The condition layer your digital twin is missing.

The two layers of a useful digital twin

A digital twin without condition data is just a 3D model of the design. Useful, but not enough. The twin becomes decision-capable when it carries:

1. Geometry layer. CAD, BIM, point cloud, 3D mesh. Echo, Omniverse, Bentley Synchro live here. Captures “what the asset is”.
2. Condition layer. Structured asset condition over time: corrosion extent, coating breakdown, biofouling, crack propagation, weld-defect evolution. Captures “what the asset is doing now”.

Operators have invested heavily in the geometry layer. The condition layer is mostly PDF reports nobody can machine-read. CoatingPassport closes that gap.

Integration patterns we support

Equinor Echo

Echo serves facility data + 3D models on mobile, tablet, PC, HoloLens. Hullproof emits CoatingPassport per inspection campaign; the passport renders into CDF (Echo consumes from Omnia / CDF). When an Echo user pulls up a jacket K-joint, the latest CoatingPassport finding for that location is in the same pane as the 3D geometry.

Cognite Atlas + Cognite Data Fusion

CoatingPassport renders to CDF Asset + Event records via the live CDF exporter. Atlas AI agents consume the same Asset hierarchy for geometry, sensor anomalies (NV-Tesseract), and visual-condition findings. One agent, three modalities, one decision pane.

NVIDIA Omniverse + sim-trained autonomy

Omniverse-grade photoreal simulation is where the next generation of autonomy is trained. CoatingPassport carries the ground-truth condition labels that fine-tune those simulations on real-asset states — biofouling distribution, coating breakdown patterns, structural defect locations. The passport becomes a sim-training data source as well as an operational data source.

Knowledge graph + RDF

Operator-internal knowledge graphs (Aker BP, Equinor, Shell run these) connect asset, document, sensor, and condition data semantically. CoatingPassport renders to RDF / Turtle so the condition layer is graph-queryable alongside the rest.

The agentic + VLA loop

When the digital twin, the agentic stack, and the autonomous-robotics layer all share the same Asset hierarchy + condition data + audit lineage, the integrity- management loop closes:

1. Robot (Spot / ANYmal / ROV / AUV) captures inspection footage.
2. Hullproof emits CoatingPassport with structured findings + lineage.
3. Passport renders to CDF / OSDU / Echo / knowledge graph.
4. Atlas AI / Echo agent reads the finding alongside geometry + sensor data.
5. Agent orchestrates next action — work order, follow-up inspection, autonomous-robot dispatch (Hullproof VLA layer).

Five steps. One canonical data object. Each layer is already shipping for major operators — Hullproof is the visual-condition modality the loop has been missing.