Read this case from your perspective
Different roles read the same programme differently. Pick the lens that matches your work, the same case study foregrounds what's most relevant to you.
Programme Director
Programme risk, regulatory accountability, audit defensibility. How does Weaver let you sign off the asset's configuration on demand, decades after a decision was made?
Asset Owner
Lifecycle TCO, vendor independence, multi-decade continuity. How does Weaver keep your asset's information valid and portable for the full operational life, independent of any single supplier's survival?
Information Architect
Federation, metamodel control, integration with existing tools. How does Weaver sit on top of Revit, Relatics, SharePoint and your document management system without forcing migration?
A new-build research reactor with a 70-year operational requirement.,
A nuclear research reactor programme procured Weaver as the Common Data Environment for the design, construction and operations phases.
The programme has a 10-year build-out and a 70-year operational design life. The regulator can demand evidence about the asset's configuration at any moment during that span, and the engineering teams who eventually answer those questions will not be the teams who specified the equipment.
Names and figures have been anonymised. The programme structure, regulatory framing and engagement shape reflect a representative Weaver implementation in the nuclear sector.
Programme details
Sector
Region
Build duration
Operational design life
Use cases
Standards applied
Integration footprint
Deployment
Users
Nuclear, research reactor programme
Europe
Approx. 10 years from contract to operational
70 years
Requirements traceability, IFC ingestion, ISO 19650 information container management, regulatory audit, lifecycle handover
ISO 19650, ISO 15926-11, IFC (ISO 16739), BCF, openCDE, RDF, SPARQL
Revit, AVEVA, Smart Plant, Plant 3D, Relatics, SharePoint-based DMS
On-premise, air-gapped from public internet
Design engineers, requirements engineers, safety case authors, regulator-facing programme office, operations & maintenance teams (post-handover)
KEY OUTCOMES
- One queryable source of asset configuration across all consortium tools.
- Inspection preparation reduced from weeks to days.
- Git-style versioning of all asset data: branches, diffs, merges, time-travel, with audit by default.
- No information-loss event at the design-to-operations handover.
Programme details
Sector
Region
Build duration
Operational design life
Use cases
Standards applied
Integration footprint
Deployment
Users
Nuclear, research reactor programme
Europe
Approx. 10 years from contract to operational
70 years
Requirements traceability, IFC ingestion, ISO 19650 information container management, regulatory audit, lifecycle handover
ISO 19650, ISO 15926-11, IFC (ISO 16739), BCF, openCDE, RDF, SPARQL
Revit, AVEVA, Smart Plant, Plant 3D, Relatics, SharePoint-based DMS
On-premise, air-gapped from public internet
Design engineers, requirements engineers, safety case authors, regulator-facing programme office, operations & maintenance teams (post-handover)
KEY OUTCOMES
- One queryable source of asset configuration across all consortium tools.
- Inspection preparation reduced from weeks to days.
- Git-style versioning of all asset data: branches, diffs, merges, time-travel, with audit by default.
- No information-loss event at the design-to-operations handover.
A multi-decade reactor programme with no margin for information loss
The programme procures a new-build research reactor with associated handling facilities, conventional and reactor-grade auxiliary systems, and a regulator-facing safety case spanning the full operational lifetime.
Construction involves a consortium of engineering, procurement and construction contractors across multiple jurisdictions. Each contractor brings its own native tooling (Revit, AVEVA, Smart Plant, Relatics, Bentley) and its own preferred documentation conventions. Information convergence is itself a sub-project.
The situation before Weaver
Information lived in three disconnected categories, and Excel was the integration tool.
- Requirements, V&V plans and asset registers were spread across Relatics and dozens of project Excel workbooks.
- Geometric information lived in Revit and AVEVA, exported to PDF or IFC for review but never unified.
- Documents lived in a SharePoint-based DMS with manual revision control.
- Issues raised in design reviews lived in BCF in some weeks, email threads in others.
- Inspection prep for the regulator routinely took 4–6 weeks of cross-tool reconciliation.
The solution after Weaver
A single Common Data Environment, ISO 19650-disciplined, ISO 15926-11-modelled, federated across the consortium's existing tools.
- All requirements, V&V plans, asset structures and design changes live as nodes in one RDF knowledge graph.
- IFC models from every authoring tool are ingested at the object level, model elements link to the requirements they satisfy and the changes that affected them.
- Documents remain in the existing DMS but appear as first-class information containers in the graph, with full ISO 19650 lifecycle states.
- Every change is preserved with who, when and what, queryable across decades.
- Inspection prep is a SPARQL query, not a reconciliation exercise.
The architectural decision was that nothing should ever require migration, not at handover, not at major release upgrades, not when the consortium contractors rotate. Open standards (RDF, ISO 19650, ISO 15926-11) were not a procurement preference; they were a technical requirement of the 70-year operational mandate.
What the implementation actually looked like
We engaged on a pilot-first, evidence-led basis. Twelve weeks from contract to operational pilot, then a phased rollout across the consortium. Throughout, Weaver was positioned as the semantic backbone above existing tools, never as a replacement for the contractors' native authoring environments.
Standard model deployment with nuclear extensions
We deployed the ISO 15926-11 reference data library as the metamodel baseline, then extended it with nuclear-specific classes (reactor vessels, primary loop systems, radiological zones, safety functions). Extensions are RDF, the customer's Certified Weavers can evolve the metamodel without engineering involvement from us.
Pre-built modules for every consortium source
Revit, AVEVA, Smart Plant, Plant 3D, Relatics, the SharePoint DMS — each was connected by configuring Weaver's pre-built data modules, not by writing integration code. IFC exports are ingested at the object level (graph nodes with properties and relationships, not opaque files). The Relatics connector synchronises requirements bidirectionally. Less common feeds (a vendor inspection CSV, a legacy planning export) are handled by configurable pipelines. Customer-side Certified Weavers own the configuration after handover.
Requirements traceability via SPARQL
Requirements and V&V activities (previously spread across Relatics and Excel) live as nodes in the graph, linked to the assets they govern. Cross-cutting questions become queries:
- Show every safety requirement, its verification status, and the model elements implementing it.
- Show every requirement that has changed since the last regulatory inspection.
- Show the lineage of this design decision back to the originating requirement.
ISO 19650 information container lifecycle
All containers (drawings, models, requirement sets, inspection reports) move through Work in Progress → Shared → Published → Archived under a defined approval workflow. Status is queryable. There is no parallel email-approval channel to bypass.
Git-style versioning across the asset's life
Every change is a commit; the history is a directed graph, not a flat log. Engineers can branch the data to explore design alternatives (a candidate cooling configuration, a proposed safety-case revision), then merge or discard with conflict resolution. Time-travel queries ("the configuration as of date X") and diffs ("what changed between submissions 7 and 8") are first-class operations. The audit trail comes for free, queryable as part of the same graph. At construction completion, there is no handover event: operations teams already use the same Weaver instance the design teams used.
The implementation worked because Weaver behaved as a semantic backbone, it federated the consortium's existing tools rather than asking them to migrate. Designers kept working in Revit and AVEVA. Requirements engineers kept working in Relatics. Document controllers kept using the existing DMS. What changed was that every output from those tools now flows into a single ISO 15926-11-modelled, ISO 19650-disciplined knowledge graph that the entire programme can query.
Predictable inspections, audit by default, no migration debt
Twelve months after go-live, the programme measured the impact across reporting effort, audit readiness, and information continuity.
Inspection prep
–80%
Time to prepare regulatory inspection packs reduced from 4–6 weeks to under one week. Reports are SPARQL queries against the graph, not cross-tool reconciliation.
Change traceability
100%
Every design change is linked to its originating requirement, its affected model elements, and its approval record. Required for regulatory defensibility; previously aspirational.
Version history
70+ yrs
Every change a commit, attributed and immutable. Branchable, diffable, mergeable, time-travellable across the full operational life. Audit trail comes for free, queryable as part of the same graph.
Qualitative outcomes
- Cross-disciplinary clarity, engineering, safety, procurement and operations work from the same configuration data, in real time.
- Confidence in regulator-facing claims, assertions about the asset's state are backed by queryable evidence, not by file collation.
- No information loss at handover, the design-to-operations transition is a continuation, not a migration.
- Vendor independence, the data structure is ISO 15926-11; the wire format is RDF. Replacing Weaver, if ever required, is a managed migration to another open-standards platform, not a reverse-engineering exercise.
We chose Weaver because the architecture answered the only question that really mattered to us: in 2090, will an inspector be able to read this data? On RDF, on ISO standards (yes. On a proprietary CDE) that's the kind of bet a 70-year programme cannot afford to take.
Chief Data Officer (anonymised composite)
How to start a programme like this
Whether the asset is a reactor, a wind farm or a major piece of infrastructure, the engagement shape is the same: pilot first, evidence before scale, customer-side ownership of the configuration by the end of the first year.
Book an introduction
A short technical conversation about your existing tooling landscape, your standards posture, and the operational life of your asset.