The situation
← Back
One turbine's bearing, another fleet's lesson
A high-speed shaft bearing starts failing 90 metres up, in turbine WTG-14. Watch Vermilion catch it 13 days out, with confidence borrowed from a fleet it has never seen, and turn a gearbox replacement into an up-tower bearing swap.
Run the incident
Watch the failure develop, and the save happen
Press play, or drag the timeline. Telemetry, asset status, the reasoning trace, and the spectrum all follow the point in time you choose.
Day 1.0
The same incident, twice
One defect, two operators
Same asset, same defect, same clock. Operator 1 runs a routine preventative maintenance program. Operator 2 runs Vermilion. Press play above, or scrub, and watch the same days land differently.
Inside the prediction
The five-stage reasoning trace
Every Vermilion prediction ships with an auditable trace, not a black-box score. The trace below unlocks when the alert fires in the timeline above.
Scrub the timeline past the alert to open this prediction's trace.
Why the lead time matters
Vermilion vs. a threshold alarm
For the reliability team
Under the hood
The physics, the sensors, and the raw numbers behind this scenario. Open what you want to inspect.
Envelope spectrum▶
See it on your fleet, not a demo fleet
This scenario runs on synthetic telemetry. A private demo runs on the failure modes that keep your team up at night: 30 minutes, your equipment classes, real reasoning traces.