Three steps. No risk to your operations.
Everything starts from data you already have. Nothing is installed, nothing is connected to your control system, and no alarm is ever closed by software.
1. Send one month of data
An export from your historian for one line: pressures, flows, pump speeds, valve commands, alarm log. Under NDA.
2. We replay it
Offline, read-only, on an encrypted machine that can stay on your premises. You see what we would have caught, where, how fast, and every alarm with its cause.
3. Run it alongside
A pilot on one line, reading SCADA through a one-way link. Your operators stay in charge. We report false alarms per month, whether good or bad.
Read-only. Explained. Approved by a person.
Read-only
We read SCADA and historian data. We never write to them.
Outbound only
Nodes, if you add them, only send. Nothing can command the line.
Never cleared
Software annotates alarms. It never closes, suppresses or hides one. It says UNEXPLAINED when it does not know.
A person approves
Recommendations pass a fixed safety rulebook, then a named person decides.
What we need from you
- Pipe length, diameter, elevation, valve and pump positions
- Sensor list with positions
- One to three months of history, 1-second if you have it
- Alarm and event logs, and any known events: tests, incidents, maintenance
- A technical contact who knows the line
What to expect
- Historian data holds 1-second values, so location from it is hundreds of metres at best
- Slow leaks and seeps are outside what any pressure or flow method can see
- Everything we show is labelled: measured, inferred, assumed or predicted
- All our figures so far come from simulation; yours will be the first real ones
Ready when you are.
A short call to agree the data list and the NDA is enough to start.