Historian Trend Looks Smooth—Did It Hide the Spike? Check the Query Mode
A smooth trend can reflect interpolation or aggregation rather than every recorded measurement. Before concluding a spike never happened, inspect both data acquisition and the query used to draw the chart. A historical display is a processed view with settings that matter.
Original data example; historian settings and query semantics depend on the version and platform.
What is happening?
Historian sampling, deadband, storage, aggregation, and interpolation affect different stages. A value never sampled cannot be recovered by querying raw records. Conversely, a stored short excursion can disappear from a chart that displays averages over wide intervals. Choose the evidence view to match the diagnostic question.
Check these five things
-
Record tag acquisition and historical sampling settings, including quality and deadband behavior.
-
Inspect the chart's time range, result size, aggregation, and interpolation options.
-
Compare a supported raw-data query with the displayed trend around the event.
-
Use minimum and maximum or another suitable aggregation when the question concerns excursions, according to platform capabilities.
-
Validate the full pipeline with a known simulated pulse and preserve the results.
Worked example
Ten recorded samples contain nine zeros and one value of 100. Their average is 10, while the maximum is 100. A chart displaying only the average communicates a different fact from the raw samples. If the pulse was never sampled, neither query can reconstruct it reliably.
Quick diagnostic reference
| Stage | Potential loss or transformation |
|---|---|
| Acquisition | Event missed before storage |
| Storage policy | Sample omitted or compressed |
| Query/display | Aggregation or interpolation |
The mistake to avoid
Increasing chart resolution cannot recover data that the historian never collected. Diagnose the acquisition and display stages separately.
How to verify the fix
Show how the known test event appears in stored data and each required display, with quality and timing interpreted correctly.
Field-work boundary: These are educational diagnostic guides. Use the exact equipment manuals and approved site procedures. Electrical testing and machinery changes belong to qualified, authorized personnel; control hazardous energy and validate affected functions before release.
Technical reference
Inductive Automation: tag history configuration, deadbands, and query behavior. The examples and diagnostic tables above are original teaching material; they do not replace the product manual.