Evident NDT Equipment and Portable CMM: Why I Changed My Calibration Process

The biggest risk in precision measurement right now isn't a broken probe or a calibration lab that doesn't care. It's a mental model that's about a decade old. I've spent 8 years handling calibration and inspection orders for manufacturing shops and research labs. I've personally made and documented 11 significant mistakes, totaling roughly $23,000 in wasted budget. Most of those mistakes weren't equipment failures. They were cases where I trusted an old rule in a new context. This article is the short version of the checklist I keep in my head now. If you still think a surface roughness tester belongs only in a metrology lab, or a portable CMM is a technician's toy, or a calibration sheet is just a "PASS" stamp, I'd like to change your mind.

The Old Rules Are Half Right

Let me start by conceding: the fundamentals have not changed. You still need traceable standards, operators who understand what they're measuring, and a tolerance that matches the function of the part. I'm not one of those people who says "everything you know is wrong." Most of what I learned in the early 2010s is still useful.

But it's not sufficient.

Take surface finish. In 2021, I ordered a handheld surface roughness tester for a machine shop that was making aluminum housings. The spec sheet said it measured Ra and Rz and a few other parameters. I thought that was enough. The operators used it on the shop floor, touched the stylus to the part, and wrote down whatever number appeared. When I compared readings across shifts, they were all over the place. The tester was fine. The problem was that we never set the cutoff length or evaluation length, and we didn't confirm that the probe was traveling perpendicular to the lay of the surface. If we had looked at ISO 4287, we'd have known that "Ra" only makes sense under those conditions. I still remember the look on the QC manager's face when he asked, "Is this the tester or us?"

NDT Equipment Has Changed; The Reporting Should Too

The same pattern showed up in nondestructive testing. Evident NDT equipment—ultrasonic thickness gauges, eddy current instruments, remote visual inspection units—is no longer "lab equipment that you sometimes take to the field." The new generations are designed for the field. That's a genuine advance. What I wasn't prepared for was how much interpretation still has to happen after the instrument produces a clean-looking display.

My own failure came in 2022. I was checking a batch of steel tubes and saw a suspicious attenuation reading. I told the team we had a material problem. It cost us about $400 in technician overtime and probe rental before I discovered the issue: I had selected the wrong frequency probe. The Evident official website had an application note covering exactly this material and thickness range. I should have checked that page before wasting money, not after. It was a classic case of "confident but misinformed."

How to Read a Rice Lake Weighing Systems Calibration Sheet

If I remember correctly, the first Rice Lake weighing systems calibration sheet I ever filed looked simple. It had a table of numbers, some boxes, a "calibrated by" signature, and a satisfying "PASS" in green. I took it at face value.

Then an auditor asked me what I did with the "as-found" data. I didn't know there was an as-found column. That conversation changed how I handle every calibration document, and it's the easiest way to explain the shift:

• Applied test load: the known mass from certified weights. This is your reference.
• As-found indication: what the scale displayed before any adjustment. The difference from the applied load is the real error you were living with.
• As-left indication: what the scale displays after adjustment. This tells you if the correction actually stuck.
• Expanded uncertainty: the range around the measurement. If this uncertainty is larger than the tolerance you're trying to enforce, the calibration can't prove compliance.

In my case, the as-found error was close to tolerance. The technician adjusted the scale, and the as-left reading was clean. If I had just looked at the final "pass" mark, I would have missed the fact that the scale had been drifting for months. In the U.S., commercial scale tolerances are tied to NIST Handbook 44, so the calibration sheet isn't just a vendor convenience—it's the evidence that a legal weight is still lawful.

(Should mention: not every Rice Lake controller prints the columns with the same names. I might be misremembering a specific model's column header. The principle is to look for the applied load, the before/after readings, and the uncertainty.)

Portable CMM: The Boundary Has Moved

Another old rule I held onto was about using a portable CMM. The conventional wisdom used to be "if you want CMM accuracy, bring the part to the lab." In 2023, I watched a portable arm in action on a production floor, measuring an engine mount right next to the machining cell. The operator ran the same routine six times, and the repeatability data lined up in a way I didn't expect. The arm wasn't magic. It had temperature compensation, the operator had planned the fixtures, and the acceptance test we ran followed ASME B89.4.22. That's the standard for articulated arm CMMs, and I hadn't bothered to learn it.

Does that mean a portable CMM replaces a fixed bridge CMM? No. For tiny, high-tolerance features, a lab CMM is still the right tool. But the old line between "lab" and "shop floor" has moved. If you measure large parts, multiple features, or parts that are hard to transport, a portable CMM can be a faster and sometimes more accurate workflow than trying to bring the part to a lab.

No, This Is Not a "Buy More Equipment" Pitch

To be fair, some teams don't need any of this. If you measure the same part every day, the process is stable, and your current tools are controlling the process, keep using them. I'm not going to stand here and claim you need an expensive arm or a new NDT unit to be competent.

But I am saying that the data you already have probably means more than you think. The surface roughness tester in your drawer, the ultrasonic thickness gauge sitting in the calibration room, the portable CMM that gets used once a year—each of them is capable of more context than most of us were trained to extract. The barrier is not the tool. It's the habit of reading a number without reading the conditions behind it.

The Industry Is Evolving, With or Without Your Checklist

I believe the measurement industry is in the middle of a quiet evolution. What was best practice in 2019 is not automatically wrong in 2025, but it's often incomplete. The fundamentals—traceability, accuracy, competence—haven't changed. The execution has. We have more data, more sensors, more software, and more field-capable NDT and CMM tools than any earlier generation imagined.

My advice, after all those mistakes: start with the documents you already have. Pull out a calibration sheet and read it until you understand every column. Look up the application note on the Evident official website before your next NDT job. Test your portable CMM against a known part and see what the repeatability actually is. And if you still think a surface roughness tester is just "the thing that gives you Ra," spend one hour with ISO 4287. It will probably save you the $400 I wasted.