Visual Assessment Scope Disclosure Score (VASDS)

GMI-WT-038Water TreatmentPilot-Ready

Definition

VASDS measures the percentage of an asset's total condition dimensions that an AI visual assessment system can actually evaluate.

The Measurement Gap

The Measurement Gap

What Legacy Metrics See:
AI inspection platforms report defect detection rates, classification accuracy, processing speed, and cost per segment. Marketing materials list the defect types the system can identify. Traditional inspection programs track coverage (percentage of system inspected) and condition distribution (percentage of pipes in each PACP overall grade).

What Legacy Metrics Miss:

No standard metric quantifies what the AI assessment actually covers relative to the full condition picture. An AI platform reporting 94% accuracy for structural defects is telling you it's very good at the things it can see, but not telling you how much of the overall condition assessment it can see. A utility relying solely on AI inspection may have excellent data on surface-visible defects while remaining blind to wall thickness degradation, soil-side corrosion, external loading conditions, and other factors that determine remaining useful life. The danger isn't that the AI is wrong about what it assesses: it's that the utility confuses "assessed by AI" with "fully assessed."

The Decision This KPI Enables:
VASDS enables the utility to design complementary inspection strategies that fill the gaps AI cannot cover. When VASDS shows that AI covers 55% of condition dimensions for metallic pipe, the utility knows that 45% of the condition picture requires other assessment methods: acoustic inspection, soil resistivity testing, coupon sampling, or physical excavation. VASDS transforms the AI inspection conversation from "is the AI good enough?" to "what does the AI cover and what else do we need?", a more productive and honest question.

Relevant Research:

NASSCO's PACP system defines condition dimensions across structural, operational, and maintenance categories. Published research on AI-assisted pipe inspection consistently shows that visual AI excels at surface-visible defects but cannot assess conditions requiring non-visual data (wall thickness, external loading, soil conditions, cathodic protection effectiveness). Multi-sensor inspection research demonstrates that comprehensive condition assessment requires combining visual, acoustic, electromagnetic, and sometimes destructive testing methods.

Formula

Plain language: For each asset type the AI inspects, enumerate the total engineering inspection dimensions required by the relevant standard. Then identify which of those dimensions the AI visual method can actually assess. The ratio is the scope disclosure score.

Notation:
VASDS = AI_assessable_dimensions / Total_required_dimensions

Where:

  • AI_assessable_dimensions = number of inspection dimensions the AI visual method can assess for this asset type (e.g., surface corrosion, coating condition, structural deformation)
  • Total_required_dimensions = total inspection dimensions required by the engineering standard for this asset type (e.g., also includes wall thickness, seam integrity, foundation condition, cathodic protection effectiveness)

Units: Percentage per asset type. The number is the disclosure. A steel water tank: AI visual can assess coating condition, surface corrosion, structural deformation (3 of 7 dimensions = 43%). The other 4 require physical inspection. The goal is not 100% but transparency about what is and is not covered.

Proposed formula, pending validation pilots.

At a Glance

Category
Inspection/Classification Fidelity
← Emerging Contaminant Liability Accumulation Rate (ECLAR)Baseline Integrity Confidence (BIC) →

Assess Your Organization

The AI-GPS Assessment evaluates your AI governance maturity and recommends KPIs like this one for your implementation roadmap.

Take the Free AI-GPS Assessment