How Should Representative Samples Be Selected for Multiple Configurations?

The practical issue

A commercial product family can contain more than one regulatory type, while one regulatory type can include several variants that do not change the type-defining characteristics. Mirrors, glazing, tyres, lamps and wheels use different type criteria and different test-sensitive parameters, so there is no universal sampling rule across products.

What should be checked?

• Build the full product universe: part numbers, dimensions, options, materials, source and production site.
• Map the applicable “same type” criteria for the correct QCVN.
• Separate variants that cross a type boundary instead of grouping them for commercial convenience.
• Build a test-to-parameter matrix showing which tests are sensitive to dimension, load, material, optical design, thermal condition, structure or installation.
• Identify a candidate worst case for each test. One configuration may not represent every test.
• Confirm sample quantity and preparation requirements under the applicable QCVN.
• Freeze the sample drawing/BOM/software/marking and maintain traceability to series production.
• Record coverage, exclusions and the technical rationale for applying results to other variants.

Regulatory basis

FACT

Each applicable QCVN defines its own type criteria and sample requirements. QCVN 28 requires four mirror samples; QCVN 125 requires three lamp samples; QCVN 34 and QCVN 36 require two tyre samples; and QCVN 78 and QCVN 113 allocate wheel or rim samples by test item. These requirements are product-specific and must not be replaced by a universal sampling rule.

LEGAL REQUIREMENT

Under Circular 55/2024/TT-BGTVT, Article 3 defines representative samples and products of the same type, Article 4 addresses preparation of the representative sample and technical documentation, and Annex I defines same-type criteria. For imported vehicles, Annex I to Circular 54/2024/TT-BGTVT defines vehicles of the same type, while the same-type criteria for components refer to the QCVNs listed in Annex XXI. These instruments must be read together with amendments in force.

Expert interpretation

“Worst case” is a technical conclusion for a specific test or failure mode, not a universal label. A higher wheel load may be adverse for one durability test but says nothing by itself about every installation variant. A different lens, coating or light source may change lamp type even if the outer housing is similar.

A defensible matrix should show which variant is covered by which sample, for which test, against which regulatory criterion, and with what supporting evidence. If that relationship cannot be explained, the proposed type family is not ready to be frozen. A representative sample selected for type testing should also be distinguished from a production sample selected for COP or surveillance; the purpose of sampling changes the selection logic.

EXPERIENCE-BASED INSIGHT

LotusTSE does not assume one “worst case” for all tests. Samples are selected according to the characteristics that affect each test, the expected failure mode and the intended certification scope. A candidate may be conservative for one test and unrepresentative for another. The sampling proposal must show which sample covers which variant, for which test and under what technical conditions. LotusTSE commits to the configuration matrix, selection rationale and gap list, but does not guarantee an external party’s acceptance decision.

Common risks

• Grouping a family by marketing name rather than regulatory type criteria.
• Selecting the “largest” or “highest-load” version for every test.
• Sending a sample that no longer matches the final drawing/BOM.
• Failing to retain rationale, making later extensions or changes difficult to assess.
• Missing sample quantity, fixtures or preparation requirements and losing a test slot.

Recommended approach

1. Build a version-controlled configuration matrix.

2. Flag type-defining and test-sensitive parameters.

3. Propose representative samples test by test with technical rationale.

4. Review the matrix, drawings and approval-mark information in advance with the relevant testing body and certification authority, according to their respective roles, before producing or shipping samples.

5. After testing, freeze traceability and document coverage/exclusions.

Related LotusTSE services

Test planning; representative-sample strategy; vehicle type-approval support; component certification support.

Knowledge Graph

Related QCVN: QCVN 28, 32, 33, 34, 36, 78, 113, 125; QCVN 09/14 where applicable.

Related procedure: Circular 54/2024; Circular 55/2024; Circular 45/2026 and effective amendments.

Related UNECE: R30, R43, R46, R54, R75, R124, R149 depending on the product.

Related Knowledge: P5 Product Change; P10 report scope vs certification scope; P15 family strategy.

Structured-data recommendation: Article + HowTo for the five-step process; do not use Product schema.

NEEDS VERIFICATION

The final representative-sample plan must be confirmed against the exact QCVN, test plan, laboratory capability and intended project scope.