Coordinate Measuring Machine (CMM) Guide: Types, Accuracy, ISO 10360 & Selection

A Coordinate Measuring Machine (CMM) is the precision-measurement instrument that sits at the top of every quality-controlled manufacturing operation. Where micrometers, calipers and height gauges measure single dimensions one at a time, a CMM probes a workpiece in three-dimensional space and reports the actual position of every measured point against the designed geometry. Cylinder bores, bolt-circle pitch diameters, perpendicularity of milled faces, profile of complex curved surfaces, total runout on rotating assemblies — a CMM does all of it in one fixture, in minutes, at micrometre-level accuracy.

This guide covers what CMMs are, the five major construction types (bridge, gantry, cantilever, horizontal arm, portable arm) plus the optical and multi-sensor variants, how to read ISO 10360 accuracy specifications, the practitioner-validated reality of CMM software (PC-DMIS, MCOSMOS, Calypso), the environmental conditions a CMM actually needs to hit its specification, and a framework for the capex selection decision. It is written for Australian quality engineers, inspection managers, and manufacturing leaders evaluating a CMM purchase or upgrade — and it does not soften the awkward truths the salespeople will skip over.

What a CMM is — and where it sits in industrial measurement

A Coordinate Measuring Machine is a precision metrology instrument that measures the physical geometry of a workpiece by sampling discrete points on its surface and computing the dimensions, form, orientation and location of features from those points. The machine moves a probe in three orthogonal axes (X, Y, Z) over the workpiece, records the precise spatial coordinates of each probed point, and software constructs the measured geometry — circles, planes, cylinders, spheres, lines — from those coordinate sets. The measured geometry is then compared to the designed geometry on the engineering drawing or CAD model, and tolerance compliance is reported.

The defining capability of a CMM is 3D coordinate measurement against a single reference frame. A height gauge measures one dimension at a time, requires re-fixturing for every face of the workpiece, and stacks operator error on each setup. A CMM measures every feature of the workpiece — top, bottom, side, internal, external, angular — from one fixed datum origin, in one continuous program, with no operator-introduced re-fixturing error between features. For parts with more than a handful of toleranced dimensions, that capability is the difference between a one-day inspection and a fifteen-minute inspection.

CMMs sit at the top of a measurement-instrument hierarchy that runs roughly: caliper (±0.02 mm, 4 minutes per dimension), micrometer (±2 µm, 1 minute per dimension), dial indicator on surface plate (±2 µm, requires fixturing per feature), height gauge / Linear Height (±1.7 µm, single-instrument 2D capability), then CMM (down to ±0.3 µm + L/1000 on top-tier inspection-lab equipment, full 3D capability, one fixture). Each tier costs roughly an order of magnitude more than the previous and unlocks geometry the previous tier could not measure.

In Australia, CMMs are central to first-article inspection (FAI) for aerospace and defence parts, supplier-acceptance inspection on imported tooling and machined assemblies, in-process QC on automotive components, batch-acceptance sampling under AS 9100 / IATF 16949 / ISO 9001 quality systems, and dimensional verification of critical components in mining, rail, energy and medical device manufacturing.

CMM ≠ CMM ASX (stock ticker) — quick disambiguation

If you searched the head term "CMM" and landed here, you are almost certainly looking for one of three different things and only one of them is what this article covers:

• Coordinate Measuring Machine — the precision metrology instrument this guide covers. Read on.
• CMM ASX (Capricorn Metals Limited, ASX ticker code CMM) — an Australian gold-mining company listed on the ASX. If you want share price, market cap or company news, search ASX or your broker, not metrology resources.
• UQ CMM (Centre for Microscopy and Microanalysis at the University of Queensland) — an academic research facility. If you want booking, training or staff contact, see the UQ website.

Three different things, same acronym. The rest of this article is about the metrology instrument only.

How a CMM works — XYZ axes, touch probe and the Abbe principle

A CMM works by moving a precision probe through 3D space along three orthogonal linear axes — X (horizontal, typically left-right), Y (horizontal, typically front-back), and Z (vertical, up-down). Each axis has a precision linear encoder (typically optical, glass-scale) that reports the position of the moving carriage to sub-micron resolution. When the probe contacts the workpiece surface, the spatial position (X, Y, Z) is captured. Software collects these points from features the operator (or program) defines — hole circumferences, plane surfaces, edge lines — and computes the measured geometry from the point cloud.

The probe itself is the contact element. A standard touch-trigger probe (the Renishaw TP20 / TP200 family is the most common reference) closes a switching mechanism when the stylus tip deflects, locking in the X/Y/Z coordinates at the moment of contact. A scanning probe (Renishaw SP25M, SP80, REVO) maintains continuous contact and streams thousands of points per second along a feature surface — essential for form measurement (roundness, cylindricity, flatness, surface profile) that requires dense point sampling.

The encoder system reports machine position; the probe reports contact. The instrument's accuracy depends on both being well-aligned to the same coordinate frame. This is where the Abbe principle matters: measurement error is proportional to the distance between the measurement axis (where the encoder reads) and the measured feature. A CMM design that places the encoders close to the work zone reduces Abbe error; designs with long offset (carriage moving on far rails while probe extends out into the work) accumulate more error per millimetre of probe extension.

The machine column and bridge are typically machined granite — a material chosen for its low thermal expansion coefficient (around 5 × 10⁻⁶ per °C, less than half that of steel), excellent vibration damping, and dimensional stability over decades. Lightweight aluminium frames are cheaper but the practitioner consensus from Practical Machinist machining forums is unambiguous: aluminium is fine for shop-floor entry-level units that don't need to hold sub-10 µm accuracy, but granite is non-negotiable for any inspection-grade CMM intended to keep specification across temperature swings and over years of service.

CMM types: bridge, gantry, cantilever, horizontal arm, portable arm — the decision matrix

There are five major construction types for fixed and portable CMMs, plus optical and multi-sensor hybrid systems. The right type depends on workpiece size, weight, accuracy class needed, and how much shop-floor flexibility matters vs lab-grade accuracy.

Type	Workpiece size	Accuracy class	Best for
Bridge (moving bridge)	400–4,000 mm parts	±1.5–3 µm + L/1000	Most-common workshop QC tier — production engineering, automotive components, machined parts
Gantry (fixed gantry)	3,000–8,000 mm parts	±3–10 µm + L/1000	Large, heavy parts that cannot be lifted onto a bridge table — aerospace, large fab, ship assemblies
Cantilever	Small parts	±2–5 µm + L/1000	Shop-floor environments needing easy load/unload from three sides
Horizontal arm	Long flat panels	±5–10 µm + L/1000	Car body panels, sheet metal assemblies, long welded structures — increasingly replaced by laser scanners
Portable arm	Parts in-situ	±20–80 µm	Onsite measurement where the part can't move — large fab, installed equipment, inspection at customer site
Optical / vision	Small, delicate parts	Sub-micron on small features	Electronics, medical devices, soft or fragile parts where contact damages the surface
Multi-sensor / hybrid	Mixed	Combines best of touch + vision + laser	Production environments needing both speed and accuracy across diverse part families

Bridge CMM deep-dive — the production workshop default

The bridge CMM is the most common construction type in Australian industrial manufacturing and the default starting point for any new capex evaluation. The design features a fixed granite table with the workpiece mounted on it, and a precision-ground bridge that spans across the table and moves along one horizontal axis. A carriage rides along the bridge in the second horizontal axis, and a vertical ram carries the probe down to the workpiece in the third axis.

The advantages of the bridge design are stability, accuracy, and price-to-performance. The bridge structure is closed on all four sides (two bridge legs + the granite base + the cross-beam at the top), which delivers excellent rigidity per kilogram of machine mass. This rigidity translates to the practical reality that bridge CMMs hold their accuracy over years of production use better than less-rigid designs.

The Hexagon GLOBAL series and ALPHA series, Zeiss CONTURA and PRISMO series, and Mitutoyo CRYSTA-Plus, CRYSTA-Apex and LEGEX ranges are all bridge-type. Specifications across this tier:

• Hexagon ALPHA — premium bridge, typically configured around 1200 × 1500 × 1000 mm work volume, ±1.5 µm accuracy class, tactile + scanning + optical sensors, automation-ready
• Hexagon GLOBAL Touch+ — workshop-tier with touchscreen interface, ±2 µm accuracy, PC-DMIS standard
• Zeiss PRISMO — high-precision bridge, ±0.9 µm + L/300 typical, used in metrology labs and aerospace QC
• Zeiss XENOS — top-of-market linear-drive bridge, 0.3 + L/1000 µm accuracy, ±0.2 µm repeatability, 0.001 µm resolution — the reference instrument when sub-micron is non-negotiable
• Zeiss Duramax — entry-tier shop-floor bridge; flagged in Practical Machinist threads as appropriate for shop-floor use but "not a good choice as a daily-driver or only-CMM" for a lab
• Mitutoyo CRYSTA-Plus M544 — cost-effective bridge, 500 × 400 × 400 mm volume, 2.2 + 3L/1000 µm accuracy, general-purpose
• Mitutoyo CRYSTA-Apex V — current production standard, Smart Factory ready with remote monitoring and predictive maintenance
• Mitutoyo MiSTAR 555 — shop-floor bridge with thermal and vibration compensation, 500 × 500 × 500 mm volume, 1.7 + 3L/1000 µm accuracy — optimised for production environments without dedicated climate control
• Mitutoyo LEGEX — high-precision fixed bridge for ultra-tight metrology; sub-micron territory

The MiSTAR 555 deserves a specific call-out for Australian manufacturers. Practical Machinist threads on real-world performance consistently note that the MiSTAR delivers better real-world accuracy in shop conditions than competitors with slightly tighter paper specs, because the thermal compensation and vibration isolation handle the conditions a typical AU workshop actually presents. A machine with ±1.5 µm spec in a controlled lab can perform worse than a 1.7 µm machine with proper environmental compensation when both are placed next to a stamping press at 30°C.

Gantry CMM deep-dive — large heavy parts

A gantry CMM is structurally similar to a bridge but supported by four columns mounted directly to the factory floor (or a dedicated isolated foundation) rather than spanning over a granite table. The workpiece is loaded directly onto the floor plate beneath the gantry, eliminating the need to lift heavy parts onto a measuring table. Measurement volumes typically range from 3,000 × 2,000 × 1,500 mm up to 8,000 × 4,000 × 3,000 mm or larger.

The advantages over bridge design at large scale:

• Load handling — workpieces of several tonnes can be placed directly on the floor plate via crane, with no risk of damaging the CMM measuring table
• Operator access — programmers walk around the part on the factory floor rather than climbing over a bridge structure
• Multi-axis access — large parts with features on five faces can be probed without reorientation
• Safety — no overhead lifting of heavy parts

The trade-offs are accuracy and capex. Gantry CMMs are typically less accurate than the equivalent bridge design (the larger structure has more thermal expansion path-length and more axis-misalignment opportunity), and they require dedicated foundations — typically a concrete pad isolated from surrounding factory vibration, with environmental control. Mitutoyo STRATO-Apex G and Hexagon ALPHA Gantry are the most common AU-market gantry systems, used by aerospace component manufacturers, ship-component fab, rail and heavy-vehicle assemblers.

Foundation requirements matter. A poorly-installed gantry CMM can deliver worse-than-spec performance for years if vibration from nearby machines, traffic, or HVAC reaches the measuring volume. Foundation design is a specialist engineering exercise — not a self-install project.

Cantilever CMM — shop-floor friction-free loading

A cantilever CMM has the measuring probe attached to one end of a horizontal arm that extends from a vertical column. The arm moves in X and Y, the column moves in Z. The defining geometry: three of the four sides of the measuring table are open and accessible. Operators load and unload parts from any of three directions without obstruction.

The advantages are operator throughput and shop-floor integration. For high-volume production environments where a part is measured every few minutes, the cantilever design saves seconds per part on loading and removes the operator-error opportunities created by lifting parts up and over a bridge structure.

The trade-off is rigidity. A cantilever beam is less stiff than a closed bridge of the same mass, so cantilever CMMs are typically built smaller and to less ambitious accuracy specs than bridge CMMs. They are best for smaller parts (typical work envelope 500 × 500 × 400 mm or below) requiring modest precision (±5 µm class). At larger sizes or tighter accuracy classes, bridge construction is the right answer.

Horizontal arm CMM — long flat parts, increasingly replaced

A horizontal arm CMM has the probe carried on a horizontal arm that travels in the X axis along a precision-ground rail, with vertical (Z) and depth (Y) motion at the probe end. The geometry suits long, flat parts where the workpiece is wider than it is tall — car body panels, sheet metal assemblies, automotive door frames, aerospace fuselage panels.

The Mitutoyo CARBstrato is a representative example: a large dual-arm horizontal CMM used in automotive body-in-white inspection, designed for measuring car bodies in two stations simultaneously. The Mitutoyo MACH-3A is a single-arm horizontal CMM optimised for in-line production inspection at high throughput — maximum slider speed 1,212 mm/s, acceleration 11,882 mm/s², on-line cycle-time-friendly.

The honest scope-out: horizontal arm CMMs are increasingly being replaced by non-contact laser-scanning systems and structured-light scanners for the body-panel and large-panel inspection work they historically owned. The reason is throughput — a laser scanner captures the entire panel surface in seconds where a contact probe captures discrete points. The horizontal arm tier is mature but shrinking. For new capex in 2026, the question is usually "horizontal arm or laser scanner?" not "horizontal arm or nothing?"

Portable arm CMM — onsite measurement, 50-100 µm reality

A portable arm CMM is a hand-operated articulated measuring arm with six or seven rotational joints, each instrumented with a precision encoder. The operator moves the probe by hand to touch the workpiece, and software computes the probe position from the joint angles. The FARO Arm, Hexagon Absolute Arm, and Romer Absolute Arm are the dominant brands.

The defining advantage is portability and accessibility. Portable arms can be transported to a part that cannot be moved (large welded structures, installed equipment, components inside production machinery), set up in minutes, and used to inspect features that a fixed CMM could never reach.

The honest accuracy reality: portable arm CMMs typically deliver ±20-80 µm volumetric accuracy at full reach — an order of magnitude less precise than a fixed bridge CMM at the same nominal size. The figure-of-merit is roughly ±50 µm at 2.5 m reach for a workshop-tier articulated arm. For tolerances above 0.1 mm, portable arms are excellent. For sub-50 µm tolerances, they are the wrong instrument and the part needs to come to a fixed CMM.

The KEYENCE XM Series represents an alternative portable-CMM concept: a caliper-form handheld probe with measurement volume up to 2 m, designed to be used by non-specialist operators outside the measurement room. It bridges the gap between traditional articulated arms and bench-top measurement, suitable for shop-floor verification rather than first-article inspection.

Optical and multi-sensor CMM — beyond touch probes

An optical CMM uses vision (camera + lighting + image processing) and/or laser-scanning sensors instead of (or in addition to) a touch-trigger contact probe. Optical CMMs measure without touching the workpiece, which makes them suitable for soft, flexible, fragile, or very small parts that a contact probe would deflect or damage.

Three optical/multi-sensor architectures dominate:

• Pure vision CMM — camera + telecentric lens + LED back-lighting and ring-lighting, measuring 2D features (hole positions, edge profiles, gear-tooth geometry) on small parts. Used heavily in electronics manufacturing and small-component QC. Hexagon Optiv and Mitutoyo Quick Vision Apex are representative.
• Multi-sensor (touch + vision) — a single instrument with both a touch-trigger probe and a vision system, switching between them under program control. Allows inspection of both contact-friendly metal features (with touch) and contact-sensitive features (with vision) on the same part without re-fixturing. Hexagon Optiv Performance is the AU reference for this tier.
• Laser-scanning CMM — a laser-line or laser-stripe scanner attached to a CMM ram (often replacing or alongside a touch probe), capturing thousands of points per second from a non-contact distance. Used for reverse-engineering, complex surface inspection, and large-part scanning where touch sampling is too slow.

Optical and multi-sensor CMMs trade accuracy for capability — they are typically less accurate per-point than a touch-trigger probe on the same CMM body, but they unlock geometry that contact probes physically cannot measure. The choice is driven by what the workpiece needs, not by which spec sheet looks better.

ISO 10360 — the international accuracy standard

ISO 10360 is the international standard for acceptance testing and reverification of CMMs, developed by the International Organization for Standardization. Where a manufacturer's brochure might claim "high accuracy," ISO 10360 specifies exactly what the machine must demonstrate and how. The standard is structured as a series of parts, each covering a specific test methodology:

• ISO 10360-1 — vocabulary and general terms used across the series
• ISO 10360-2 — acceptance and reverification tests for the volumetric length measuring error (E) and the probing error (P). This is the headline accuracy spec most buyers see.
• ISO 10360-3 — CMMs with rotary table as the fourth axis
• ISO 10360-4 — CMMs used in scanning measuring mode (continuous probe contact)
• ISO 10360-5 — CMMs using single and multiple stylus contacting probing systems
• ISO 10360-7 — CMMs with imaging (vision) probing systems
• ISO 10360-8 — CMMs with optical distance sensors
• ISO 10360-9 — CMMs with multiple probing systems
• ISO 10360-12 — articulated arm CMMs

The critical thing about ISO 10360 that separates it from less rigorous standards: 100% of measurement results must fall within the specified maximum permissible error (MPE). Not 95%, not "typical." Every single one of the 105 measurements in the ISO 10360-2 test (5 calibrated length gauges, measured 3 times in each of 7 spatial orientations) must be within the MPE. One outlier fails the test. This is why a CMM with an ISO 10360-compliant accuracy statement is fundamentally more trustworthy than one with a manufacturer's "typical" accuracy claim.

In Australia, ISO 10360 is the de-facto standard. NATA-accredited calibration laboratories use it as the basis for periodic CMM verification, and procurement specifications for aerospace, defence, and high-end manufacturing capex routinely require ISO 10360 acceptance reporting with the new machine.

CMM accuracy specifications decoded — MPE_E = A + L/K

The headline accuracy figure on a CMM specification is typically written as:

MPE_E = A + L/K (µm)

Where:

• MPE_E is the Maximum Permissible Error for length measurement, in micrometres
• A is the fixed (length-independent) error component, in µm — typically the probing and geometry errors
• L is the measured length in millimetres
• K is the length-error denominator (a unitless number — bigger = better)

Worked examples on common formulations:

Specification	At 100 mm	At 500 mm	At 1000 mm	Class
0.3 + L/1000 µm (Zeiss XENOS)	0.4 µm	0.8 µm	1.3 µm	Ultra-premium reference
0.9 + L/350 µm (Zeiss PRISMO)	1.2 µm	2.3 µm	3.8 µm	Aerospace / metrology lab
1.5 + L/333 µm (Hexagon ALPHA)	1.8 µm	3.0 µm	4.5 µm	Production premium
1.7 + 3L/1000 µm (Mitutoyo MiSTAR 555)	2.0 µm	3.2 µm	4.7 µm	Shop-floor with thermal comp
2.0 + L/300 µm (Hexagon GLOBAL Touch+)	2.3 µm	3.7 µm	5.3 µm	Workshop production
2.2 + 3L/1000 µm (Mitutoyo CRYSTA-Plus)	2.5 µm	3.7 µm	5.2 µm	Cost-effective workshop

The reality check the salespeople will not volunteer: these MPE figures only hold under the manufacturer's specified operating conditions. Typically that means 20°C ±1°C ambient temperature, no vibration above ISO 230-5 limits, no airflow disturbance, with the workpiece thermally equilibrated. A machine specified at 1.7 + 3L/1000 µm sitting in a 30°C workshop next to a forging press, with a workpiece that came off the mill ten minutes ago at 45°C, will deliver real-world accuracy that bears little relation to the spec sheet — typically 5-10× worse.

The second reality check from Practical Machinist consensus: do not confuse repeatability with accuracy. A machine that measures the same workpiece five times and gets the same answer to 0.5 µm has 0.5 µm repeatability. That number says nothing about whether the answer is correct. Accuracy is the agreement with the true (NATA-traceable) value. Sales conversations routinely conflate the two; specifications must distinguish them.

The third reality check, from the same source: single-axis numbers are meaningless unless the part is being measured along that one axis only. The volumetric length error (E) integrates all axis errors plus the probing error plus geometry errors into one figure. Asking the vendor for the ISO 10360-2 ball-bar test report is the way to verify the spec sheet against measured performance.

Probes, styli and probing systems — the indexing head is the single highest-ROI accessory

The probe is the contact element of the CMM and the single biggest determinant of what the instrument can actually measure. Three categories cover most of the AU market:

• Touch-trigger probes — switching probes that close a contact when the stylus tip deflects (Renishaw TP20, TP200 series). Lower cost, simple, robust. Best for measuring discrete points — hole positions, edge locations, plane heights. Most common probe on workshop and production CMMs.
• Scanning probes — continuous-contact probes that maintain stylus pressure on the workpiece while moving along the surface (Renishaw SP25M, SP80, REVO). Capture thousands of points per second along a feature, essential for form measurement — roundness, cylindricity, flatness, surface profile. Significantly more expensive than touch-trigger.
• Indexing probe heads — articulating heads that rotate the probe to one of several pre-calibrated orientations (Renishaw PH10M / PH10MQ / PH20). Allow the same probe to measure features on different faces of the workpiece without re-fixturing.

The single most-cited piece of buying advice from practitioner forums (Practical Machinist, multiple threads): pay for the indexing head. The line from a working metrologist is direct: "Pay for the indexing head. It's worth it. A CMM is an investment in reducing the cost of good quality, time spent trying to program around limited probes is an increase in that cost." The Renishaw PH10M-class indexing head adds substantial cost to a CMM purchase but pays for itself in programmer time saved over the first year of production use. A REVO five-axis measurement head is overspec for most workshop applications, but PH10 / PH10M is the standard recommendation for production CMMs.

Stylus selection is the other key decision. The basic dimensions:

• Stylus length — longer reaches deeper into bores and around obstructions but bends more under contact force, reducing accuracy. Keep stylus length minimum-needed for the application.
• Stylus tip diameter — typically 1, 2, 3, 5, 6, 8 mm ruby ball. Larger tip averages out surface roughness (good for plane measurement on rough castings) but cannot enter small features. Smaller tip resolves small features but is more affected by surface texture.
• Stylus material — ruby (industry standard, hard, low-wear, low-friction), zirconia (for ferrous workpieces — ruby can wear when scanning steel surfaces continuously), silicon nitride, or tungsten carbide for specific applications.
• Star probes and multi-tip clusters — multiple stylus tips on one head, allowing measurement from multiple directions without head re-indexing. Good for simple side-hole and bore measurements; the practitioner advice from Practical Machinist is that "engine blocks need fancy stuff" — at the more complex end, an indexing head with single-tip configurations beats a star probe.

Practitioner warning on probe management: articulating (indexing or motorised) probes add calibration time and small additional measurement error. Each indexed orientation needs to be calibrated against a reference sphere, and the calibration is only valid until the next probe collision or re-mounting. Build the calibration time into program development estimates.

CMM software — PC-DMIS / MCOSMOS / Calypso reality

CMM software is a bigger decision than CMM hardware for most buyers, and one where the marketing language and the practitioner reality diverge sharply. Three platforms dominate the AU market:

• PC-DMIS (Hexagon) — the world's most-installed CMM software platform. Used on Hexagon machines and many third-party CMMs. Programmer-pool advantage: more PC-DMIS-trained operators in the AU labour market than any other platform, and more training providers (TAFE-level metrology courses default to PC-DMIS). The practitioner reality, almost universally stated: PC-DMIS is not user-friendly. The line from a working machinist: "PC-DMIS is a lot of things, user friendly isn't one of them. However, you'll have the best luck finding programmers and training." Buyers choose it because the hiring market makes the choice for them, not because the software is preferred.
• MCOSMOS (Mitutoyo) — Mitutoyo's modular CMM platform with separate modules for standard measurement, scanning, CAD comparison, and SPC integration. The defining practical advantage cited by Mitutoyo users: no annual support charges after purchase. PC-DMIS effectively requires annual licence renewal to retain Hexagon support and update access; MCOSMOS does not. Over a 10-year CMM lifecycle, the cost difference compounds into meaningful capex+opex differential. Mitutoyo MCOSMOS is well-regarded by Mitutoyo CMM users but less broadly skilled in the AU labour market.
• Calypso (Zeiss) — Zeiss's CMM platform, used on Zeiss machines. Strong CAD-integration capability, sound technical foundation. Mid-market positioning between PC-DMIS dominance and Mitutoyo's free-support model. Zeiss support and software has improved in recent years but practitioner consensus is that Zeiss running/service costs remain higher than competitors over the machine lifetime.
• Modus (Renishaw) — used primarily on Renishaw Equator gauging systems and some smaller CMM installations. Well-regarded by users but not common; programmer pool small.
• CMM Manager (Nikon Metrology) — modern platform, well-regarded in practitioner reports, but less common than the big three.

The TCO trap most buyers don't see at quote time: annual software maintenance contracts can run 10-15% of the original software capex per year. Over a 10-year CMM lifecycle, software maintenance can equal or exceed the original software capex. On a CMM where software is bundled with hardware, this cost is often invisible at purchase and only becomes apparent two years in when the first renewal arrives. Ask the vendor for the 10-year software-maintenance projection in writing as part of the procurement quote.

The second TCO trap: some manufacturers will not provide technical support without an active software licence. If the machine breaks down and the software maintenance has lapsed, the manufacturer may quote both software-renewal back-pay and a service call to get the machine running again. This is most flagged for Hexagon (PC-DMIS) but applies in varying degrees to all platforms.

Environment — temperature, vibration and the "ghost error" reality

Environment is where most CMMs underperform their spec sheets. The published accuracy figures assume 20°C ±1°C ambient temperature, controlled humidity (40-60% relative), vibration below ISO 230-5 limits, no airflow over the work zone, and thermally-stabilised workpieces. Most AU industrial environments do not meet any of those conditions, and the gap between spec-sheet performance and shop-floor reality is the source of most "the CMM gives different answers every time" frustration.

The four environmental factors that dominate "ghost error" — measurement drift that appears progressively through a shift without a clear cause:

• Temperature gradient and rate of change — what kills CMM accuracy is not absolute temperature but changes in temperature, particularly gradient (top of granite warmer than bottom, or one side warmer than the other). Practitioner guidance: monitor temperature at the granite level, not at the wall thermostat. Maintain rate-of-change below 1°C per hour. A granite plate that's 25°C uniform-top-to-bottom can deliver near-spec accuracy; the same plate at 22°C with a 3°C gradient will not.
• Workpiece thermal equilibration — a part that came off the machine tool at 40°C and is measured 5 minutes later at 25°C ambient is contracting at micrometres per minute. Leave parts to equilibrate in the measurement room — minimum 30 minutes for small parts, hours for large heavy parts.
• Vibration — stamping presses, forklifts, HVAC compressors, traffic outside the building — all contribute to floor vibration that the CMM cannot fully isolate. A CMM with passive isolation typically rejects vibrations above 10 Hz; below 10 Hz, structural building modes can transmit straight through. Active isolation systems exist for sub-micron metrology labs and add to the capex.
• Airflow and dust — direct airflow over a probe stylus from a ceiling vent can introduce contact-force variation. Dust and swarf on the granite or stylus tip introduces single-point spike errors that look exactly like a random measurement glitch.

The practical implication for AU buyers: if the CMM is going into a workshop without dedicated environmental control, choose a machine with full thermal compensation and active vibration isolation (the Mitutoyo MiSTAR 555 is specifically engineered for this scenario), or build a dedicated metrology room around the machine. The capex difference between a machine + environmental room and a bare machine in the workshop is much smaller than the multi-year accuracy difference between them.

The other environmental practice that pays for itself: daily verification with a NATA-traceable artifact. A reference ball or gauge block kept in the metrology room and measured at the start of each shift catches drift before it produces bad data. Five minutes per shift, calibration-traceable verification.

Capex selection framework — accuracy, throughput, training, calibration

A CMM is a capital equipment decision with a typical 10-15 year useful life and total cost of ownership 2-3× the original capex by year 10 (calibration, software maintenance, training, environmental control, probe replacement). The selection framework below is the order of priority that experienced AU metrology managers apply:

1. Define the tightest tolerance you actually need to measure — and then double the accuracy budget. A workpiece with ±0.025 mm tolerance needs a CMM measuring at ±5 µm or better (rule of thumb: instrument accuracy = tolerance ÷ 5). Stricter rules of thumb (÷10) push you into more expensive tiers. Most buyers under-spec accuracy at purchase and over-pay for upgrades later.
2. Define the largest part you'll measure in the next 10 years — and then add 30%. The practitioner consensus: under-sized work envelopes are the most-regretted CMM purchase decision. A larger machine can measure smaller parts; a smaller machine cannot measure larger parts.
3. Decide CNC or manual — the practitioner rule: don't buy a CNC system if you only measure each part once. CNC programming time outweighs inspection time on one-off work. CNC pays back fast on repeat production, FAI workflows, and overnight unattended measurement. Most production CMMs are CNC; metrology-lab one-off work is more often manual or hybrid.
4. Probe configuration is not optional — budget for an indexing head (PH10M class minimum) from day one. The cost of trying to program around limited probes outweighs the head cost within months.
5. Software choice tied to hiring market — if your area's metrology hiring pool is primarily PC-DMIS trained, choose Hexagon. If you are running a small operation and value the no-support-charge MCOSMOS model, Mitutoyo. If your CAD workflow is Zeiss-aligned, Calypso. The software outlives the hardware in most cases.
6. Environment first, machine second — a ±2 µm machine in a controlled environment outperforms a ±0.5 µm machine in a hot workshop. Get the environment right or buy a machine that compensates for it.
7. 10-year TCO, not capex — software maintenance, calibration, training, probe replacement, environmental control. Software maintenance alone can equal capex over a decade.
8. In-house vs outsource decision — for low-volume inspection work (less than 5-10 hours per week of CMM time), outsourcing to a NATA-accredited calibration and inspection service is usually more cost-effective than in-house ownership. The break-even depends on hourly utilisation; calculate honestly.
9. NATA traceability — for AU customers requiring NATA-accredited inspection certificates (aerospace, defence, regulated industries), the CMM purchase must include NATA-accredited calibration certification at delivery and annually thereafter. Confirm the supplier's calibration chain.
10. Training capacity — the machine is only as good as the operator. Budget for training as part of capex, not an afterthought. Mitutoyo Australia and Hexagon Manufacturing Intelligence both run AU training programs; factor 2-4 weeks of operator training per new CMM into the project plan.

AIMS Industrial — for AU buyers evaluating a CMM purchase

AIMS Industrial is not the direct AU distributor of new CMM hardware. The Mitutoyo Australia authorised CMM dealer is Automated Solutions Australia (ASA), who handle the CRYSTA-Apex, MiSTAR, STRATO and LEGEX ranges plus on-site demonstration, application engineering and service. Hexagon Manufacturing Intelligence Australia handles the Hexagon range, Zeiss Industrial Quality Solutions handles Zeiss. For new CMM hardware capex, those are the direct channels.

Where AIMS Industrial fits the CMM-buyer journey:

• Adjacent inspection-lab kit — gauge blocks (Mitutoyo Series 516), surface plates (Mitutoyo Series 517), dial indicators, test indicators, micrometers, calipers, height gauges, pin gauges, surface roughness comparators — all the everyday metrology hand-tools that surround a CMM and that we hold in stock or bring in through the Mitutoyo Australia supply chain
• Mitutoyo Australia referrals — for CMM hardware enquiries we can connect you to ASA (Automated Solutions Australia) or Mitutoyo Australia direct, depending on the path that fits your buying process
• Calibration support — Mitutoyo Australia operates NATA-accredited calibration laboratories for CMMs, gauge blocks and reference standards; we can route enquiries to the right Mitutoyo calibration service contact
• Application advice — our team has practical metrology background and can sense-check capex specifications, accuracy class requirements, and TCO considerations before you commit to a quote

If you're evaluating a CMM purchase and want to talk through the decision before going direct to the hardware vendors, call our team on (02) 9773 0122 or contact us via aimsindustrial.com.au/contact-us. We'll route the conversation to the right Mitutoyo Australia channel and help you spec the adjacent metrology kit you'll need around the CMM itself.

---

Frequently Asked Questions

What is a CMM (coordinate measuring machine)?

A Coordinate Measuring Machine is a precision metrology instrument that measures the physical geometry of a workpiece by probing discrete points on its surface in 3D space and computing the dimensions, form, orientation and location of features from those points. CMMs measure features against a single reference frame in one fixture — top, bottom, side, internal, external — eliminating the operator-introduced re-fixturing error of single-dimension instruments. They are the standard tool for first-article inspection (FAI), supplier acceptance, in-process QC, and any dimensional verification requiring micrometre-level accuracy across multiple features on the same part.

What does CMM stand for?

CMM stands for Coordinate Measuring Machine. The acronym is also used by Capricorn Metals Limited (ASX ticker code CMM), an Australian gold-mining company, and by the University of Queensland Centre for Microscopy and Microanalysis. The three are completely unrelated; this guide covers the metrology instrument only.

What types of CMM are there?

The five main construction types are bridge (moving bridge, the most-common workshop type), gantry (fixed gantry over a floor-mounted workpiece, for large heavy parts), cantilever (three-side-open access for shop-floor loading), horizontal arm (long flat parts like car body panels), and portable arm (articulated hand-held arm for onsite measurement). Plus optical/vision CMMs for small, soft, or fragile parts, and multi-sensor hybrid CMMs combining touch + vision + laser scanning. The choice is driven by workpiece size, accuracy class needed, and whether the part can come to the machine or the machine must go to the part.

What's the difference between a bridge CMM and a gantry CMM?

Bridge CMMs span a precision-ground granite table with a bridge structure that moves over the workpiece — best for parts up to about 4 metres that can be lifted onto the table, with accuracy typically in the ±1.5 to ±3 µm class. Gantry CMMs are supported by four columns mounted directly to the factory floor, with the workpiece loaded onto a floor plate beneath — best for parts 3 to 8 metres or larger that are too heavy to lift onto a bridge table, with accuracy typically ±3 to ±10 µm. Bridge designs are more accurate and more compact; gantry designs handle bigger and heavier parts at lower accuracy. Gantry CMMs also require dedicated isolated foundations.

What is ISO 10360?

ISO 10360 is the international standard for CMM acceptance testing and reverification, developed by the International Organization for Standardization. It specifies exactly what tests a CMM must pass to demonstrate its accuracy claim. The most-cited part of the series is ISO 10360-2, which defines the volumetric length measuring error (E) test using calibrated length gauges measured in multiple spatial orientations. The critical feature: 100% of measurement results must fall within the specified Maximum Permissible Error (MPE) — not 95%, not "typical." One outlier fails the test. ISO 10360 is the de-facto AU standard, used by NATA-accredited calibration labs and required by most aerospace, defence and high-end manufacturing procurement specifications.

What does MPE_E mean on a CMM specification?

MPE_E stands for Maximum Permissible Error for length measurement, defined under ISO 10360-2 in micrometres. The formula is typically written as MPE_E = A + L/K where A is the fixed length-independent error component (µm), L is the measured length (mm), and K is the length-error denominator (unitless, bigger is better). For example, a Hexagon ALPHA at 1.5 + L/333 µm delivers 1.8 µm at 100 mm, 3.0 µm at 500 mm, and 4.5 µm at 1000 mm. The spec only holds under specified operating conditions — typically 20°C ±1°C ambient, no vibration, thermally-equilibrated workpiece. Real-world workshop accuracy is often 5-10× worse when those conditions are not met.

How accurate is a CMM, really?

Accuracy varies dramatically by class. Top-of-market metrology-lab CMMs (Zeiss XENOS) deliver 0.3 + L/1000 µm — around ±1.3 µm at 1 metre measurement length. Premium production CMMs (Hexagon ALPHA, Zeiss PRISMO) sit in the ±1.5 to ±2 µm range. Workshop-tier production CMMs (Mitutoyo CRYSTA-Apex, Hexagon GLOBAL Touch+) are typically ±2 to ±3 µm. Shop-floor-rated CMMs with thermal compensation (Mitutoyo MiSTAR 555) deliver 1.7 + 3L/1000 µm in environments where lab-class machines would not hit spec. Portable arm CMMs (FARO, Hexagon Absolute Arm) are ±20-80 µm depending on reach. Critically: accuracy is not the same as repeatability. A machine can be repeatable to 0.5 µm but inaccurate by 10 µm — repeatability means same answer every time, accuracy means correct answer. The relevant figure on a CMM is volumetric accuracy under ISO 10360-2, not single-axis or repeatability spec.

What's the difference between a touch-trigger probe and a scanning probe?

A touch-trigger probe (Renishaw TP20, TP200) closes an electrical switch when the stylus tip deflects, locking in a single X/Y/Z coordinate at the moment of contact. The probe touches the workpiece, records one point, lifts off, and moves to the next point. Touch-trigger probes are lower cost, simple, and robust — the standard for production CMMs measuring discrete features (hole positions, edge locations, plane heights). A scanning probe (Renishaw SP25M, SP80, REVO) maintains continuous stylus contact with the workpiece while moving along the surface, streaming thousands of points per second. Scanning is essential for form measurement — roundness, cylindricity, flatness, surface profile — that requires dense point sampling. Scanning probes cost significantly more than touch-trigger; the buying decision is whether the workpieces have form features that justify the upgrade.

Do I need a CNC CMM or is a manual CMM enough?

The practitioner rule is direct: don't buy a CNC system if you only measure each part once. CNC programming time outweighs inspection time on one-off measurement work — for a manufacturer doing first-article inspection on prototypes and tooling, a manual CMM is often more productive. For repeat production work where the same part is measured tens or hundreds of times against a stored program, CNC pays back fast. CNC also enables overnight unattended measurement (load a batch at end of shift, collect data the next morning) which is a major throughput advantage in production environments. Most workshop CMMs are CNC; one-off metrology-lab work skews more toward manual or hybrid. If your operation runs both repeat-production and one-off work, hybrid CMMs (motorised positioning with manual programming) are a common compromise.

What's PC-DMIS and why does everyone use it even though they complain about it?

PC-DMIS is Hexagon's CMM software platform, the world's most-installed CMM programming and operating environment. The reason it dominates despite practitioner complaints is the labour-market network effect: more PC-DMIS-trained metrologists are available for hire than any other platform, more TAFE-level metrology courses default to PC-DMIS, and more third-party CMM software ecosystems integrate with it. A working machinist's summary that captures the consensus: "PC-DMIS is a lot of things, user friendly isn't one of them. However, you'll have the best luck finding programmers and training." Buyers choose it because the hiring market makes the choice. Alternative platforms — Mitutoyo MCOSMOS, Zeiss Calypso, Renishaw Modus, Nikon CMM Manager — are technically capable but trained operator pools are smaller. The TCO trap with PC-DMIS to watch: annual software maintenance is often required for support and updates, and the cumulative cost over a 10-year machine life can equal the original software capex.

How much does a CMM cost?

CMM capex spans a wide range. Entry-tier workshop bridge CMMs (Mitutoyo CRYSTA-Plus class, Hexagon GLOBAL S class, Zeiss Duramax) sit in the lower-five-figure to low-six-figure bracket. Mid-tier production CMMs with indexing heads, scanning probes and CNC (Mitutoyo CRYSTA-Apex V, Hexagon GLOBAL Advantage, Zeiss CONTURA) move into the mid-six-figure range. Premium metrology-lab CMMs (Zeiss PRISMO, Hexagon LEITZ Reference) are upper-six-figure to seven-figure. Large gantry CMMs and dual-arm horizontal-arm systems sit at the top of the range. Beyond the headline capex, budget for software maintenance (annual contracts at 10-15% of software capex), NATA-accredited calibration (annually), training (2-4 weeks per new operator), and environmental control (dedicated metrology room with HVAC). Total cost of ownership over 10 years is typically 2-3× the original capex.

What is a portable arm CMM and when does it make sense?

A portable arm CMM is a hand-operated articulated measuring arm with six or seven rotational joints, each instrumented with a precision encoder. The operator moves the probe by hand to touch the workpiece, and software computes the probe position from joint angles. FARO Arm, Hexagon Absolute Arm and Romer Absolute Arm dominate the AU market. The defining use case is onsite measurement of parts that cannot be moved to a fixed CMM — large welded structures, installed equipment, components inside production machinery, customer-site inspection. Accuracy is typically ±20-80 µm at full reach (roughly ±50 µm at 2.5 m reach for a workshop-tier articulated arm) — an order of magnitude less precise than a fixed bridge CMM at the same nominal size. For tolerances above 0.1 mm, portable arms are excellent. For sub-50 µm tolerances, the part needs to come to a fixed CMM. They are a complement to a fixed CMM, not a replacement.

What is multi-sensor or hybrid CMM?

A multi-sensor or hybrid CMM combines two or more measurement technologies in one instrument — typically a touch-trigger probe plus a vision system plus optionally a laser scanner, all sharing the same CMM body and coordinate frame. The advantage is being able to measure features that benefit from contact (precise metal surfaces, hole edges) using the touch probe, and features that benefit from non-contact (soft, fragile, very small, or optically clearer than mechanically clear) using vision or laser — all on the same part without re-fixturing. Hexagon Optiv Performance is the AU reference for this tier, with similar offerings from Mitutoyo and Zeiss. Multi-sensor CMMs are increasingly common in electronics manufacturing, medical device inspection, and any production environment with diverse part families. The trade-off is capex — multi-sensor systems cost more than single-sensor equivalents, and the complexity adds calibration time and operator training burden.

Can I outsource CMM inspection instead of buying one?

For low-volume CMM work (less than 5-10 hours per week of measurement time), outsourcing to a NATA-accredited calibration and inspection service is often more cost-effective than in-house ownership. AU-based contract metrology services include Mitutoyo Australia's own inspection bureau plus numerous independent NATA-accredited labs. The break-even calculation: weekly hours required × hourly inspection rate vs annualised capex + software maintenance + calibration + training + environmental control. Outsourcing is also valuable when you need NATA-traceable inspection certificates for export documentation, AS 9100 / IATF 16949 audit evidence, or supplier acceptance reporting and don't want to maintain in-house NATA accreditation. The other use case: in-house CMM for daily production work + outsourced for one-off high-accuracy first-article inspections that exceed in-house machine spec.

How often does a CMM need calibration and what does NATA-accredited mean?

The standard CMM calibration interval is annual for production use, with daily verification against an in-house traceable reference artifact (reference sphere or gauge block kept in the metrology room). Critical applications — aerospace, defence, pharma, regulated medical devices — may require 6-monthly calibration. NATA accreditation is the Australian framework administered by the National Association of Testing Authorities; a NATA-accredited calibration certificate documents traceability of the CMM's reported measurements to the international system of units (SI) via a documented unbroken calibration chain. For Australian aerospace, defence, regulated and export-oriented manufacturing, NATA accreditation is the de-facto requirement. Mitutoyo Australia operates NATA-accredited calibration labs for CMMs, gauge blocks and reference standards; Hexagon and Zeiss have their own AU calibration services; independent NATA-accredited labs also offer CMM calibration. Annual calibration cost depends on machine size and accuracy class — confirm with your calibration provider during procurement.

Pair this with our GD&T Symbols Guide for the AS/NZS 1100 and ASME Y14.5 symbol reference.