Digital Twin Fidelity Index (DTFI)

GMI-WT-007Water TreatmentPilot-Ready

Definition

Measures the aggregate agreement between a continuously calibrated digital twin's predictions and observed sensor values across the network, reported as a score from 0 (no predictive value) to 1 (perfect match).

The Measurement Gap

Hydraulic digital twins drift from reality as field conditions change: partially closed valves discovered during emergencies, deposit buildup reducing C-factors, undocumented connections from construction, pressure reducing valves drifting from their set points. No metric tracks the aggregate gap between model predictions and observed system behavior across the network. System pressure at monitoring points is a spot check; it does not validate the model's pressure predictions at the vast majority of nodes without sensors. When decisions rely on twin simulations (emergency isolation, flushing optimization, fire flow analysis, pressure management) and the twin has drifted, those decisions are based on fiction. The gap between fiction and reality is invisible until an emergency exposes it.

Decision Supported:

Can you trust the digital twin's scenario simulations for emergency response, flushing optimization, fire flow analysis, or pressure management, and where in the network has the twin drifted most?

Formula

Plain language: At every point where you have both a twin prediction and a sensor observation (pressure monitoring points, flow meters, tank levels), calculate the percentage error. Average across all comparison points. Subtract from 1 so that 1 means the twin matches reality and 0 means it is useless.

Notation:

DTFI = 1 − ( 1/S ) × Σ |Twinₛ − Observedₛ| / |Observedₛ|

Variables:

Variable Definition Unit
Twinₛ Digital twin's predicted value at sensor point s PSI (pressure), GPM (flow), or ft (level)
Observedₛ Actual sensor reading at point s Same units as Twinₛ
S Total number of comparison points with both twin prediction and sensor observation Count
s Index across comparison points n/a

Proposed formula, pending validation pilots.

At a Glance

Unit
Score (0 to 1)
Trend Direction
Toward 1.0
Reporting Frequency
Weekly with zone-level detail
Category
AI Decision Quality
Inspection Classification Equity Index (ICEI) →

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