CoatingPassport · schema v1.1Demo · Pipeline

NCS-PIPELINE-DEMO-T17 — Coating Passport

Same canonical CoatingPassport data product as every other Hullproof vertical, applied to a pipeline. Domain knowledge (applicable standards, severity calibration) is automatically applied via the domain registry — adding a new asset class is a directory, not a rewrite.

View as JSON →Download PDF →Cognite CDF export →OSDU export →RDF / Turtle →DNV Veracity export →GeoJSON / ArcGIS →OPC-UA NodeSet →Condition risk (JSON) →

Condition Risk Index

indicative · fully traceable

62.9high/ 100 · evidence conf 88 %

Axes

Surface62.9

Structure—

Performance—

Indicative commercial signal

Insurance premium multiplier

×1.39 – ×1.96

Annual fuel penalty (ISO 19030)

— (no performance value)

Top drivers — every point traces to a finding or metric

severe corrosion+5.41
moderate pitting+2.114
moderate coating_breakdown+1.847
minor weld_defect+0.533

Every point of `index` traces to a finding_id or SHM metric in `drivers`; recompute the same passport to reproduce. Indicative only — Hullproof sets no premium and takes no risk; this is the signal a market prices against.

Data provenance

how each number was obtained

4 findingsVLM hullproof-vlm-v2.1.3, mean confidence 84 %modeled
Inspection duration18000 sdeclared
Frames analysed540,000measured

Measured = read from a sensor / instrument / file header. Estimated = derived from a proxy. Modeled = AI / algorithm output. Declared = operator-supplied. Nothing is presented as stronger than its provenance.

Asset

Asset ID: NCS-PIPELINE-DEMO-T17
Type: Pipeline
Material: coated_steel
Operator: Demo Midstream AS

Inspection

Timestamp: Fri, 03 Apr 2026 05:20:00 GMT
Footage source: crawler
Duration: 300 min
Frames analyzed: 540,000
Operator tier: enterprise_user
Model version: hullproof-vlm-v2.1.3

Findings summary (4)

critical

severe

moderate

minor

Findings detail

corrosion

confidence 92%1287 framesimg-quality 84%severe

General corrosion with 58% wall-thickness loss over a 1.2 m × 0.4 m area at top-of-line. Consistent with TLC (top-of-line corrosion) signature from sour-gas service.

Location: KP 8+420, 12 o'clock orientation (8420, 0, 0)

Recommended action: DNV-RP-F101 / B31G fitness-for-service assessment; consider pressure derating until repair scheduled; plan composite-wrap reinforcement at next inhibitor campaign.

pitting

confidence 87%412 framesimg-quality 81%moderate

Pitting cluster at low point, max depth 3.2 mm. Consistent with water-accumulation at sag bend.

Location: KP 11+200, 6 o'clock (bottom) (11200, 0, 0)

Recommended action: UT thickness confirmation at next dive; monitor sag-bend pigging strategy.

coating breakdown

confidence 81%198 framesimg-quality 76%moderate

FBE (fusion-bonded epoxy) coating disbondment over a 0.5 m circumferential band. Consistent with prior pipeline movement; cathodic-protection compensation adequate per UT.

Location: KP 4+150, full circumference, 0.5 m length (4150, 0, 0)

Recommended action: Re-inspection within 24 months; consider in-situ patch coating if disbondment grows >0.8 m circumferentially.

weld defect

confidence 74%89 framesimg-quality 72%minor

Single inclusion indication at girth-weld root, ~3 mm. Within ISO 19232 / API 1104 acceptance for in-service inspection.

Location: KP 6+885 girth weld (6885, 0, 0)

Recommended action: No action this cycle; document for trending.

Compliance

EU AI Act class: high_risk_ai_system
Review workflow: Pending

Applicable standards

• DNV-ST-F101 (Submarine pipeline systems)
• DNV-RP-F101 (Corroded pipelines)
• API 1160 (Managing system integrity for hazardous liquid pipelines)
• API 1163 (In-line inspection systems qualification)
• ASME B31.4 (Liquid pipeline transportation)
• ASME B31.8 (Gas transmission and distribution)
• PSA Norway Facilities Regulations §61

Interoperability

Cognite CDF compatible: Yes
OSDU mapping: osdu:reference-data--AssetType:Pipeline:1.0.0
Veracity export: Not exported
Knowledge graph: Not generated

The same JSON powers Cognite CDF Asset+Timeseries, OSDU exports, RDF knowledge graphs, and MCP-tool calls from agentic systems. PDF is a renderer of this; never the source.