Know what your machines actually run.

Two current sensors clip round the power cables on each machine. From that we work out when it was cutting, when it was powered up doing nothing, and when it was off. No PLC access, no rewiring, no IT project, and nothing that stops working because a machine is thirty years old.

How the measurement works

Almost everything else on this page follows from one decision: reading the power signal rather than talking to the controller.

Two sensors, not one

A single clamp on the incomer only tells you whether the machine has power. That counts a machine warming up, or sat in a fixture change with the hydraulics running, as production. So a second clamp goes on the main drive. The difference between the two is what separates powered-on from actually working.

Both are split-core, which means they clip around the existing cable. Nothing is cut, nothing is spliced, and the cable stays exactly where it is.

Calibrated on your floor

Every machine draws a different current signature, and the same model running different work draws differently again. We set the thresholds on your machines rather than shipping a guess and hoping. That's an afternoon during commissioning, and it's the difference between a number you trust and a number you argue with.

SensorSplit-core current transformer, clipped round the existing cable
Per machineTwo: one on the power incomer, one on the main drive
Edge unitArduino Opta, DIN rail mounted, CE and UKCA marked
NeedsPower and a network connection. Nothing else.
Controller accessNone. Nothing connects to the PLC or CNC control.
Install timeAround ten minutes a machine, by your own electrician
Optional inputsSpare inputs for a proximity switch or photo-eye where you want a true part count
HostingAWS, Ireland. Your data stays in the EU.

Three states, and why the middle one matters

Most shops measure the first and third and quietly count the second as production. That's where the missing hours live.

Cutting

Drive current above the threshold we set for that machine. This is the only state that counts as productive time.

Powered, not cutting

Machine is on but the spindle isn't loaded. Setups, first-off inspection, waiting on material, waiting on an operator, or simply left running through a break.

Stopped

No power draw. Breakdowns, planned maintenance, or nobody on that machine this shift.

If a stop lasts longer than the threshold, the operator gets a prompt to tap a reason. One tap, not a form. The utilisation number is right whether they bother or not; the reason codes are what turn it into a Pareto worth acting on.

What you get to look at

Four views, on a desk, a phone, or a screen hung on the shop floor.

Live shop floor dashboard showing utilisation, machines running now, downtime hours and OEE availability across fourteen machines

Live status. Every machine, its current state, how long it's been in it, and the reason where one was logged.

Downtime Pareto

Lost hours ranked by cause, so the argument about what to fix first has an answer. The unattributed row is deliberate: it shows you how much you don't know, rather than folding it into something that looks tidier.

Downtime Pareto chart ranking causes of lost hours

Utilisation by machine

The whole floor ranked. This is the view that shows you the spread — the machines you worry about are often fine, and the quiet one in the corner is the problem.

Utilisation by machine, ranked from 88 per cent down to 12 per cent

Production timeline

The shift replayed as a strip of states. Patterns that are invisible in a daily average are obvious here: the machine that never starts before eight, the gap that appears every Friday afternoon.

Production timeline showing each machine's running, idle and stopped periods across a shift

Improvement tracker

Every monitoring system gives you a Pareto. Almost none of them show whether the fix worked. Each action is tracked against the cause it targeted, with the hours it gave back and whether the gain held or quietly slipped.

Improvement tracker showing weekly utilisation rising alongside open improvement actions and hours recovered

What it won't do

Worth knowing before you spend anything, rather than after.

It can't count parts on its own
Current draw tells you the machine was cutting, not how many pieces came off it. Add a proximity switch to a spare input and you get a true count. Without one, you get run time.
It can't see inside the controller
No programme names, no feeds and speeds, no alarm codes. That's the trade for never touching the CNC control and working on machines that have no data port at all.
It won't tell you why on its own
The clamps see that a machine stopped. Someone has to say why, or you have to infer it. That's what the reason prompt is for, and why asset tracking exists as the other half.
It needs a sparky for an hour
Fitting a clamp round a live power cable is electrical work. Your own electrician can do it, but it isn't a plug-in-and-go device and we won't pretend otherwise.

Three machines, thirty days

Find out what the floor actually does before you build a business case, not after. Free if you go ahead afterwards, £250 if you don't.

First

Site visit

We walk the floor and agree which three machines to instrument. Pick the ones you reckon are busiest.

Then

Install and calibrate

Your electrician fits the sensors. We commission the units and set the thresholds on your machines.

For a month

Thirty days of data

Live dashboard from day one, with a check-in each week. Ignore the first fortnight; it takes that long to settle.

At the end

We go through it

Your real utilisation figure and your downtime Pareto, and an honest view on whether doing the rest of the floor is worth it.

Book a 15-minute call

Most shops guess eighty. Then they measure.

Fifteen minutes on the phone. We'll ask what machines you've got and tell you straight whether this is worth doing on your floor.

Book a 15-minute call