A coordinate measuring machine purchase for incoming-inspection cells is governed less by brand and more by five quantitative gates: ISO 10360 accuracy class, temperature envelope of 20±2°C with 40-60% humidity, axis travel versus the worst-case part envelope, probe strategy (trigger vs. scanning vs. multi-sensor), and bridge/gantry/portable/horizontal-arm architecture matched to workflow [S1].
Modern CMMs trace to DEA and Ferranti units in the 1960s, migrated from manual hard-probe tables to CNC trigger-probe systems, and now include portable arms that can also pass ISO 10360 acceptance [S1]. For an incoming-inspection buyer the practical question is which of the five architectures maps to the dominant part type.
The five CMM architectures and where each wins
Bridge CMMs remain the most common in mold, machining, and stamping cells because the moving bridge gives high stiffness per dollar and supports scanning digitisation in production volumes [S1]. Cantilever units suit smaller measuring tools and short main parts where the open side allows operator loading, while gantry CMMs cover large molds and heavy castings that exceed bridge travel, and horizontal-arm machines are the standard pick for high-throughput aerospace, defence, and appliance batches [S1].
Portable CMMs (PCMM) carry 3D and GD&T measurement to the part, support CAD comparison and automatic inspection reports, and are increasingly specified where shop-floor conditions rule out a fixed lab cell [S2][S4]. For incoming inspection, the rule of thumb is: if the part is heavy and rarely moves, gantry; if it is small and high-mix, bridge or cantilever; if it is large and low-volume, horizontal-arm; if it must come to the receiving dock, portable.
Accuracy, environment, and the ISO 10360 envelope
Every global CMM must comply with ISO 10360, which is the gating acceptance standard for linear accuracy, probe performance, and length-measurement error [S1]. The standard environment for full-spec acceptance is 20±2°C, 40-60% relative humidity, and a vibration-isolated foundation; outside that envelope, even a high-accuracy machine will be downgraded by thermal expansion on the part, the fixture, and the machine structure itself [S1].
Probe calibration is the first measurement step and uses a reference sphere with at minimum a five-point method (one pole point plus four equatorial points), with the probe head, stylus, and sphere rigidly fixed and the sphere surface cleaned [S1]. Skipping this step, or using a dirty standard ball, injects systematic error that propagates into every downstream dimension and is the most common root cause of false rejects in incoming inspection.
Fixed-shop, shop-floor, and ultra-accuracy tiers
Reps and OEM line cards in 2026 segment CMMs into three operational tiers: ultra-accuracy lab machines for sub-micron work, shop-hardened fixed CMMs for cells next to machining, and portable arms for at-line measurement [S2]. The ultra-accuracy tier typically couples a thermally stabilised granite bridge with scanning probes and rotary tables, and is appropriate when the incoming tolerance window is single-digit microns on tight-tolerance aerospace or medical components [S2].
Shop-hardened CMMs trade peak accuracy for thermal robustness and rigidity so the machine can sit near the production cell; portable arms trade absolute accuracy for accessibility and are most often used for first-article checks, fixture verification, and large-parts inspection on the shop floor [S2][S4]. When selecting, match the tier to the smallest tolerance band on the inspection drawing, not to the nominal machine spec sheet.
Probe strategy: trigger, scanning, and multi-sensor
For incoming inspection of prismatic machined parts, a fixed trigger probe (touch probe) is the default; for freeform surfaces such as turbine blades, plastic trim, or stamped panels, a scanning probe collects dense point clouds that the software fits to CAD [S2]. Multi-sensor CMMs (optical, laser, touch, and scanning) consolidate these workflows into one platform and are attractive when the cell must cover a mix of connector housings, turned shafts, and sheet-metal brackets [S2].
Structured-light scanners added to a CMM line give fast, non-contact measurement of soft or complex features where a touch probe would deform the part or take too many points [S2]. Selecting a probe strategy without first listing the dominant feature types in the incoming queue is the single most common spec error, because a touch-probe-only CMM on a blade family will bottleneck the cell.
Selection criteria mapped to incoming-inspection needs
The four decision criteria a quality manager should score before issuing a PO are: (1) ISO 10360 length-measurement error versus the tightest part tolerance, (2) axis travel versus the largest part envelope plus fixture, (3) probe type versus the dominant feature family, and (4) environmental robustness versus the actual cell conditions [S1][S2]. On those criteria, a bridge CMM with a scanning probe scores high for mold and mid-volume machining; a gantry scores high for large castings; a portable arm scores high for dock-side and on-machine checks; a horizontal-arm scores high for aerospace and appliance batches [S1].
For comparison, a buyer evaluating the four against a typical incoming tolerance of ±0.05 mm on a 400 mm prismatic part: bridge and gantry CMMs comfortably pass ISO 10360 and have full scanning options; horizontal-arm machines cover the workflow but require part-specific fixturing; portable arms meet the tolerance but need skilled operators and a controlled local environment to keep the result auditable [S1][S4]. Multi-sensor systems add vision and laser to the touch and scanning channel and are the best fit when incoming mix changes weekly [S2].
Acceptance, calibration, and when to escalate
Acceptance test and periodic re-verification of a CMM are performed against reference objects (standard sphere, ring gauge, step gauge, ball plate), with the procedure consolidated in the international framework for calibration, acceptance, and periodic inspection of CMMs [S5]. A CMM that fails the volumetric length test, or that shows probe-qual repeatability drift above the ISO 10360 budget, should be taken offline and re-qualified before the next lot is released, because the alternative is shipping parts that look accepted on screen but are out of tolerance in the customer fixture [S1][S5].
Portable CMMs must be calibrated in the same posture and with the same shank length used in production, otherwise the kinematic model is invalid for that setup, and this is the most frequent field failure mode when portable arms enter a metrology room for the first time [S4]. If a part is beyond machine travel, do not attempt to re-fixture outside the verified envelope: re-fixture outside the calibrated volume is equivalent to running the part on an uncalibrated machine, and the lot should be sent to a larger CMM or split into multiple setups with documented join error.
Trackable signals to watch: ISO 10360 re-test results on a fixed quarterly cadence, probe-qual drift logged per shift, and a written rule that any fixture change is followed by a verification run on a calibrated master before the next production lot is signed off [S1][S5]. For buyers comparing options side by side, the Cutter Machine and Coding Machine reference pages on this site set out the equivalent selection-discipline workflow for adjacent cells, and the related Sorting System Selection for Automotive Parts Logistics guide covers the downstream material-flow side of the same incoming-inspection loop.
Component reference pages worth checking: contour measuring machine.