CBM Interval Outcome Validation Rate (CIOVR)

GMI-RE-048Renewable EnergyPilot-ReadyPattern O-1: Confidence Calibration

Definition

Proportion of CBM-extended intervals where the asset completed the extended interval within acceptable condition parameters, without unplanned failure, emergency maintenance, or threshold exceedance. Closes the loop on CBM: not whether the model was confident, but whether the extension actually worked.

The Measurement Gap

The Measurement Gap

CBM programs are typically evaluated on input metrics: how many assets have condition monitoring, how many interval extensions were recommended, how much maintenance cost was avoided through extended intervals. These input metrics demonstrate CBM program activity but do not validate CBM program effectiveness. An extension recommended by the model and accepted by the planner counts as a success in input metrics regardless of whether the asset survived the extended interval or failed catastrophically three weeks before the rescheduled maintenance.

Without outcome validation, CBM programs operate on an unfounded assumption that all recommended extensions are safe. The cost savings from extended intervals are documented and reported. The cost of failures during extended intervals is absorbed into general failure statistics without being attributed to the CBM decision that created the exposure. An organization might report $500K in CBM savings from extended intervals while experiencing $300K in failures during those same intervals, netting only $200K but believing the savings are $500K because nobody connects the failure costs back to the extension decisions.

The Decision It Enables

CIOVR provides the outcome validation that transforms CBM program evaluation from activity measurement to effectiveness measurement. When CIOVR is high (above 95%), the CBM model's interval extensions are safe and the maintenance cost savings are real. When CIOVR is moderate (90 to 95%), some extension decisions are producing failures, and the model's extension aggressiveness should be reviewed for specific asset types and failure modes. When CIOVR is low (below 90%), the CBM model is extending too aggressively and maximum extension limits should be reduced.

The segmentation capability is critical: CIOVR computed at the fleet level may show 96%, but segmented by asset type and PM task it may reveal that gearbox oil change extensions have 99% CIOVR while blade inspection extensions have 82% CIOVR. The fleet number masks a problem that is concentrated in specific PM types where the CBM model's confidence is not calibrated to actual outcomes.

Formula

Primary Formula:

CIOVR = Successful_Extensions / Total_Extensions x 100

Variable Definitions:

Variable Definition
Successful_Extensions the number of extended-interval PMs where the asset reached the next scheduled maintenance without any of the following: an unplanned failure of the failure mode the PM addresses, emergency or urge...
Total_Extensions all PM intervals extended by CBM recommendation during the measurement period that have completed the extended interval

Calculation Example:

Over a 12-month measurement period, an organization's CBM system recommends 180 interval extensions across its wind fleet. Of these, 145 have completed the extended interval and reached the next scheduled maintenance. Of the 145 completed extensions: 137 assets reached the next maintenance without incident (successful), 5 experienced unplanned failures of the target failure mode during the extended interval, and 3 arrived at the next maintenance with conditions exceeding acceptable thresholds. CIOVR = 137 / 145 x 100 = 94.5%.

Edge Cases and Measurement Boundaries:

Assets that received other maintenance during the extended period for a different failure mode should be counted as successful for the CBM-extended mode. Extensions where the technician arrived and found the asset in worse condition than expected should be counted as failures. Trivial extensions of less than 25% increase over the original interval should be excluded because they represent minimal additional exposure.

Proposed formula, pending validation pilots.

At a Glance

Unit
Percentage
Trend Direction
Higher; more successful extensions
Reporting Frequency
Quarterly by PM type
Category
Condition-Based Maintenance Integrity
← Multi-Service Warranty Exposure Index (MSWEI)CBM Sensor Health Correlation Index (CSHCI) →

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