Guide
Portable CMM arm vs bridge CMM: when is an arm accurate enough for your part?
A portable CMM is a coordinate measuring machine that is taken to the part, most often an articulated arm with a touch probe or laser scanner, and its stated error is several times larger than a good bridge CMM’s, and more so on small features. That makes an arm a sound choice for large weldments, frames and castings with tolerances in the tenths of a millimeter, and a poor one for bearing bores, close-tolerance hole patterns and small machined parts.
By Saola Manufacturing engineering team · Published · Updated
Buyers usually meet this question in one of two ways: a supplier offers an arm report where the purchase order said “CMM report,” or a part is too big to fit on any bridge machine nearby. This guide compares the two machine types by their published test figures, explains what those figures leave out, and sets out when an arm result should be accepted. How a CMM report is read, and when a purchase order should ask for one at all, is covered in the CMM inspection guide.
On this page
How does an articulated arm differ from a bridge CMM?
A bridge CMM moves a probe along three linear axes on a rigid frame in a controlled room; an arm is a chain of rotary joints with encoders, moved by hand, whose tip position is calculated from the joint angles.
That structure explains most of the accuracy gap. Every joint angle error is multiplied by the distance to the probe tip, the operator’s hand loads the arm while measuring, and the arm is usually mounted on a stand or magnetic base next to a part on the shop floor rather than in a temperature-controlled room. In exchange, the arm reaches around and inside large parts, can be carried to a fixture, a weld cell or an assembly, and with a laser line scanner collects dense surface data quickly.
The two machine types are also tested differently. Bridge CMMs are accepted and reverified for length measurement under ISO 10360-2. Articulated arms have their own part of the series, ISO 10360-12, and in the US the older ASME B89.4.22 method for evaluating articulated arm CMMs, reaffirmed in 2024. Because the tests differ, the numbers below show the order of magnitude, not a like-for-like ranking.
| Item | Bridge CMM (CRYSTA-Apex V series) | Articulated arm (Absolute Arm 8525-7) |
|---|---|---|
| Test standard | ISO 10360-2 | ISO 10360-12:2016 |
| Stated length error | (1.7 + 3L/1000) µm, L in mm | 0.031 mm (31 µm) maximum |
| Worked example at L = 1,000 mm | 4.7 µm | 31 µm (single figure over the working volume) |
| Working volume or reach | Fixed measuring range of the frame | 2.98 m reach |
| Temperature condition | Accuracy specified over 16–26 °C with compensation | Shop-floor use; check the maker’s stated range |
How much less accurate is a portable CMM arm in practice?
On published figures, a 3 m class arm is several times less accurate than a premium bridge CMM over a meter of travel, and the gap widens on small features, where the bridge machine’s error shrinks and the arm’s does not.
Mitutoyo specifies its CRYSTA-Apex V bridge machines at an E0,MPE of 1.7 + 3L/1000 µm, against 2.5 + 4L/1000 µm for what it calls a standard CMM, over 16 to 26 °C with temperature compensation (Mitutoyo Bulletin 2297(2), 2022). TU Darmstadt lists its 2.98 m arm at a maximum permissible length error of 0.031 mm per ISO 10360-12:2016 (TU Darmstadt, PTU). For a 100 mm feature the bridge figure falls toward 2 µm, while the arm figure is a single value for the whole volume, per ISO 10360-2 and ISO 10360-12 figures as published by the two sources above.
Then add what the test does not include. An arm’s acceptance figure is measured by a trained operator on calibrated artifacts; on a real part, operator technique, probe force, base stiffness, re-positioning (“leapfrogging”) to reach the far side and shop temperature all add error. A scanner adds its own error on top of the arm’s, and shiny or dark machined surfaces scan worse than matte ones.
When can you accept an arm report instead of a bridge CMM report?
Accept it when the arm’s uncertainty on that feature is a small fraction of the tolerance, which in practice means large parts with loose tolerances: weldments, frames, brackets, fixtures and large castings checked for location, profile and fit-up.
A common rule is that the measuring uncertainty should be no more than about a tenth of the tolerance, and a quarter at the outside; the reasoning, and when it can be relaxed, is in the 10-to-1 gauge accuracy guide. Apply that to the figures above and the boundary becomes clear without any shop-specific claim: an arm whose stated error is in the hundredths of a millimeter fits tolerances in the tenths, not tolerances in the hundredths.
- Usually fine for an arm: hole locations on a welded frame, flatness of a large mounting face measured for fit-up, the profile of a casting before machining, reverse-engineering a worn part.
- Usually not: bearing bore diameters and their position, a tight hole pattern on a machined plate, true position at MMC on small parts, thread and small-bore features.
- Gray zone: large machined parts with a few tight features. Measure the large-scale geometry with the arm and the tight features with dedicated gauges or a bridge CMM.
Why are portable arms common on welded and large parts?
Because the part cannot go to a bridge machine, and the tolerances on weldments and large structures are usually loose enough for the arm.
A welded frame several meters long distorts as it cools and is often measured in its fixture, before and after stress relief, to decide where to machine. Moving it to a CMM room is impractical, and a bridge machine big enough is rare. Weldment drawings also tend to carry general tolerances and profile tolerances in the tenths of a millimeter, which suits the arm. The machined interfaces on the same weldment, such as bores for shafts, are the exception, and are better checked with bore gauges or after machining on a fixed machine.
One more practical point: an arm measures the part as it sits. If the part is restrained in a fixture, say whether the drawing applies in the free state or restrained, and record which in the report.
What should the purchase order and the arm report state?
The order should say which features may be measured with a portable arm, and the report should identify the machine, its last verification and the setup, so you can judge the result.
Ask for these in the report header or notes:
- Machine type, model and reach, and whether a contact probe or scanner was used for each feature.
- Date and standard of the last verification (ISO 10360-12 or ASME B89.4.22), with the stated error.
- How the part was aligned to the drawing datums, and whether the arm was repositioned during the job.
- Part and room temperature at the time of measurement.
- Restraint condition: free state or in a named fixture.
On the purchase order, list the features where an arm is acceptable, for example “portable CMM acceptable for weldment profile and hole location; machined bores by bore gauge.” That prevents a supplier from measuring everything with the arm and prevents a receiving inspector from rejecting a valid result for using the wrong machine.
What does this mean for a Saola order?
Our lead CNC plant has no CMM, fixed or portable, so a drawing that requires a CMM report has to be discussed at drawing review before quoting.
Parts are machined at the lead plant in Vietnam on CNC milling and CNC turning, and the plant also does welding. Its inspection equipment includes a 2D optical (vision) measuring instrument, a height gauge and a hardness tester, with thread go/no-go gauges on turned parts. Every new part gets a first-article inspection report and every lot a dimensional inspection record. If your drawing or purchase order calls for CMM data, say which features need it; at drawing review we will tell you plainly what can be measured with the plant’s equipment and agree how the rest is verified. Send the drawing for a quote.
Frequently asked
Is a portable CMM the same as a laser tracker?
No. Both are portable coordinate measuring systems, but a tracker measures to a reflector over much longer distances and suits very large structures; an arm works within its reach and carries a touch probe or scanner. Each has its own test method.
Can an arm with a laser scanner check GD&T?
It can evaluate profile, flatness and position from scan data, and it is good at showing deviation maps on castings and weldments. The scanner adds error, so tight features are still better probed or gauged.
Does a bigger arm mean less accuracy?
Generally yes. Longer arms have longer links, so the same joint error moves the tip further, and makers publish larger error figures for the longer models of a series. Use the shortest arm that reaches the features.
Should first article inspection always be on a bridge CMM?
No. First article inspection needs every characteristic measured with an adequate method. For a large weldment, an arm plus hand gauges may be the right method; for a small precision part, a bridge CMM or dedicated gauges usually are.
Sources
- CRYSTA-Apex V Series CNC Coordinate Measuring Machines, Bulletin No. 2297(2), Mitutoyo America Corporation, November 2022
- Hexagon Absolute Arm, articulated arm coordinate measuring machine and laser scanner, Institute of Production Engineering and Forming Machines (PTU), TU Darmstadt, accessed 2026
- ASME B89.4.22-2004, Methods for Performance Evaluation of Articulated Arm Coordinate Measuring Machines, ASME, 2004 (catalog listing, BSI Knowledge, reaffirmed 2024)