Twin-Physical Divergence Index (TPDI)

GMI-RAIL-012Rail TransportationInfrastructure-Dependent

Definition

The magnitude and direction of divergence between a digital twin's predicted asset state and the physical asset's actual measured state at each validation point.

The Measurement Gap

A bridge digital twin's stress model assumes material properties that have degraded in reality. The twin estimates 15 years of remaining life; the physical structure has 8. Maintenance decisions based on the twin systematically underinvest in the asset. Bridge condition ratings from inspection are periodic (typically biennial). Between inspections, the twin's estimate is the decision basis and no metric tracks twin-versus-physical divergence.

Decision Supported:

How much should we trust this digital twin's assessment? Is it drifting optimistic (underestimating degradation, which is dangerous) or pessimistic (overestimating degradation, which is wasteful)?

Formula

Plain language: At each validation point (physical inspection, load test, cross-sensor verification), compare the twin's predicted state parameter to the measured actual. Express as a percentage divergence with sign indicating direction.

Notation:

TPDI_abs = SUM(|P_twin_i - P_actual_i| / P_actual_i) / N_validations
TPDI_signed = SUM((P_twin_i - P_actual_i) / P_actual_i) / N_validations

Variables:

Variable Description Unit
P_twin_i Twin's predicted state parameter at validation point i Parameter-specific
P_actual_i Physically measured state parameter at validation point i Same units
N_validations Number of validation events with both twin and physical measurements Count

Sign Convention: TPDI_signed positive = twin is optimistic (dangerous). Negative = twin is pessimistic (wasteful). Track both absolute and signed.

Proposed formula, pending validation pilots.

At a Glance

Unit
Percentage
Category
Digital Twin and Recommendation Fidelity
← Recommendation Executability Score (RES)Consist-Adjusted Fuel Target Variance (CAFTV) →

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