Can 0.02 mm Scanner Accuracy Reliably Verify a ±0.02 mm Part Tolerance

The short answer is generally no—not from the scanner specification alone. A scanner advertised with 0.02 mm accuracy can be very useful for full-field inspection, process monitoring, deviation analysis, and identifying where a part may be approaching its limits. However, if the drawing tolerance is ±0.02 mm and the scanner’s stated error limit is also 0.02 mm, the nominal relationship is only about 1:1. That leaves no comfortable margin for setup, environment, surface, alignment, data processing, or decision risk.

Reliable verification requires more than matching a brochure number with a drawing. The manufacturer must estimate uncertainty for the actual feature and workflow, define an acceptance rule, and demonstrate capability. A 4:1 ratio suggests capability near ±0.005 mm, but it does not replace customer or quality-system requirements.

This article explains how to make that decision without dismissing the value of a 0.02 mm scanner. In many factories, the right answer is a staged workflow: use industrial 3D scanning for comprehensive surface coverage and rapid diagnosis, then apply a validated higher-accuracy method to critical or near-limit characteristics.
 

First, Decode the Two “0.02 mm” Numbers

The numbers look identical, but they describe different things.

A ±0.02 mm part tolerance normally defines the permitted departure of a characteristic from its nominal value. If a nominal dimension is 50.000 mm, for example, the acceptable interval may be 49.980 to 50.020 mm. The total tolerance width is 0.040 mm, while the distance from nominal to either limit is 0.020 mm.

A 0.02 mm scanner accuracy specification describes a performance characteristic of the measuring system under stated test conditions. Its exact meaning depends on the specification, test artifact, standard, operating mode, calibration status, and reporting convention. It may be an accuracy statement, a maximum permissible error, or another defined performance metric. It is not automatically the expanded uncertainty of every measurement made on a factory part.

The comparison should therefore begin with two questions:

· Does the scanner’s 0.02 mm figure represent a one-sided error magnitude that can be compared with the drawing’s one-sided tolerance limit?

· Is the feature measurement on the real part covered by the same conditions and method used to establish that scanner specification?

If the answer to either question is uncertain, a simple numerical match is not a valid proof of inspection capability.
 

Why a 1:1 Comparison Has No Practical Margin

For a preliminary screen, assume that 0.02 mm is the maximum error magnitude relevant to the measured characteristic. Comparing it with the ±0.02 mm tolerance gives a nominal accuracy-to-tolerance ratio of 1:1. In other words, the possible measurement error is as large as the distance from nominal to the specification limit.

The examples below are intentionally simplified worst-case illustrations. They are not a probability model, and they do not describe the behavior of any specific scanner. They show why an equal-number comparison cannot by itself support a reliable pass/fail decision.

 This is why a scanner that clearly detects broad surface deformation may still be unable to make a defensible acceptance decision on a feature only a few micrometers from the limit. The problem is not whether the scanner produces an attractive color map. The problem is whether the measurement result, its uncertainty, and the decision rule together control the risk of false acceptance and false rejection.
 

Use TAR and TUR as Screening Tools

Two common ratios help teams discuss measurement capability.

Test accuracy ratio (TAR) compares the tolerance with the stated accuracy of the measuring equipment. Test uncertainty ratio (TUR) compares the tolerance with the uncertainty of the complete measurement process. TAR is easier to estimate early because equipment specifications are readily available. TUR is more informative for a real conformity decision because it includes the actual uncertainty budget.

Mitutoyo’s educational guidance distinguishes the two on that basis: TAR uses equipment accuracy, while TUR uses measurement uncertainty. A NIST publication also explains the historical use of the 4:1 TUR and emphasizes that the method used to calculate uncertainty matters.

For a one-sided tolerance limit of 0.020 mm, preliminary targets look like this:

 These ratios are not automatic acceptance criteria. Some customers prescribe a ratio, some organizations use a guard-banded decision rule, and some applications require a formal measurement-system analysis or uncertainty statement. Geometric tolerances also need careful interpretation: a 0.04 mm profile zone, for example, is not necessarily equivalent to an ordinary ±0.02 mm size dimension in how it is evaluated.
 

Scanner Accuracy Is Not the Same as Measurement Uncertainty

A scanner specification is an important input, but the inspection result is created by an entire measurement process. Task-specific uncertainty may include contributions from:

· calibration and verification status;

· ambient temperature, vibration, and lighting;

· part temperature and material expansion;

· surface reflectivity, translucency, texture, and preparation;

· target placement and tracking geometry;

· scan distance, angle, coverage, and operator technique;

· accumulated registration error across multiple scan positions;

· point spacing, filtering, meshing, and hole filling;

· datum construction and alignment strategy;

· feature extraction and fitting method;

· repeatability and reproducibility across operators and shifts.

Feature type matters as well. A freeform surface comparison may be stable while a small bore, sharp edge, thin wall, recessed groove, or datum feature is more sensitive to access and point distribution. Resolution also should not be substituted for accuracy. A scanner may resolve closely spaced surface detail without locating every point to that same numerical accuracy.

The most defensible answer therefore comes from a task-specific study. Measure a representative part or calibrated artifact repeatedly under normal production conditions. Vary the operators, setups, and days that matter. Compare selected characteristics against a reference method with sufficiently better capability. Then estimate the uncertainty or perform the measurement-system analysis required by the organization’s quality plan.

ISO 14253-1 provides rules for proving conformity or nonconformity while taking measurement uncertainty into account. The current 2017 edition was confirmed by ISO in 2023. The key practical lesson is that the measured value alone is not the whole decision, especially near a specification limit.
 

Where a 0.02 mm Scanner Still Adds Major Value

A 0.02 mm scanner can be highly valuable even when it is not the sole acceptance method for a ±0.02 mm critical characteristic. Its advantage is dense, full-field information. A conventional point measurement may confirm several dimensions; a scan can show where the surrounding surface is bowed, twisted, offset, or locally deformed.

Suitable roles include:

· first-pass comparison of the complete part with CAD;

· identifying regions that require higher-accuracy verification;

· monitoring drift before a process produces nonconforming parts;

· comparing cavities, ribs, flanges, trimming lines, and formed surfaces;

· diagnosing fixture, forming, welding, machining, or assembly problems;

· documenting engineering changes and prototype iterations;

· creating repeatable digital inspection records;

· prioritizing scarce CMM or specialist metrology capacity.
 

Guard Bands Turn Uncertainty into a Decision Rule

When results approach a tolerance boundary, the organization needs an agreed decision rule. A guard band narrows the acceptance zone to account for measurement uncertainty.

Under a simple shared-risk approach, a feature with limits at ±0.020 mm and an expanded uncertainty of 0.010 mm might be accepted automatically only when the measured result remains within ±0.010 mm. Results outside the specification are rejected according to the applicable rule, while results between the acceptance boundary and the specification limit may require escalation or remeasurement. The exact rule must be defined by the contract, quality procedure, and relevant standard; this example is not a substitute for them.

Guard bands make the tradeoff visible. A wider guard band reduces false-accept risk but may send more good parts for evaluation. The appropriate rule matches the consequences of each error, the uncertainty, and the customer requirement.

If the uncertainty is equal to the one-sided tolerance, a conservative acceptance interval can collapse. That is another way of seeing why a nominal 1:1 relationship does not support routine, unambiguous acceptance near the limit.
 

How Current SCANOLOGY Systems Fit the Workflow

SCANOLOGY is the overseas subsidiary of SCANTECH (Hangzhou) Co., Ltd. and sells the company’s industrial product line. The following current systems illustrate how equipment can be assigned to different parts of a tolerance-verification workflow. Product selection should still be based on a representative-part test and the exact characteristics to be inspected.

SIMSCAN-S Gen2 for Detailed Full-Field Inspection

The SIMSCAN-S Gen2 is a palm-sized metrology-grade scanner with a stated accuracy of 0.015 mm. The current palm-sized Gen2 family includes SIMSCAN-S Gen2 and SIMSCAN-E Gen2; both use 108 blue ultra-fast laser lines, 17 parallel hyperfine laser lines, and one additional deep-hole laser line. The official specifications list a weight of 560 g and a scan area up to 700 × 600 mm.

 

For compact parts, intricate surfaces, narrow spaces, and localized deformation analysis, the SIMSCAN-S Gen2 can provide dense data and convenient handling. Its 0.015 mm stated accuracy provides a better initial equipment ratio than a 0.020 mm rating. However, 0.015 mm is still not equivalent to a proven task-level uncertainty of 0.005 mm. For a strict ±0.02 mm final decision, the workflow must still be validated for the feature, environment, operator, software settings, and decision rule.

KSCAN-E for Flexible Full-Field and Large-Part Work

The KSCAN-E is a wireless, all-in-one metrology 3D scanner with a stated accuracy of 0.020 mm. Its official specifications distinguish accuracy from volumetric accuracy and resolution; these values should not be interchanged. Integrated photogrammetry and multiple scanning modes support work across different part sizes and feature types.

 

For a ±0.02 mm tolerance, the KSCAN-E is best evaluated as part of a broader inspection plan: capture the full surface, locate deviations, monitor repeatability, and route the most critical or borderline features to an approved confirmation method. Its value is the combination of mobility, coverage, and full-field context—not the assumption that equal nominal numbers guarantee conformity decisions.

AccuArm as a Supplementary Portable CMM Option

AccuArm is a portable coordinate measuring machine, or PCMM. It can complement optical scanning by probing selected geometric features, datums, and locations that benefit from tactile measurement. The official specification table is organized by working range, S/E/C grade, six- or seven-axis configuration, and ISO 10360-12 performance metrics such as SPAT, EUni, PFORM, PSIZE, and LDIA.
 

AccuArm portable coordinate measuring machine for supplementary feature verification

Those metrics must be read in the correct configuration rather than compressed into one generic “accuracy” number. A probing option can strengthen a hybrid workflow, but the chosen arm, reach, probe, setup, and characteristic still need their own capability assessment.
 

DefinSight for Consistent Analysis and Reporting

DefinSight metrology software supports data processing, scan-to-CAD comparison, feature evaluation, deviation visualization, and inspection reporting. For a tolerance-sensitive workflow, software consistency is essential. Teams should control alignment rules, datum definitions, filters, feature-fitting settings, color-map ranges, report templates, and revision history. A repeatable analysis recipe prevents two operators from turning the same point cloud into materially different conclusions.
 

A Practical Hybrid Inspection Workflow

For critical ±0.02 mm features, a staged workflow can combine coverage with decision confidence:

1. Define the characteristic. Record the tolerance type, datum reference frame, acceptance rule, customer requirement, and consequence of a wrong decision.

2. Scan the complete part. Use a suitable current scanner and controlled procedure to capture the surface, establish the overall deviation pattern, and identify suspect areas.

3. Analyze consistently in DefinSight. Apply a locked alignment, documented fitting method, and controlled report template. Avoid changing the color scale to make results look better or worse.

4. Classify results by risk. Clearly conforming regions can support process monitoring. Near-limit characteristics, inaccessible features, and high-consequence dimensions move to confirmation.

5. Verify critical features. Use a validated method with adequate task-specific uncertainty, such as an appropriate higher-capability CMM, specialized gauge, or qualified probing process.

6. Apply the approved decision rule. Record uncertainty, guard band, retest policy, and disposition authority.

7. Feed the full-field data back to production. Use trends and spatial patterns to correct the process before the critical dimension reaches the limit.

This approach avoids treating scanning and tactile metrology as competing choices. Each method handles the part of the problem for which it provides the most useful evidence.
 

Validation Checklist Before Using Scan Data for Acceptance

Before approving a 3D scanning process for any tight tolerance, confirm the following:

· The specification’s accuracy metric and test standard are understood.

· Accuracy, volumetric accuracy, resolution, and repeatability are kept separate.

· The actual part material, finish, size, and temperature are represented in the study.

· The real fixture, target strategy, scanning distance, and alignment method are used.

· Critical features have adequate access and point distribution.

· Repeatability and reproducibility are tested across relevant operators and shifts.

· Results are compared with a traceable reference method.

· The uncertainty budget or required measurement-system study is documented.

· The TAR or TUR convention is stated clearly.

· The guard band and decision rule are approved before production decisions begin.

· Software settings and report templates are controlled.

· Periodic verification, drift checks, and revalidation triggers are defined.

If several of these items are missing, the scanner may still be excellent for engineering analysis, but the organization has not yet demonstrated a reliable acceptance process.
 

Conclusion

A 0.02 mm scanner rating should not be assumed to reliably verify a ±0.02 mm part tolerance. When the numbers are treated as comparable one-sided limits, the nominal ratio is about 1:1, leaving insufficient margin for a robust near-limit decision. A 4:1 planning target would point toward measurement capability around ±0.005 mm, but the final requirement must come from the approved quality and risk framework.

That does not make the 0.02 mm scanner unsuitable for the part. Full-field scanning can reveal deformation, accelerate root-cause analysis, monitor process drift, and guide limited high-accuracy verification to the features that matter most. Current systems such as SIMSCAN-S Gen2 and KSCAN-E can support that coverage, while DefinSight standardizes analysis and AccuArm can add portable probing where appropriate.

The reliable path is to define the characteristic, validate the complete measurement process, estimate uncertainty, establish guard bands, and use a hybrid workflow when the decision margin is tight. The result is faster inspection without confusing a published scanner specification with proof of conformity.
 

Frequently Asked Questions

Does 0.02 mm scanner accuracy mean every point is within ±0.02 mm?

Not necessarily. Read the manufacturer’s definition, test standard, operating mode, and conditions. A published system-performance metric is not automatically a guarantee for every point, feature, surface, and factory setup.

Is scanner resolution the same as accuracy?

No. Resolution describes the system’s ability to distinguish or sample fine detail. Accuracy describes closeness to a reference value under defined conditions. High resolution cannot compensate for inadequate measurement accuracy or an unvalidated workflow.

Can a SIMSCAN-S Gen2 automatically approve a ±0.02 mm feature because its stated accuracy is 0.015 mm?

No automatic conclusion should be drawn from the two specifications alone. The 0.015 mm rating is relevant and useful, but final acceptance depends on task-specific uncertainty, feature accessibility, alignment, environment, repeatability, and the agreed decision rule.

Can a KSCAN-E inspect parts with ±0.02 mm tolerances?

It can capture and analyze such parts, provide full-field deviation information, and help identify critical locations. Whether its data alone can make the final pass/fail decision must be demonstrated for the actual characteristic and process. Near-limit results may require confirmation with a higher-capability validated method.

When should probing be added to a scan workflow?

Probing is useful when selected datums, bores, planes, or geometric features need targeted confirmation; when optical access is limited; or when the approved inspection plan requires tactile measurement. AccuArm can support this type of portable hybrid workflow when the selected configuration meets the application’s requirements.

Is a 4:1 ratio mandatory?

Not universally. It is a widely used planning convention with historical precedent, but contracts, standards, customer requirements, uncertainty policy, and risk may specify another approach. The organization should document why its chosen ratio and decision rule are suitable.