How it works

Three diagrams. No IT project.

A pod on the asset, a cellular modem in the pod, a thirty-day baseline before anyone talks about a spend-up.

Cellular firewall bypassPod reports from the asset on a built-in cellular modem straight to the Thrumcast cloud, never touching the plant network.Asset tagMotor · pumpPress · conveyorPodThrumcastcloudAnomaly timelinePlant networkVLAN · firewall · active directoryPlant network — never connectedPlant IT is never asked, never looped in. If their network is down for a patch, the pod keeps reporting.

Step one

Plant IT is never in the loop.

The pod reads vibration, current draw, and skin temperature. It reports on a built-in LTE-M modem straight to the Thrumcast cloud. The plant network never sees the traffic.

  • ✕
    No VLAN request.
  • ✕
    No firewall exception.
  • ✕
    No port opening.

If the corporate network is down for a patch, the pod keeps reporting. Cellular out, not LAN up.

Step two

Four steps. Under ten minutes.

No laptop, no PLC, no tech on-site. The crew on the floor does the whole thing in a coffee break.

Four-step pod installation: bolt on, power, pair, baseline01Bolt on02Power03Pair04BaselineUnder ten minutes per asset · no laptop · no PLC · no IT ticket
  1. Step 1

    Bolt on.

    Epoxy-mount or magnetic mount on the bearing housing. Two bolts on most assets, three on a conveyor idler. No shaft coupling and no tapped holes.

  2. Step 2

    Power.

    A single AA cell pack. Cold-rated for an unheated building. We ship the first pack; you swap it every nine months.

  3. Step 3

    Pair.

    The foreman scans the QR on the pod with a phone, types in the asset tag, and confirms it shows up on the dashboard. That’s the pairing step. No provisioning server.

  4. Step 4

    Baseline.

    The pod starts its thirty-day timer from the moment it shows green on the dashboard. The first teachable day is day one.

Step three

Four weeks of unsupervised learning, then the alert box turns on.

Here is what the pod is doing during the first thirty days, and when the first real alert can reach someone.

Thirty-day baseline timelineA horizontal timeline of what the pod does during the first thirty days: calibration, baseline training, quiet drift checks, then the first anomaly alerts.Day 0Day 7Day 14Day 25Day 30First alert ≥6σDays 1–6Calibration driftDays 7–14Baseline model trainsDays 15–25Quiet drift checksDays 26–30First anomaly alertsMost pilots see the first real alert between day 28 and day 35. A clean first month is also a result — we'll tell you either way.
  1. Days 1–6

    Calibration drift only. We discard the first week; the bearing temperatures are still bedding in.

  2. Days 7–14

    Baseline model trains. The pod learns what “normal” looks like at this duty cycle, this load, this shift pattern.

  3. Days 15–25

    Quiet drift checks. If a signal wanders outside the model, the pod flags it but does not alert — your engineer reviews the chart.

  4. Days 26–30

    First anomaly alerts. Anything that breaks the trained envelope by ≥6σ pages the on-call contact you gave us at onboarding.

Ready when you are

Ready for one line?

Send three fields — name, email, what you'd like us to look at. We come back inside one business day with pricing for your tier and the install date for your first pod. No plant-IT meeting required.

View pricing Read the security posture