Across Swedish industry, data sits in incompatible systems with the context that gives it meaning stripped away. Engineers misread each other's records, partners can't reconcile data across organisational boundaries, and AI systems fail silently on inputs they can't truly interpret. We're researching and disseminating the methods, means needed, maturity, and shared understanding that let any Swedish industrial actor make their data interpretable — by humans and machines alike.
The same datapoint can — and should — mean different things in different contexts. Semantic Data Readiness is the discipline of making that context explicit enough that AI systems, partner organisations, and future engineers can interpret data correctly even when their viewpoint differs from the one that produced it.
It is what lets an organisation climb the ladder from raw data to wisdom at enterprise scale — without losing meaning between rungs.
Raw signals, records, measurements — numbers and strings without anchored meaning.
Data placed in context: linked to entities, time, location, ownership — made queryable.
Patterns, rules, and constraints made explicit — reusable across systems, teams, and time.
Judgement at scale: machines and humans deciding what to do — and when not to act — on shared, trusted meaning.
A bearing-temperature reading from one factory means something different to every system that touches it. Without explicit context, AI can't reason across sources and humans waste time reconciling data that should already agree. This is one symptom among many — the underlying gap shows up across maintenance, traceability, compliance, engineering and AI workflows alike.
The capability to make context explicit: a maturity model, validated methods, reference architectures, and a national arena where Swedish industry learns together how to get there — not a single tool, but the means to create them.
With Semantic Data Readiness comes the ability, at enterprise scale, to:
… unambiguously and interoperably, for humans as well as machines.
— Tobias Ljungkvist, Project Manager
This project is being carried out with support from Vinnova (Sweden's Innovation Agency) within the programme Advanced Digitalization. Vinnova defines this as a system-changing initiative — a strategic project that changes structures in industry and society through advanced digital technology. Not a product. A national capability shift: shared methods, standards, and ways of working that let Swedish industry actually realise the value of its data.
Vinnova's six-year effect goal for this initiative is that common, scalable methods for industrial dataflows are nationally anchored and implemented in selected value chains — strengthening interoperability, data quality, and the competitiveness of Swedish industry. This project builds the foundation for that shift during Stage 1: methods, maturity model, reference architectures, and a coordinating arena. Stage 2 carries it into broad adoption.
The primary knowledge-generating package. Consolidates results from WP3 and WP4 into a generalisable framework: capability model, operationalised maturity scale, systemic failure-mode analysis, and a value model — all technology-agnostic.
How semantic structure affects whether agentic AI can reliably reason over maintenance data — and act on it.
→Making materials, batches, and processing steps describable across organisational boundaries in a circular economy.
→When does semantic metadata make visual datasets actually transferable between sites — and when doesn't it?
→How early semantic design lets regulated operational data scale into secondary AI and analytics use safely.
→Bridging mechanical, electrical, software, and systems-engineering ontologies in complex product development.
→Maintaining semantic stability across decades, supplier networks, and security constraints in defence systems.
→Semantic Data Readiness as foundation for DPP compliance and meaningful sustainability data across supplier networks.
→What semantic structure design configurators need to scale across product families without breaking.
→How large organisations manage semantic governance across thousands of contributors, products, and regions.
→Making continuous manufacturing data usable by AI optimisation across multiple plants without per-site rework.
→Semantic Data Readiness as enabler of meaningful sustainability reporting and compliance across complex value chains.
→External organisations contribute cases through the Forum. Get in touch: info@semanticdatareadiness.org
An open arena for dialogue, exchange, and feedback on the project's emerging methods. No agreement, no cost — just active participation. We expect to engage 30–50 external organisations during Etapp 1.