Inspection Scope Coverage Ratio (ISCR)

GMI-RE-013Renewable EnergyPilot-ReadyPattern O-7: Inspection Coverage

Definition

Proportion of failure modes that regulators and the OEM manual require inspection to detect which the AI inspection method is structurally capable of detecting. Drone and CV inspection can post excellent coverage numbers while being blind to failure modes their sensors cannot physically see.

The Measurement Gap

AI-powered inspection methods offer enormous efficiency advantages: a drone can survey 100 turbine blades in the time a rope access team inspects three. Thermal imaging can scan an entire solar field from the air in hours. The economic case for shifting from physical to AI-based inspection is compelling, and most operators are moving in this direction. The measurement gap is that efficiency gains have outpaced scope validation.

Every regulatory body and OEM maintenance manual specifies a set of failure modes that periodic inspection must be able to detect. Internal cracks in blade root bolting. Subsurface delamination in composite structures. Corrosion on structural fasteners inside nacelles. Contact degradation in high-voltage junction boxes. Each of these failure modes requires specific detection capabilities: ultrasonic for subsurface defects, visual access to internal surfaces, thermal sensitivity thresholds for electrical faults. When AI inspection replaces physical inspection, some of these detection capabilities are maintained, some are enhanced, and some are permanently lost.

ISCR measures what percentage of the required inspection scope the AI method can structurally cover. This is not a question of detection reliability (that is what DSCS measures) but of detection capability. A drone with a visual camera can see surface cracks but cannot detect subsurface delamination. A thermal camera can identify hot junction boxes but cannot assess bolt torque. These are not calibration problems to be solved with better algorithms; they are physics limitations inherent to the sensor type.

The decision ISCR supports is whether the organization's current AI inspection program can detect all required failure modes, or whether a residual physical inspection program is needed to cover the specific gaps that AI methods cannot address.

Formula

List all failure modes that the regulatory body (OSHA, NERC, or applicable standard) and the OEM maintenance manual require inspection to detect. For each failure mode, determine whether the AI inspection method is structurally capable of detecting it based on sensor type, resolution, and access angle. Divide the detectable count by the total required count.

Formula (Notation):

ISCR = Detectable_Failure_Modes_AI / Total_Required_Failure_Modes

Variable Definitions:

Variable Definition
Detectable_Failure_Modes_AI count of failure modes from the regulatory and OEM requirement set that the AI inspection method is structurally capable of detecting
Total_Required_Failure_Modes count of all failure modes that regulatory and OEM standards require periodic inspection to detect

Edge Cases and Measurement Boundaries:

Detection capability is binary: either the sensor and method can detect the failure mode or it cannot. Do not conflate detection capability with detection reliability, which is what DSCS measures. Different AI inspection methods cover different failure modes, so compute ISCR per method and per combined-method scenario to understand how method combinations close gaps. When OEM requirements exceed regulatory requirements, use the union of both to define the total required set.

Proposed formula, pending validation pilots.

At a Glance

Unit
Ratio
Trend Direction
Higher; more coverage
Reporting Frequency
Annually per inspection method
Category
AI Inspection and Defect Management Quality
← Defect Severity Calibration Score (DSCS)Alarm Suppression Signal Loss Rate (ASSLR) →

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