Definition
The share of AI-generated failure analysis hypotheses that were confirmed by the work actually performed on the asset.
The Measurement Gap
There is no metric that tells you what proportion of AI failure-analysis hypotheses were confirmed by the work actually performed, which means the diagnostic accuracy of the recommendation engine is never scored.
Formula
FHCR = H_conf / (H_conf + H_disconf)
Proposed formula, pending validation pilots.
At a Glance
Decision Supported
Whether to accept an insight's root cause reading as the basis for a work order, and whether the recommendation engine's diagnostic reasoning is reliable for this asset class.