Fault Isolation Accuracy-Speed Trade-off Index (FIASTI)

GMI-AVI-038AviationIntegration-Required

Definition

The Pearson correlation between the length of the AI-generated fault isolation path (number of diagnostic steps taken before a conclusion is reached) and the subsequent repeat defect rate for those events. A strong negative correlation indicates that shorter AI fault isolation paths produce more repeat defects: evidence that the algorithm is trading diagnostic thoroughness for speed.

The Measurement Gap

Traditional fault isolation metrics track mean time to isolate, first-time fix rate, and repeat defect rates as independent measurements. None of them reveal whether the AI's diagnostic process is introducing a trade-off between speed and accuracy. An organization might see a 10% improvement in mean time to isolate and a 5% increase in repeat defect rates in the same quarter and interpret each as separate issues. FIASTI connects them. It tests whether the AI is reaching faster conclusions by taking shorter diagnostic paths, and whether those shorter paths are leaving root causes unconfirmed.

This distinction matters because AI fault isolation systems face an inherent pressure toward brevity. Shorter paths feel more efficient. They are faster to present, faster to act on, and easier to explain to maintenance supervisors under turn pressure. But aviation fault isolation exists to ensure the correct root cause is identified before the aircraft is returned to service. If the AI is systematically under-testing its conclusions when time pressure is high, shorter paths become a safety signal, not an efficiency win. FIASTI is the metric that converts this suspicion into a measurable, monitorable relationship.

Formula

Plain language: Calculate the Pearson correlation between path length and repeat defect outcome across all AI-assisted fault isolation events in the period. Supplement with a delta rate comparing repeat defect rates for short versus long paths.

Primary notation:
FIASTI_ρ = ρ(Path_length_i, Repeat_defect_i) for all events i in period

Supplemental delta:
FIASTI_Δ = Repeat_rate_short_path − Repeat_rate_long_path

Variables:

Variable Definition Source
Path_length_i Number of diagnostic steps in the AI fault isolation sequence for event i AI diagnostic system log
Repeat_defect_i Binary (1/0) indicator of repeat defect within 30 days for event i CMMS/MRO system
Repeat_rate_short_path Repeat defect rate for events with path length below the period median Calculated
Repeat_rate_long_path Repeat defect rate for events with path length at or above the period median Calculated

Proposed formula, pending validation pilots.

At a Glance

Category
Operational Decision Quality Under Pressure
← Check Duration Pressure Index (CDPI)MEL Deferral Operational Context Score (MDOCS) →

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