputRole ;
  sulo:refersTo :AmoxicillinAdminCondition .
:OutputRole a sulo:Role .
:AmoxicillinAdminCondition a sulo:Feature ;
  owl:sameAs rxnorm:723 .

# inferred: DrugAdmin hasParticipant Patient
# via hasParticipant ∘ isFeatureOf⁻¹ → hasParticipant
One shared pattern. Any standard can map to it. Machines reason consistently.
The Solution

SULO: a minimal grammar of meaning.

SULO (Simplified Upper Level Ontology) provides a small, carefully chosen set of building blocks — simple enough to adopt correctly, formal enough for machines to reason over.

01 — MINIMAL
8 classes. 5 relations.
SULO's power comes from its minimalism. Object, Process, Feature, Collection, Quantity, Role, InformationObject, Time. That's it. No sprawling hierarchies. No ambiguity through over-specification.
02 — FORMAL
Machine-readable and verifiable.
Expressed in OWL 2 — the W3C web standard for ontologies. Data aligned with SULO can be automatically validated and reasoned over. Inconsistencies are caught before they cause problems downstream.
03 — REUSABLE
Forces consistency through shared vocabulary.
Instead of inventing new relations ("hasDiagnosis", "hasIngredient"), everyone expresses knowledge using the same small vocabulary. Reuse, not reinvention — that's how interoperability scales.
04 — OPEN
CC0 license. No barriers.
SULO is released under CC0 — the most permissive open license. Any institution, project, or product can adopt it without restrictions, fees, or attribution requirements. Built for public benefit.
CHALLENGE WITHOUT SULO WITH SULO
Cross-institutional data sharing Custom mappings per partner pair
Months of engineering per integration		 Map once to SULO
Reuse across all SULO-aligned partners
AI training across hospitals Inconsistent input → brittle models
Poor generalization, limited trust		 Consistent semantic structure
Robust, explainable, cross-site AI
Regulatory reporting & audit Manual reconciliation of formats
Error-prone, expensive		 Automated validation
Consistent, traceable, machine-verifiable
Secondary use of health data Data silos limit research scope
Reuse requires re-engineering		 Semantically FAIR data
Findable, accessible, interoperable, reusable
EU Policy Alignment

Built for Europe's data strategy.

SULO directly enables the semantic interoperability layer that European data legislation demands but does not yet provide out of the box.

EUROPEAN HEALTH DATA SPACE · EHDS
The EHDS mandates cross-border health data access.
The EHDS Regulation requires Member States to enable secondary use of electronic health data across borders for research, policy, and public health. This only works if data from different EHR systems, registries, and hospital networks carries the same meaning.

SULO provides the semantic alignment layer the EHDS needs: a common vocabulary that maps FHIR, OMOP, SPHN, and other national standards to a shared ontological foundation — without replacing them.
EHDS Regulation → ec.europa.eu ↗
EUROPEAN DATA SPACES · DIGITAL STRATEGY
Data Spaces require machine-understandable semantics.
The European data space strategy (health, mobility, agriculture, energy, and others) requires that data shared across space boundaries is not just technically interoperable but semantically interoperable — machines must understand what the data means, not just its format.

SULO is precisely the kind of lightweight, domain-agnostic ontology that data space connectors need to attach meaning to the data they exchange.
European Data Spaces → digital-strategy.ec.europa.eu ↗
HOW SULO SUPPORTS EU DATA GOALS
FAIR Data Principles
SULO-aligned data is Findable, Accessible, Interoperable, and Reusable — the FAIR principles the EU mandates for funded research data.
GDPR & Data Governance Act
Formal semantics support data lineage and provenance tracking — making it easier to demonstrate compliance with processing constraints.
AI Act Transparency
AI systems trained on SULO-structured data can more easily explain their reasoning — supporting the explainability requirements of the AI Act.
Digital Health Priority
SULO is developed under the EU Horizon AIDAVA project, directly addressing the EU's digital health and AI-in-health strategic priorities.
Community & Governance

Open by design. Community-led.

SULO is not a product — it's a community standard, developed transparently and governed by its users.

Monthly working sessions

Every first Wednesday of the month, the SULO community convenes to work through modeling challenges, validate use cases, and evolve the ontology. Participation is open to institutions, researchers, clinicians, and policy experts.

Join Next Meeting Technical Documentation →
CORE TEAM · Supported by EU Horizon AIDAVA (GA 101057062)
🇪🇺
This work is supported by the AIDAVA project (Grant Agreement 101057062), part of the European Union's Horizon Europe research and innovation programme.
aidava.eu ↗