Device data analytics

What you get

• Fault and anomaly detection running against your existing telemetry
• A written account of which faults are detectable in the data you already have, and which would need more sensing
• Detection running against live data, with alerts

Who this is for

Fleets already sending data that nobody has time to read. The failure is usually visible in the data weeks or months before it happens — nobody is looking, or the alerting is threshold-based and either silent or drowning in false positives.

This needs no new hardware. If the answer turns out to be that the data cannot support the detection you want, we tell you that and what sensing would fix it.

What we need from you

A data export covering at least one period containing a known fault, and a description of what that fault was. Without a labelled example we can find anomalies but cannot tell you which ones matter.