Fleet insights

Main-line delay hours avoided

1,940

rolling 90 days · 22,000 route miles monitored

Detector events triaged

48,217

+12% vs prior 90 days

Handled without a call

96.4%

+1.8 pts vs prior 90 days

Median time to dispatch

3m 41s

−7m 12s vs prior 90 days

First-call resolution

89%

+4 pts vs prior 90 days

Detector events

Last 30 days

Share needing a call

Same window — held flat while volume rose

Events by detector type

Rolling 90 days

Severity mix

What actually needed a human

Minor86.4%
auto-logged, no call
Serious12.3%
train master dispatched
Critical1.3%
escalated for approval

When defects surface

Detector events by weekday and hour

0003060912151821

Mon

Tue

Wed

Thu

Fri

Sat

Sun

fewermore

Busiest corridors

Ranked by detector volume

Corridor	Division	Events	Calls	Median dispatch	Delay avoided
Shenandoah Line	Virginia	4,820	214	3m 12s	312h
Fostoria District	Lake	4,390	188	3m 48s	268h
Pittsburgh District	Pittsburgh	3,910	176	4m 02s	241h
Harrisburg District	Harrisburg	3,640	151	3m 30s	205h
Piedmont District	Piedmont	3,180	143	3m 55s	188h
Georgia District	Georgia	2,870	129	4m 18s	164h

How criticals resolved

Rolling 90 days

Context layer

Three patterns the fleet has learned

▍

48,217 events · 19 months of history

Precursor

Established

Acoustic roughness predicts a hot bearing

1,100 mimedian warning before the critical

A roughness index above 2.4 preceded 71% of hot-bearing criticals.

Learned from 4,812 bearing failures since 2024

93%

Correlation

Established

Cold snaps drive bearing failures

2.7×criticals within 48h of a 20 degree drop

Holds across three winters on the Shenandoah and Pocahontas lines.

Learned from 3 winters · 11 subdivisions

87%

Linkage

Emerging

Some repairs don't hold

3.1×re-flag rate within 90 days

One shop's released cars return to a detector far sooner than fleet average.

Learned from 9,140 work orders joined to later detector hits

74%

Illustrative figures. This view is not connected to the live demo data.