The Short Version: Your CMM Isn't the Weak Link
If you're responsible for measurement quality, the weakest link in your lab is almost never the instrument brand. It's the training and verification discipline around the hardware.
A Hexagon CMM without structured Hexagon CMM training is an expensive doorstop with a probe. A Hexagon laser tracker that's trusted on paper but never verified in the field is a source of false confidence—which I'd argue is worse than no data at all. And the boring stuff—a clogged vacuum filter, an unstable power circuit, an uncalibrated hand tool—causes more bad measurements than any single piece of hardware failure.
I've spent the last four years reviewing inspection results before they reach customers. Roughly 200+ unique instrument measurements pass through my review annually—maybe 180, I'd have to check the actual log. In our Q1 2024 quality audit, I rejected 11% of first-delivery inspection reports. Not because the instruments were bad. Because the people running them hadn't been trained to catch their own setup errors.
Why I Get to Have an Opinion
I'm a quality and compliance manager for a contract inspection company. Every calibration certificate, every CMM program, every inspection report that goes out the door crosses my desk first. When we acquire equipment or write acceptance criteria, I'm the one who signs off.
In 2022, I implemented our laser tracker field verification protocol after an incident I'll describe below. Before that, I spent two years on the receiving end of bad data from otherwise good machines. If I remember correctly, the ratio of setup-error to hardware-failure across our corrective actions is about 6 to 1. I might be misremembering the exact figure, but that's the direction of the imbalance.
My point: I don't write about this from a service manual. I write about it from rejection notices and redo costs.
The Equipment Matters Less Than the System Around It
Hexagon CMM Training: What Actually Changes
The most common question I get is "which CMM should we buy?" The harder question—the one people should be asking—is "how do we make sure the operators can run it correctly?"
Here's something vendors won't tell you: the machine's accuracy specification assumes the operator knows how to perform probe qualification, choose the right alignment strategy, and manage temperature effects. None of that comes standard. A probe tip that's been dropped and micro-nicked, a star probe that wasn't dressed correctly, a datum that was misidentified—these are operator problems, not machine problems.
We run Hexagon CMMs with PC-DMIS, and the difference between operators who've been through formal Hexagon CMM training and those who "learned on the job" is stark. Trained operators catch errors in the pre-inspection phase: a dirty probe tip, a part not fully seated on the fixture, a temperature log that hasn't stabilized. Self-taught operators develop consistent, reproducible errors—which are the worst kind, because they look correct. The measurements are repeatable. They're just wrong.
People think expensive instruments deliver better measurements. Actually, verified instruments deliver better measurements—the price follows the verification. The causation runs the other way.
Hexagon Laser Tracker: Verify in the Field, Not Just on Paper
The laser tracker is a different beast from a CMM. It's portable, so it ends up on factory floors where the temperature isn't controlled, the floor vibrates, and people walk through the measurement volume. What most people don't realize is that the tracker's specified accuracy assumes stable temperature, minimal vibration, and clear line of sight. The performance standard for laser trackers—ASME B89.4.19 or ISO 10360-10—defines what the instrument can do under controlled conditions. Field conditions are not controlled conditions. Real conditions are rarely that tidy.
Here's what happened in late 2023. Our field team used a Hexagon laser tracker to verify a large fixture that we had solid reason to believe was good. The tracker reported the fixture was off by 0.4 mm. We re-ran it three times; same result. The machine was within its factory calibration spec. The equations said "bad part." My gut said "bad setup." This was a direct conflict, and the frustrating part was that we had no protocol to distinguish between them at the time.
The culprit turned out to be a thermal gradient. A loading dock door had been left open, and the sun created a temperature difference of about 6 °C across the measurement volume. The tracker's compensation software corrected for some of it—but not all, because the gradient was uneven and moving. We rejected a good fixture based on a bad environment, told the client their part was out of tolerance, and then had to retract it. That mistake cost us a $22,000 redo and four days of delay.
Since then, our protocol is simple: log environmental conditions before every tracker session and run a verification against a known reference length. If the environment is outside the tracker's compensation range, we don't measure. (Should mention: that adds about 15 minutes to each setup. Worth it, every time.)
The Supporting Tools Nobody Talks About
There's a whole class of measurement errors that has nothing to do with the CMM or the tracker itself. Three tools catch most of them.
Vacuum filter on the air supply line. CMMs with air bearings need clean, dry air at consistent pressure. A clogged vacuum filter (or coalescing filter) causes pressure drops so small you don't see them on the shop gauge, but the CMM drifts. We had a lab blaming a CMM for three weeks of unstable readings. The actual problem was a filter element that hadn't been changed in 11 months; the manual said to inspect it monthly. Nobody did, because nobody considered it measurement equipment. It is. Change the filter, check the dryer. This is the most boring advice I'll give you, and it's probably the most valuable.
Electrical supply. Before you chase a measurement error, verify the power that feeds the machine. We use a 374 FC True-RMS AC/DC clamp meter to check voltage stability at the machine intake. The True-RMS capability matters because average-responding meters understate distorted waveforms. We had a CMM that developed sporadic Z-axis error. A 374 FC logging on the circuit caught voltage dips every time a compressor kicked in; the CMM was on the same line. We moved it to a dedicated circuit and the error disappeared. If we'd chased the Z-axis without checking power, we'd still be swapping scales.
Starrett angle finder. For all the expensive equipment, I still keep Starrett angle finders in every inspector's toolbox. If you're wondering how to use a Starrett angle finder correctly in a QC context, here's the workflow: clean the reference surface first—a 0.05 mm burr will throw off a small tool. Zero the blade against a certified surface plate or a known square. Place the base against the feature, rotate the blade until it contacts the work surface, lock the blade, and read the angle. Digital models feel more reliable than analog ones, but I've seen a digital unit drift by 0.2° after bouncing around in a toolbox for months—so zero it before each use. It's not a substitute for a CMM. It's a 30-second sanity check that catches gross setup errors before they waste coordinate measuring machine time.
When This Advice Doesn't Apply
I recommend formal Hexagon CMM training for anyone running a CMM weekly. But if you only run a few inspection programs per quarter, the training investment probably won't pay back—outsource the first few programs to a metrology service provider and have them supervise your operator instead. That's a legitimate alternative, not a compromise.
If your laser tracker works only in a climate-controlled metrology lab, the full environmental verification protocol I described is overkill. Do the factory-recommended checks and stop there.
The Starrett angle finder is fine as a setup aid, but it won't give you traceability. Don't let a $150 tool make you forget why you bought a $200,000 one. And the 374 FC clamp meter, as useful as it is, won't tell you about grounding quality beyond what it's designed to measure.
No single tool or course covers every situation. The system around the equipment—filters, power quality, environmental logging, verification routines—determines whether your data means anything. The training makes sure you know what to look for; the tools make sure you catch it; the discipline makes sure you act on it.
