Troubleshooting

Calibration drift, baseline and beyond.

A scannable checklist for the first thirty days and the first month of real alerts.

The first six days — ignore the readings

Bearings are still bedding in after install. Drift noise here is normal. The pod is learning the geometry of YOUR mount, YOUR bolt pattern, YOUR asset. Don’t chase numbers in this window.

Days 7 – 25 — the baseline trains

The model learns what “normal” looks like at your duty cycle, your load, your shift pattern. Drift checks are quiet — the pod detects anomalies but does not alert yet. Your engineer can review the chart any time.

Day 26 onwards — the alert box turns on

Anything that breaks the trained envelope by ≥6σ pages the on-call contact. Below 6σ, the dashboard shows soft warning bands but does not wake anyone up.

If a pod shows continuous drift above 6σ after baseline

  • Confirm the mounting surface is clean and the bolts are torqued to spec.
  • Confirm the ambient-temperature log is not spiking — > 40 °C or < 0 °C is worth a call.
  • Check the cellular RSSI on the pod card in the dashboard. A weak signal forces retransmits and skews the noise floor.
  • Re-baseline as a last resort: from the dashboard’s “Pod actions” menu, an engineer can request a 14-day re-baseline. Use this when the asset itself has changed, not as a routine.

When to call us

If the same pod drifts high across two weekly reviews, write to the address below with the pod ID and the chart screenshot. We respond inside one business day with a remote diagnosis — and dispatch a swap pod if the unit itself is the problem.

Still stuck?

Want a one-line answer specific to your plant?

Send three fields. We come back inside one business day with the answer tailored to Thrumcast on your worst line.

← Back to the knowledge base index