The agrifood sector has invested heavily in generating data. The return on that investment depends heavily on whether data from different systems can be understood together. Right now, it mostly cannot.
Across the agrifood value chain, the same term routinely means different things in different systems. 'Yield' may refer to fresh weight or dry weight, field-level or enterprise-level, pre- or post-storage loss. 'Organic' carries different legal meanings in different jurisdictions. 'Variety', 'cultivar', 'lot', and 'batch' are used interchangeably in some systems and with distinct technical precision in others. These inconsistencies are not a matter of poor data quality in the conventional sense, since the measurements may be accurate. The problem is that the concepts themselves have not been formally defined and the relationships between them have not been agreed upon.
Why this matters — and why now
The consequences are concrete and costly. Data cannot be combined reliably across systems without manual curation. AI and machine learning models trained on inconsistently defined data produce unreliable outputs. Traceability chains break at the points where one organisation's system meets another's. Sustainability calculations yield different answers depending on whose definitions are used. Regulatory compliance data cannot be automatically validated. And the cost of integration (building and maintaining the translation layers between systems) falls on every participant in the chain, repeatedly.
Ontologies are the established formal solution to this class of problem. They provide machine-readable definitions of concepts, properties, and the relationships between them; a shared vocabulary that systems can reason with, not just read. Substantial ontology work already exists across agriculture, food systems, sustainability, and related domains. The barrier is that it remains fragmented, inconsistently applied, and not yet anchored to international standards that give it authority and durability.
AHG 13 has been established to change that.
"The digital transformation of agrifood systems is not constrained by a shortage of data or analytics capability. It is constrained by a shortage of agreed meaning. Ontologies are the instrument for supplying it."
What this work will make possible
AHG 13 will review existing agrifood ontologies, identify where current coverage is inadequate or fragmented, develop a Technical Report establishing the authoritative landscape, and where clear gaps exist, draft New Work Item Proposals for international standards. The use cases below illustrate where agreed ontologies deliver measurable value across the stakeholder community.
For farmers and field-level operators
- Farm management software, advisory apps, and precision agriculture platforms can exchange data without manual re-entry or translation — because the concepts they use (field, crop, variety, input, event) are defined consistently and in a way that can be machine interpreted without ambiguity.
- Decision-support tools and AI-powered recommendations become more reliable when the data they are trained and operated on carries consistent, formally defined meaning across sources and regions.
- Record-keeping for traceability, subsidy claims, and certification is simplified when farm systems share the same ontological foundation as the downstream systems that consume those records.
For technology and platform developers
- Build data models and APIs against internationally standardised ontological foundations, reducing the cost of integration with partner systems and third-party data sources.
- Enable interoperability with any platform that aligns with the same ontology, expanding the addressable market without the overhead of bespoke connector development for each integration.
- Train AI and machine learning models on datasets whose semantic consistency is guaranteed by ontological standards, improving reliability and reducing the cost of data preparation.
- Reduce the burden of regulatory compliance by aligning data models to the same ontological definitions used in official reporting and certification frameworks.
For supply chain operators — packers, processors, logistics, and traders
- Receive data from suppliers (farm records, lot identifiers, quality attributes, certification status) in a form whose meaning is unambiguous, eliminating the translation layer that currently sits between most supply chain systems.
- Conduct traceability investigations faster and with greater confidence when every link in the chain uses consistently defined lot, batch, and event concepts.
- Aggregate sustainability data (carbon intensity, water use, biodiversity impact) across suppliers and geographies using a common definitional framework, rather than reconciling incompatible methodologies.
- Respond to emerging regulatory requirements for supply chain transparency (deforestation, forced labour, environmental due diligence) more efficiently when the underlying data concepts are formally defined and consistently applied.
For retailers, brands, and consumer-facing businesses
- Make credible, specific claims about product provenance, production practices, and sustainability attributes when those claims are backed by data grounded in formally defined, internationally recognised ontologies rather than loosely defined internal classifications.
- Verify supplier claims more reliably (and defend them to auditors and regulators) when the concepts in question have agreed, machine-interpretable definitions.
- Build consumer-facing transparency tools (QR code lookups, provenance displays, sustainability scores) on a data foundation that is consistent across product categories and geographies.
For regulators, certification bodies, and standard-setting organisations
- Draft regulations and certification criteria using formally defined concepts that can be automatically checked against operator data, reducing compliance cost for industry and enforcement cost for regulators.
- Aggregate and compare data submitted by regulated entities across different jurisdictions without needing to manually reconcile definitional differences.
- Align national and regional standards with international ontological frameworks through ISO, reducing the fragmentation that currently multiplies compliance obligations for businesses operating across borders.
- Ground sustainability frameworks (environmental footprint methodologies, GHG accounting standards, biodiversity impact assessments) in formally defined data concepts that make cross-study comparison and policy evaluation tractable.
For researchers and AI developers
- Access and combine datasets from different research programmes, geographies, and institutional sources without the semantic harmonisation effort that currently consumes a disproportionate share of research capacity.
- Develop AI models for crop disease detection, yield prediction, supply chain risk, and climate adaptation on datasets whose conceptual consistency is guaranteed, not assumed.
- Contribute ontology developments through a standards pathway that gives them international authority and durability, rather than leaving them as one-off research outputs that are rarely reused.
The foundation beneath everything else
AHG 13 occupies a distinctive position within ISO/TC 347's work programme. Where other AHGs (e.g. on smart apiculture, smart irrigation, livestock, crops, and field management) are developing domain-specific data models and standards, AHG 13 addresses the semantic layer that can make all of those standards coherent with one another and with the broader agrifood data ecosystem.
An ontological foundation does not replace domain or reference data standards. It complements and completes them, ensuring that when a livestock data model uses the concept 'animal', an irrigation standard uses 'field', and a traceability system uses 'lot', those terms can be linked to formally defined, machine-interpretable meanings that allow data to flow reliably across domain boundaries.
This is also why AHG 13's output takes the form of a Technical Report before any NWIPs. The landscape of existing agrifood ontologies is substantial and complex. Rushing to standardise without a thorough assessment of what exists, what works, and where the genuine gaps are would risk creating yet more fragmentation rather than resolving it. AHG 13's approach is deliberately evidence-first.
AHG 13 does not build ontologies from scratch. It establishes the framework for how existing and future agrifood ontologies can be assessed, aligned, and anchored to international standards — making the whole more than the sum of its parts.
How the work will proceed
AHG 13 operates within ISO/TC 347 — the technical committee for data-driven agrifood systems. Its co-conveners bring complementary expertise that spans the full scope of the work: Mr. Qian Heng (Chair, Standard Institute of Emerging Technologies and Innovations, Qilu University of Technology; Convener, ISO/IEC JTC1/WG11 Smart City) brings over 30 years of standards leadership in IoT, data capture, and food traceability. Dr. Yongchao Gao (Professor, Qilu University of Technology; Convenor, IEC Smart Cities Reference Architecture) brings deep expertise in knowledge engineering, ontology development, AI systems, and international standardisation, including specific work on food supply chain ontologies and city data models.
The group will begin by developing representative agrifood use cases that ground the technical work in practical reality, then conduct a systematic review of existing ontologies across agriculture, food systems, sustainability, and traceability domains. The findings will be documented in an ISO Technical Report. Where clear standardisation gaps are identified, NWIPs will follow. Membership is open to all ISO/TC 347 members and liaison organisations.
The AHG aims to have its Technical Report and any NWIPs ready before March 2027.
