Definition
Degree to which the fleet reliability data supporting a proposed maintenance interval extension is representative of the aircraft and operating conditions to which the extension will be applied: measuring whether the evidence actually supports the decision before the reliability board approves it.
The Measurement Gap
The measurement gap IEEVS addresses: reliability board submissions for interval extensions typically present statistical evidence (removal rates, in-service findings, reliability trends) without characterizing the representativeness of the fleet that generated that evidence. A proposal may show zero failures under the current interval for 5 million flight hours, but if those hours came from 60% temperate-climate aircraft and the fleet is 40% coastal-environment aircraft, the evidence is not fully representative. The reliability board has no standard metric for evaluating whether the evidence covers the full population to which the extension will apply.
This is a formalized version of survivorship bias: the evidence fleet survived under the current interval by definition. That's why data exists for them. Aircraft that experienced failures may have been removed from the population before generating the data that supports the extension. New-to-fleet aircraft with limited hours haven't been exposed to the conditions that would reveal the risk at the proposed extended interval. The evidence is real; the question is whether it applies.
The decision IEEVS supports: should the reliability board approve this interval extension as submitted, restrict it to the evidence fleet's operating profile, or require additional data collection before fleet-wide application? Without a representativeness metric, the board's only option is qualitative judgment about whether the evidence "seems sufficient." IEEVS provides a quantified basis for that judgment and a defensible documentation trail if the extension is challenged by regulators or auditors.
Formula
Plain language: For a proposed interval extension, compare the characteristics of the aircraft whose data supports the extension (the evidence fleet) to all aircraft to which the extension will be applied (the target fleet). Measure representativeness across four dimensions. Low scores indicate the evidence doesn't cover important segments of the target fleet.
Notation:
IEEVS = 1 − (1/D) × Σ Divergence_d(evidence_fleet, target_fleet)
Variables:
- D = number of representativeness dimensions (4: operating environment, aircraft age distribution, utilization rate, modification status)
- Divergence_d = statistical divergence between evidence fleet and target fleet on dimension d, normalized to 0–1 (use Kolmogorov-Smirnov statistic for continuous variables, chi-squared for categorical, normalized to 0–1 range)
Four representativeness dimensions:
- Operating environment: distribution of flight hours across environment categories (hot-and-high, coastal, cold, temperate)
- Aircraft age: distribution of airframe flight hours and cycles
- Utilization rate: average daily flight hours
- Modification status: percentage of aircraft with relevant STCs, modifications, or configuration differences affecting the component in question
Units: 0 to 1. 1.0 = evidence fleet perfectly represents the target fleet. Lower = less representative. Present IEEVS alongside individual dimension divergence scores so the reliability board can see which dimensions are underrepresented.
New-to-fleet handling: aircraft with insufficient hours to contribute to the evidence fleet represent a gap by construction. IEEVS will be appropriately low if they constitute a significant fraction of the target fleet. This is the correct signal: the evidence doesn't cover them.
Proposed formula, pending validation pilots.