Industrial remote monitoring: the operational truth machine

Walk through how most sites actually keep tabs on their equipment and you find the same machinery underneath: a clipboard round, a laminated checklist, a shared spreadsheet, perhaps a logbook by the plant-room door. Someone walks the route, reads a gauge, glances at a thermometer, ticks a box, types a number. Done honestly and diligently, this still records only one thing — the state of the asset at the instant a person stood in front of it.

That distinction matters more than it sounds. A reading taken at 08:00 and another at 16:00 tells you nothing about the eight hours in between. The spreadsheet shows two tidy rows; the asset lived through twenty-eight thousand seconds the spreadsheet knows nothing about. If pressure sagged at 10:15 and recovered by lunchtime, the log says everything was fine — because every time someone looked, it was.

Manual records also inherit every human limitation of the person keeping them. Rounds get shortened on busy days, gauges get read in poor light, numbers get transcribed wrong, and a check that has passed five hundred times in a row gets glanced at rather than read. None of this is negligence. It is what happens when you ask people to be sampling instruments, which is a job people are poorly built for and sensors are very good at.

So the honest description of a manual log is not "a record of the asset". It is a record of attendance — a series of snapshots whose spacing is set by staffing, not by risk. The asset's true history is the line that runs between the dots, and nobody has it.

Consider what actually goes wrong on a typical estate, and when. A cold store drifts upward overnight, when nobody walks past it, and the stock inside is quietly ruined hours before the morning check finds a warm room and a hard decision. A compressor trips on Friday evening and the line that depends on it is discovered dead on Monday, turning a restart into a lost shift. A pressure system bleeds down slowly over days, so each individual reading looks plausible and the trend that would have told the story never appears on paper. A loading-bay door is left ajar and a temperature-controlled space spends the night fighting the weather at full power.

Each of these has the same shape: a state change that occurred between observations, discovered late, paid for in full. The bill arrives as wasted engineer callouts to assets that were fine an hour ago and are fine again now; as stock written off because nobody can say when the excursion started; as downtime measured from when the fault happened, not from when it was found; as service-level penalties for outages that ran on silently; and as disputes that turn on whose recollection is believed.

There is a second, quieter cost: the checking itself. Sites spend skilled hours every week sending people to confirm that nothing has changed — which, most days, is exactly what they find. The rounds exist because the gap exists. Close the gap and most of that routine walking becomes unnecessary, freeing the same people for the work that actually needs hands.

The uncomfortable truth is that "we thought it was fine" is the default state of any asset monitored by sampling. It usually is fine. The whole cost of the approach is concentrated in the occasions when it wasn't, and the log couldn't tell you.

Industrial remote monitoring, in the form we build it, is deliberately simple: small battery-powered wireless sensors fixed to existing equipment, each watching one thing that matters, reporting one-way to a live dashboard. There is no control path, no intervention in the asset, and nothing for your IT estate to absorb — the sensor observes, transmits, and the dashboard and alerts do the rest.

What the sensors watch falls into two natural categories. The first is binary state: a door open or closed, mains power present or absent, a pump or motor running or stopped, a tank float above or below its mark. The second is thresholds on continuous values: pressure above or below a set band, temperature or humidity drifting outside the range a cold chain or a server room can tolerate, a level falling faster than it should. In both cases the question being answered is the one the clipboard round was always trying to answer — has anything changed? — except now it is answered continuously instead of twice a day.

When a state changes or a threshold is crossed, the change is timestamped and an alert goes out by email or SMS to the people who need it, while the dashboard keeps the running record. That is the entire proposition. It does not predict the future and we don't pretend it does; it tells you the present, immediately, which is precisely the thing the spreadsheet never could. For the specific disciplines this enables, see our pages on real-time pressure monitoring and IoT environmental monitoring linked below.

The traditional answer to this problem is SCADA: hardwired instrumentation, an RTU or PLC per area, control-room software, integration engineering, and a budget to match. For a major process line that genuinely needs supervisory control, that is the right tool. But most of the assets that quietly cost money — the cold rooms, compressors, plant-room gauges, storage tanks, remote kiosks, back-of-house doors — were never going to justify it. They are not worth wiring, not worth a dedicated controller each, and not worth the months of integration work; so under a SCADA-or-nothing model, they get nothing, and "nothing" in practice means the clipboard.

Retrofit monitoring removes that choice. A self-contained wireless tag fixes to the asset you already own — no rewiring, no panel work, no downtime to install, no software project. The sensors are low-cost enough that instrumenting the long tail of ordinary equipment finally makes economic sense, and a battery life of around five years means the estate doesn't acquire a new maintenance burden in exchange for visibility.

It also changes who can have monitoring at all. A site with no control room, no instrumentation team and no capital project on the horizon can still put real-time visibility on the equipment it depends on, asset by asset, starting with whatever hurts most. The comparison that matters is not retrofit monitoring versus SCADA; it is retrofit monitoring versus the consequence of the next event nobody saw. Monitoring should cost less than the thing it prevents you discovering late — and on existing equipment, it now does.

The alerts are what sell the system; the record is what you end up valuing. Once an asset reports its own state changes, the site owns a timestamped history of what actually happened — when the temperature left its band and when it returned, when the door opened and for how long, when power dropped and when the pump restarted. Not a reconstruction, not a best recollection under pressure, but the sequence itself.

That changes conversations that used to run on memory. An insurer asking when the excursion began gets a timestamp, not an estimate. A customer disputing storage conditions can be answered from the record. An engineer arriving on site starts from what happened and when, rather than interviewing whoever was on shift. Internal post-mortems stop being arguments about whose round it was and become a read-through of the actual sequence of events.

There is a cultural payoff too. When the record keeps itself, the people who used to keep it stop being the system's weakest component and the implicit suspects when it fails. Rounds shrink to the inspections that genuinely need eyes and hands — a point regulators themselves make about the limits of human checking, as the sources below set out.

A spreadsheet is a diary of when somebody looked. Continuous state-change monitoring is the operational truth: what the asset did, all day, every day, written down by the asset itself. Low-cost, retrofit, CE-certified hardware with a five-year battery and real-time email and SMS alerts puts that truth on the equipment you already own — usually for less than the cost of the next event you would otherwise have discovered late. If there is a piece of kit on your site whose state you are currently guessing between checks, talk to us.