Wavepinpart of OOH
Product · Explain

Every alarm, explained.

Each alarm gets a cause and the evidence behind it. Nothing is ever cleared or hidden.

Ready for a data pilotRuns offline on one month of your history. Nothing is installed.
1

Names the cause: an operation, an instrument, a transient, a leak, or “unexplained”

2

Keeps the evidence trail for every call

3

Works on your existing alarm and historian logs, offline and read-only

What it looks like

Six alarms, six causes

Sudden leak>99% sure

Leak near km 18.1

About 15.6 litres per second. Somewhere between km 17.8 and km 18.6.

Also weighed: Faulty sensor <1% · All normal <1%
Valve half shut>99% sure

Valve 1 is choking the flow

No product is lost. The valve is more closed than it should be.

Also weighed: Faulty sensor <1% · Does not add up <1%
Faulty sensor>99% sure

Bad reading from the outlet pressure gauge

The pipeline is fine. One instrument needs a check.

Also weighed: Valve problem <1% · Does not add up <1%
Weak pump>99% sure

The pump station is weak

It delivers about 8 m less head than it should. No product is lost.

Also weighed: All normal <1% · Leak <1%
Pump trip83% sure

Something does not add up

No single cause fits. A person should look at this one.

Also weighed: Leak 11% · Valve problem 3%
Nothing wrong90% sure

All normal

Nothing to do. Keep running.

Also weighed: Leak 2% · Faulty sensor 2%
Flag: the line is not behaving like the model.

Simulated cases. Each card is the system's verdict at the end of the case, with the runners-up it weighed.

What it needs

  • Alarm and event logs
  • Pump and valve commands
  • Pressure and flow history

What it is not

  • It never closes an alarm. A person decides.
  • Two faults at once are still often read as one
Works with
Ready for a data pilot

Leak

Find the leak. Know where.

Open →
Proven in simulation

Decide

What if we did this?

Open →
Proven in simulation

Twin

The whole line, every metre.

Open →
Proven in simulation

Place

Where the next sensor pays back.

Open →