Telematics-Log Concordance Index (TLCI)

GMI-TF-010Fleet & TransportationIntegration-Required

Definition

Percentage of operator-reported log problems (DVIR equivalent) on a DOT-regulated component where the concurrent telematics data supports rather than contradicts the operator's report: measuring whether AI trusting telematics over driver reports creates regulatory exposure.

The Measurement Gap

There is no metric that measures the concordance rate between telematics-generated condition assessments and operator-reported log problems on the same asset during the same period. Fleet operations have two parallel condition assessment pipelines: automated sensor data from telematics systems and operator-reported trip problems (the DVIR equivalent). When these pipelines generate conflicting signals (a driver reports "brakes feel soft" while telematics shows brake stroke within tolerance) AI-driven maintenance prioritization trusts the sensor data over the subjective report and deprioritizes the work order. DVIR Severity Classification Accuracy (DSCA) is partially blind here because it measures NLP classification of the driver's report text, not what happens when telematics data contradicts that report. Telematics-based health scores are blind because they are the source creating the override. DVIR Defect Resolution Time is blind because the resolution may be recorded as "telematics assessment confirms no defect" rather than "mechanic examined and certified." The regulatory consequence is severe: 49 CFR 396.11 requires drivers to report vehicle condition, and 49 CFR 396.13 requires mechanics to certify repairs before return to service. If AI deprioritizes a log-generated work order because telematics contradicts the driver, the carrier has a documentary record showing the driver reported a defect, no mechanic examined it, and the vehicle returned to service.

Decision Supported:

When telematics data contradicts a driver's defect report, is the AI creating a regulatory exposure by trusting sensors over the legally required driver inspection process?

Formula

Plain language: When a driver reports a problem in the log application AND telematics data exists for the same component during the same period, how often do the two sources agree?

Notation:

TLCI = (ConcordantReports / TotalOverlappingReports) × 100

Critical sub-metric: Discordant Exposure Rate:

DiscordantExposureRate = (DriverReportWithHealthyTelematics / TotalOverlappingReports) × 100

Variables:

Variable Definition Unit
ConcordantReports Log problem reports where telematics data for the same asset and component during the same period also indicates an anomaly (out of tolerance, alert, or trending toward failure) Count
TotalOverlappingReports All log problem reports where concurrent telematics data exists for the same asset Count
DriverReportWithHealthyTelematics Log problem reports where the driver reported a defect but telematics showed the component as healthy: the regulatory exposure cases Count

The Discordant Exposure Rate is the critical regulatory metric. A TLCI of 80% means 20% discordance, but not all discordance creates regulatory risk. Only cases where the driver reports a problem and telematics says "healthy" create the exposure pattern under 49 CFR 396.13. Target for DiscordantExposureRate: below 5%. Above 10% means AI should not use telematics to deprioritize driver-reported problems.

Proposed formula, pending validation pilots.

At a Glance

Unit
Percentage
Trend Direction
Higher
Reporting Frequency
Monthly
Category
Condition Assessment Discordance
← Risk Score Currency Index (RSCUI)Motor Pool Maintenance Funding Ratio (MPMFR) →

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