A factory audit for a headlamp supplier is not a tour. It is twelve specific checks on raw materials, moulds, assembly, ageing and final test — and the checks that catch the expensive problems are the ones a visitor rarely looks at.

By Lily, Technical Director · Published 30 October 2026 · 11 min read · Last verified: 2026-10-30

The short answer: run the audit against the published production process, because it tells you where the failures live. Our headlamp product line for wholesale buyers and our factory advantages page describe that process in detail, and the dual power supply Type-C charging sensor headlamp is the worked example of a specification a buyer has to verify on the line.

Points 1–4: Materials, Tooling and Assembly

The first four checks establish whether the factory controls the things that determine whether two units from the same order will behave the same way. They are also the checks most often skipped, because raw material stores and mould rooms are not on the visitor route.

1. Raw material sourcing and supply stability

Because a single-supplier critical component is the most common cause of a mid-season delivery failure in this category, the supplier list is one of the few audit items that is worth copying verbatim into the purchase agreement.

2. Injection moulding is in house

A single-cavity mould produces a consistent part slowly. A multi-cavity mould produces a fast part that varies from cavity to cavity. Because cavity-to-cavity variation is invisible on a sample and visible only across a run, the mould is the only place where that variation can be assessed.

3. Auxiliary part and circuit board assembly

Beyond the plastic case, a headlamp requires circuit boards, cables, bulbs and other parts, and the published assembly step is where workers combine the components according to the design requirements to ensure the suitability and reliability of all parts.

Watch it rather than listen to it. What is worth seeing is whether the operator is checking, and what is being checked. A station with a defined check and a recorded result is different from a station where the operator is fast because the pieces go together one way.

4. The design and development capability, verified

Published product development and design capability is described as a core competency, with a team of senior engineers and designers working from market research and product planning through design and testing, and the factory actively working with customers to provide customized solutions.

The audit question is simple and it is about evidence rather than intent: ask to see the tooling or the electronic design files for one of your own models, and see how long it takes them to produce them. A factory that has not done it before will take days; a factory that has will take minutes, and the difference tells you what your second order will cost.

Points 5–8: Ageing, Final Test and the Product Claims

5. The ageing and performance test is real

Three things to establish. That the equipment exists and is running. That the cycle count and the ambient temperature conditions are recorded rather than assumed. And that a proportion of units from your batch go through it rather than a separate production run.

Because the ageing test is where an assembly fault, a marginal battery cell and a thermal design problem all reveal themselves before they reach a customer, skipping it is the single largest avoidable risk in a headlamp programme.

6. Final inspection is done on real equipment

The published quality approach monitors the whole production process through 20 sets of production equipment, and then before the product is delivered uses 30 testing equipment to conduct a final inspection and inspection of the product.

The number is worth checking against the line rate. A hundred pieces a day passing through thirty test stations is a different operation from a hundred pieces a day passing through three, and the published figures are the kind of detail a factory will show you if the process is there.

Because final inspection is the only gate between a non-conforming unit and your customer, the coverage rate of that gate is what determines the defect rate you will actually experience.

7. The waterproof claim, on the specific model

Published waterproof performance for the range meets IPX4, and the products are described as waterproof and shock-proof so they keep functioning in rain and on rough mountain roads. IPX4 is a splash rating, not a submersion rating, and it means protection against water splashing against the body from any direction.

The check is not whether the factory owns an IPX4 tester. It is whether the model you are buying has been tested, in that configuration, with the lens, the switch and the USB port closed as they will ship.

Because the sealing detail that fails in the field is almost always the port cover or the switch, and because a rating claimed for the range is not a rating for your model, the test report has to name the model.

8. The power source options are actually built

This one catches people because the specification reads well and the failure is invisible on a showroom sample. A dual power supply design combines an internal polymer battery with a 3xAAA fallback, and the published specification is explicit that the AAA cells are excluded from the supplied kit.

Two things to check on the line. That the internal battery is the correct cell for the stated capacity, and that the fallback path actually works, which means the contact design and the switch sequence are right rather than merely present.

Because a dual power supply headlamp is sold on the promise that it keeps working when the rechargeable cell is flat, and because the fallback is the part most often quietly omitted to save a few cents, this is the check that separates a dual power product from a single power product with a battery compartment.

Points 9–12: Claims, Customisation, Non-Conformance and Acceptance

The last four checks are the ones that decide what happens after the audit, and they are the ones a buyer most often does not write down at all.

9. The quoted specifications match the packed product

Weigh it. Measure it. Count the LEDs and count the modes. A published weight that is wrong by 30 g changes the freight calculation, and a published lumen figure that is optimistic changes your marketing.

Because every published number becomes a term in your own business case, a number that does not survive a spot check on the packing bench is a number that should not be in the business case either.

10. The customisation route, with tooling priced separately

Published customisation covers logo, headband customisation including colour, material and pattern, and packaging customisation across colour box, bubble packaging and display box packaging. All three are real capabilities and all three have tooling implications.

Because a logo mould, a headband in a new material and a new print colour are three separate toolings rather than one “customisation” line, and because a quotation that bundles them hides which one is expensive, each should be quoted as its own line.

11. What happens to a non-conforming batch

The published process ends with packaging and delivery, where workers package the product with protective materials and labels and load it into a transport vehicle. The more useful question is not what happens at the end but what happens when something fails in the middle.

Ask what happens to a batch that fails final inspection. Is it quarantined, reworked, re-tested, or scrapped? Is there a written non-conformance procedure? Is the customer notified?

Because a factory with a real non-conformance process will tell you about it without being asked, and because the absence of an answer is itself the answer, this question takes five minutes and tells you more than an hour of walking the floor.

12. The audit ends in written acceptance criteria

This is the point that decides whether the audit was worth the trip. An audit produces a list of observations. What makes it useful is a written set of criteria the factory is measured against, that you can reference on a purchase order and that both sides have agreed before anyone signs.

Because the audit is only worth anything if it ends in criteria the factory is measured against, a checklist without acceptance criteria is a tour with a clipboard and it will not stop the next shipment.

Frequently Asked Questions

Which Approach Should You Take?

Run the audit backwards from a finished unit, spend most of your time in the ageing area, and leave with written acceptance criteria.

Because the forward walk confirms what the factory does well and the backward trace confirms what your order will actually get, the difference between the two is the entire value of an audit.

MENTING was founded in 2014 and specialises in the development and production of USB flashlights, headlamps, camping lights, work lights, bicycle lights and other outdoor lighting equipment. Because a factory that publishes its process and its equipment has already had that process audited by someone, a buyer who asks for the records is only asking to see what a competent customer was already shown.