CapabilitiesData quality and workflow observability
RAVEN · Ultron · GRID · GRAYLOCK

Data quality and workflow observability

Identify whether an unusual result came from the machine, the instrument, the data connection or the software workflow. Give the engineer evidence from the first affected point and name the person responsible for recovery.

Product documentation · Updated 15 November 2026

Who uses it
Instrumentation engineers, deployment engineers and application support

#Assess the measurement

Check time since last update, missing intervals, a flatline, impossible values and source quality. In a simulated frozen-transmitter case, Ultron flagged the reading as invalid and directed the user to compare it with a local gauge. Associate a data-quality event with the sensor so it can be investigated independently of machine condition.

#Trace a workflow

For an inspection request, trace the sequence from finding to submitted request, destination queue and technician status. For an AI-assisted action, identify which permitted tool ran, what result it returned and whether the target system acknowledged the action. Keep trace detail appropriate to the site’s data-handling agreement.

#Verify recovery

After a connection returns, distinguish new readings from delayed history. Record which time range is incomplete and check that stale values do not reappear as current. Review recurring data-quality issues separately from confirmed equipment faults.

#Outcome and verification

What the team gets
A measurement-quality finding or a workflow trace with source, time, failure point and recovery owner.
How to verify it
Induce a labelled test gap and a separate destination failure. The operator should see different symptoms and be able to route each to the correct owner.