My phone rang at 2:47 on a rainy Tuesday night. It was the night-shift supervisor at a cold-storage warehouse, two hours from my workshop. Their 500 kVA standby set — the machine backing up a warehouse full of frozen food — had just failed to start.
He knew because a small black box wired to the controller had texted him twelve minutes earlier. That’s remote monitoring, the whole of it. Everything else is sales talk.
Clients ask me about that little module on almost every quote. Usually right after they see the price. Here are the three questions that actually matter.
Mostly, no. It’s not a crystal ball, and anyone selling it as one is lying to you.
What it does is narrower — and, honestly, more useful. It tells you things the moment they happen instead of the moment the power fails. Failed auto-start. Common fault. Battery voltage sliding downhill week after week. Fuel level dropping.
Here’s the thing most people miss. A standby generator sits there for months. NFPA 110 wants emergency systems run-tested every month, and most commercial sites I walk into would fail that test on paper before the engine even turns over. I’ve asked site managers, face to face, “when did this set last run?” and heard “uh… last year?” more times than I can count. Not dishonest people. Busy people.
The box tells you the weekly auto-test failed on a quiet Tuesday morning. Not during the blackout on Friday night. That time gap — weeks of it — is the entire product. You’re buying time, not prediction.
And one warning, from a job I did a few years back. A client got a “mains failure — generator running” alert at 3 a.m., drove to the site, and found the set humming along happily. The building was still dark. The ATS hadn’t transferred the load. The alert told the truth: the generator ran. It took a human standing in front of the switchboard to see the rest. The box reports. It doesn’t think.
This is where my opinion gets strong, so take it or leave it: most cloud dashboards are toys you’ll open twice. Vendors love subscriptions. What has value is the alert reaching a phone at 2 a.m. — not forty-seven parameters rendered as pretty graphs.
For one standby set at a site where someone is nearby most days, a basic GSM module that sends a text on failed start, common fault, low battery, and low fuel is plenty. The hardware runs a few hundred dollars in the projects I’ve priced, plus a cheap SIM with a data plan that costs less than a sandwich a month. That’s it. That’s the whole monitoring system.
Now, if you run several sites, unmanned sites, or a rental fleet — different story. Then telemetry earns its keep. Run hours feed your maintenance schedule, and your invoices if you rent the sets out. Fuel levels across five sites on one screen beat five phone calls. A single dashboard for ten sets an hour apart is genuinely worth paying for.
What I’d never pay extra for is the “AI predictive maintenance” label some module makers slap on the box. Smile and move on. What predicts a failure is oil analysis, a proper load-bank test, and a battery you actually replace on time — not an algorithm staring at coolant temperature graphs. An old technician I worked with says remote monitoring makes people lazy, checking a phone instead of walking the set. He’s got half a point. The screen is no substitute for ears and a flashlight. But “half right” is the most you can say for him, because the half he misses is the 2 a.m. text.
Whatever tier you pick, the box has to do four things. Nothing more is required:
Sure. If you actually will. You won’t.
Nobody does. I’ve seen maintenance logs that look perfect — dates filled in, boxes ticked — and an air filter behind them that hadn’t been touched since installation. The walk-over plan fails the same way every diet fails: not because it’s wrong, but because Tuesday gets busy.
A proper maintenance routine still has to happen with human hands, module or not. The box can’t change oil. It can’t spot the bird’s nest inside the canopy — I’ve found two, one of them with actual eggs. It can’t hear a bearing going rough or smell coolant cooking. What it does is force honesty: the test ran or it didn’t, the battery is at 12.4 volts or it isn’t, and the numbers are sitting on someone’s phone instead of in a logbook nobody reads.
My favorite cautionary tale is a site with the full setup. Beautiful dashboard. Color-coded everything. And a low-coolant alarm sitting unacknowledged for three weeks. The screen was perfect. The radiator was nearly empty. Technology doesn’t fix not caring — it just documents it.
So here’s where I land, after twelve years of these conversations. If the generator matters — hospital, cold storage, a production line where a failed start costs real money — put the box on it and wire it to the four things above. The controller it talks to is worth understanding too; a good controller with the right module is a combination that quietly pays for itself. If the set is a backup for a backup, the basic SMS module is plenty, and anyone telling you otherwise is selling.
And whatever you buy — this is the part no module can do — put the phone number of the person who’ll actually drive out there at 2 a.m. in the alert list. Not the facilities manager’s office line. The guy with the keys. If you’re sizing a system for a site like a hospital, where there’s no second chance, that number matters more than any feature on the datasheet. Talk to us if you want help picking one — Pavo’s engineers spec monitoring with every set we build, and we’ll tell you straight which tier your site actually needs.
Email：[email protected]
Work：Monday to Friday,9:30-18:30
