Technology · Cognitive sensors
A sensor that flags its own signal problems.
A model is only as good as its input. That is why every node also watches its own sensors.
Signal health
Level, RMS and clipping are computed for every segment. A dropout or a clipping input is flagged in the record.
Dropouts
Gaps in the recording are noted in the record. No result quietly rests on incomplete data.
An example
The input of a vibration sensor clips. The Vibration Node counts the clipped samples in every segment and stores the count with the recording, so the clipped segment is flagged instead of passing silently into analysis.
Next step
Questions about measurement quality?
We show you how signal health and dropouts appear on the dashboard.