Interoperability

“Two systems are not interoperable because they can send each other files; they are interoperable when meaning survives the transfer. ”

Interoperability is often reduced to an API. One platform can send data and another can receive it, so the systems are declared connected. Yet successful transmission does not prove that the receiving organisation understands the same farmer, unit, status or event. Technical exchange is necessary, but interoperability exists only when the information can be used correctly.

The European Interoperability Framework defines interoperability as the ability of organisations to interact toward mutually beneficial goals through sharing information and knowledge, supported by business processes and data exchange between ICT systems. The definition begins with organisations and goals, not software. Technology serves an agreed relationship. The Framework describes several layers.

Legal interoperability concerns whether laws, contracts and permissions allow exchange and reuse.

Organisational interoperability aligns responsibilities and business processes. Semantic interoperability preserves meaning. Technical interoperability concerns interfaces, protocols and infrastructure. Failure at any layer can stop the data from working. Consider a farmer record moving from a cooperative platform to a buyer. Both systems may have a field called farm size.

One stores hectares measured by GPS; the other stores farmer-reported acres. The exchange succeeds technically and fails semantically. A risk model may then calculate an implausible yield or land-use result with no visible error. Identifiers create similar problems. Names are unstable across spelling, language and household convention. One system may identify a household, another a person and another a farm business.

Matching records by name can merge different people or duplicate the same one.

Interoperability requires shared identifier rules or explicit methods for resolving identity. Business meaning also depends on time. Certified may refer to a current certificate, a historical status at delivery or a claim inherited through chain-of-custody accounting. If date and scope do not travel with the field, the receiving system can turn valid data into a false claim.

Metadata are part of the information, not optional documentation. GS1's EPCIS illustrates the value of a common event language. Trading partners can share information about what happened, when, where and why using agreed structures. The standard does not force every organisation to use the same internal software. It creates a common boundary through which diverse systems can communicate.

Interoperability is therefore not the same as centralisation. A single platform may simplify exchange and concentrate control, create dependency or require every participant to adopt one vendor's model. Federated systems can preserve autonomy where standards, governance and trust allow data to move. The appropriate architecture depends on purpose, risk and power. Privacy and confidentiality remain constraints.

The ability to exchange data does not create permission to share it. Legal basis, purpose limitation, access control and commercial rights should travel with the information. Interoperable systems need policy enforcement as well as technical compatibility. Versioning is another test. Standards, schemas and code lists change. An interface may continue operating while meaning drifts between versions.

Governance should define backward compatibility, deprecation, validation and responsibility for failed translations. The discipline is to test use, not connection. Can the receiving organisation make the intended decision with the same meaning, unit, identity, time and permissions? Interoperability is achieved when data remain trustworthy across the institutional boundary, not when the transfer log shows success.

Practical application

Define the shared goal and map legal, organisational, semantic and technical requirements. Agree identifiers, vocabularies, units, dates, statuses, metadata and error handling before building interfaces. Use open standards where suitable and document deviations. Test exchanges with real edge cases, not only ideal records. Maintain schema versions, validation rules and responsibility for correction.

Share only data necessary for the purpose and attach permissions, provenance and confidence information.

Why it matters

Sustainability evidence is distributed across farms, governments, schemes, traders and buyers. Without interoperability, organisations repeatedly recollect data, lose meaning and create inconsistent claims. Good interoperability reduces burden and supports decisions across systems without requiring one owner to control all data.

Common misconception

Interoperability is often equated with integration or file exchange. Systems can be connected and still misunderstand each other. Interoperability requires alignment of law, process, meaning and technology so that exchanged data can be used correctly.

Connections

Traceability Systems generate and exchange event and master data. Data Governance assigns rights and responsibilities across organisations. Data Quality and standardised Sampling determine whether transferred information remains fit for purpose.

A question worth asking

When your systems exchange a sustainability field, what proves that both sides mean the same entity, unit, period, status and permitted use?

Selected references

European Commission. 2017. New European Interoperability Framework. European Union. 2024. Regulation (EU) 2024/903 Laying Down Measures for a High Level of Public Sector Interoperability across the Union. GS1. 2022. EPCIS and Core Business Vocabulary 2. 0. World Wide Web Consortium. 2017. Data on the Web Best Practices. ISO/IEC 19941:2017. Cloud Computing - Interoperability and Portability.