Recovery Equity Variance (REV)

GMI-AVI-003AviationInfrastructure-Dependent

Definition

Variance in passenger recovery outcome quality across defined segments (loyalty tier, fare class, route type, or originating station) measuring whether the AI recovery engine systematically produces better outcomes for some passenger populations than others.

The Measurement Gap

Standard IROPS reporting aggregates recovery outcomes across all passengers, making systematic inequity structurally invisible. Overall completion rate, average delay, and aggregate PJDI all report the central tendency: they mask which segments consistently receive good outcomes and which consistently receive poor ones. Operational research models used in IROPS recovery are trained on historical decisions that often reflect historical prioritization: premium passengers were protected, specific routes were favored, and certain hubs received better treatment. When AI systems learn from this history, they inherit and may amplify these patterns without any deliberate discriminatory design. Regulatory scrutiny of AI fairness is increasing: DOT consumer protection rules and EU AI Act provisions both address algorithmic decision-making that produces disparate consumer outcomes. An airline cannot know whether its AI recovery engine is treating all passengers equitably unless it measures the distribution of recovery quality across segments. A carrier that sees PJDI = 0.18 overall could have PJDI = 0.08 for elite passengers and PJDI = 0.52 for economy passengers: the same recovery engine, systematically advantaging premium segments while the aggregate hides the problem. REV makes that distribution visible before it surfaces in complaint trends or regulatory inquiries.

Decision Supported:

Is the IROPS recovery AI introducing systematic bias in who receives good recovery outcomes and who receives poor ones?

Formula

Plain language: Compute PJDI separately for each defined passenger segment. Loyalty tier, fare class, route type, or originating station. Measure the variance across those segment-level PJDI scores. Higher variance means more unequal recovery quality across passenger groups.

Notation:

REV = σ²(PJDI_segment) across all defined segments

Variables:

Variable Definition Unit
PJDI_segment Passenger Journey Degradation Index computed separately for each defined segment (e.g., loyalty tier × fare class, route type, or originating station) 0–1 score per segment
σ² Statistical variance of PJDI_segment scores across all defined segments in the measurement period Variance units (PJDI²)

Segmentation strategy: begin with loyalty tier (elite vs. standard) and fare class (premium cabin vs. economy) as the highest-priority segmentation dimensions. Add route type (hub-to-hub vs. hub-to-spoke) and originating station as secondary dimensions once the primary computation is established.

Proposed formula, pending validation pilots.

At a Glance

Unit
Statistic
Trend Direction
Lower
Reporting Frequency
Per-event; rolling 30-day summary
Category
Disruption Recovery Intelligence
← Passenger Journey Degradation Index (PJDI)Recovery Resilience Consumption Index (RRCI) →

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