Ask most shop owners what their spindle utilization is, and the number they give you tends to run high, half the day or better. Then they measure it, and the real figure comes in well under that. Industry surveys put average spindle utilization somewhere in the 20 to 30 percent range, which means the machine you paid six figures for is cutting metal less than a third of the time it's scheduled to run.
That gap between what you assume and what's actually happening is worth real money. Before you spend on another machine to add capacity, it's worth knowing how much of the capacity you already own is sitting idle. You don't need a smart-factory overhaul to find out, either. You need to know when each spindle is running and when it isn't, and there's a way to get that off every machine on your floor, including the ones built long before anyone thought to put them on a network.
There's a temptation to jump straight to OEE. It's the number everyone talks about, and it's a useful one, but it isn't where you begin. OEE folds in performance and quality alongside availability, and pulling all of that together turns a quick answer into a long project.
Utilization is the simpler place to start. How much of the scheduled shift is the spindle actually cutting? That one figure tells you where your real capacity is going. It also helps to know why a machine is sitting, because a spindle waiting on a setup is a different problem than one waiting on a part or an operator, and you can't fix what you've never measured.
Every machine can tell you whether it's running or sitting. What changes from one to the next is how easily you can get at that, and most floors have a mix.
On a newer machine, the control is already built to share what it's doing. It speaks a common language, usually MTConnect or OPC-UA, that any monitoring system can read. You connect it to the network and it reports its own status, running or idle, with nothing else added. That's the easy end of the floor, and the older machines just take a little more work.
An older machine that's still on the network usually can't hand its data over quite so directly, but the data is there. A Fanuc control, for example, holds it in its own protocol, FOCAS, and a small adapter translates that into the same place as everything else.
The one people assume they can't touch is the older machine with a control that shares nothing at all. You don't pull data off it, because there's none to pull, so you watch it from the outside instead. A clamp on the spindle's power line picks up when it's drawing current, which tells you it's cutting, or you read the same stack light the machine already uses to show its status, all without touching the control itself.
That last case matters more than it sounds, because it's the reason a machine from the 1990s or older doesn't get left out. If it has power going to it, you can tell whether it's working.
The usual reason shops hold off on monitoring a mixed floor is the belief that everything has to be standardized first, and it doesn't. You take each machine as it is, use whatever method suits its age, and the data lands in one place no matter the machine brand or what year it was built.
It stays practical because the external-sensor route costs little per machine and installs without downtime, so nothing comes off the floor and no control gets rewired. For a shop running new machining centers alongside equipment that's twenty or thirty years old, that's the difference between seeing the whole floor and seeing only the newer end of it.
The data is only worth as much as what you do with it. Once you can see where spindle time actually goes, the decisions get easier. You can find the real bottleneck instead of guessing at it, put real numbers behind a capital purchase, or hold off on one you don't need because the capacity was there all along. You can schedule around your actual constraint, and spot the machine that's quietly down more often than anyone realized.
None of it takes new equipment, just a clear picture of what you already have running.
If you're working a mixed floor and want to know what it would take to see all of it, that's a conversation worth having. DiPaolo services every brand and age of machine tool, which is exactly what it takes to walk a floor and pick the right way to pull data from each machine on it.
Need help deciphering it all? Check out our ROI Calculator.