Case study

Not a pilot.
A working field.

Argus runs every day on producing wells for a heavy-oil operator in Trinidad — progressing-cavity pumps on variable-frequency drives, satellite uplink, no control room. The dashboard on this site is the same software that runs that multi-pad field. Screens marked SIM carry sample figures: drive and energy values are measured, production volumes come from the crew’s tank readings.

Monitoring has run continuously since June 2026, brought up pad by pad. In the interest of being straight with you: this field is where Argus was built, not an arm’s-length customer — we explain the relationship here.

Jun 2026Running since

90 daysHistory on file

0On-site servers

AfternoonPer pad to install

Before Argus

Wells you can only
see by driving to them.

012 a.m. tripsA pump could trip at 2 a.m. and go unnoticed until the morning round.

02A truck per readingEvery reading meant a truck, fuel and a drive to the pad.

03Slow failuresSlow mechanical trouble showed up only once it had already broken something.

04No historyNo number, no history, no way to compare one well against another.

What was deployed

An afternoon per pad.
No trenching, no server.

01Field cabinet per padA sealed enclosure reads each variable-frequency drive over Modbus and turns it into clean telemetry. Nothing was replaced; the existing drives stayed exactly as they were.

02Uplink over StarlinkThe pads sit behind satellite and carrier-grade NAT. Argus connects outbound through an encrypted mesh, so there is no static IP, no port forwarding and no on-site PC.

03Live in the cloudTelemetry from every well lands in one dashboard at a 2-second update, with history, comparison across wells and reports.

04Alarms to the phoneDrive faults, power loss, pump-off and torque anomalies push straight to the operator, typically inside a minute.

What it caught

The failures that
used to arrive late.

These are events from the running system, not scenarios.

A slow torque creep that fixed limits miss

On one well torque climbed from 12 to 23 percent over about two hours while motor current barely moved, 16.8 to 17.4 A. A current-based limit would never have fired. Argus runs a slower reference window specifically for this shape and raised it as a torque rise at 18:07, with the from and to values; the crew logged “no flow” at 20:30, 2 h 23 min later.

Power loss told apart from a lost link

A satellite blip and a genuine outage look similar from a distance. Communication loss is classed separately, so a dropped uplink never masquerades as an equipment failure, and a real power loss sends a recovery message once it clears.

Wells that stop without anyone noticing

A stopped or offline well shows its real state, runtime and energy instead of quietly disappearing from the view.

Where it stands

Running, watched
and still growing.

The field runs in monitoring mode with remote control gated behind a site-wide safe lock, so the wells stay under local authority while the platform proves itself day after day. Downhole level and surface pressure instrumentation is going in next, and the same kit extends to further pads without adding infrastructure on site.

Want the same view of your own field? Email info@argusautomation.tech or request access below.

Put your field online.

First well live in days, not quarters.

Not a pilot.A working field.

Wells you can onlysee by driving to them.

An afternoon per pad.No trenching, no server.

The failures thatused to arrive late.