A 40HQ load of brake rotors reaches the consolidation warehouse at 07:40. By ten o'clock, two pallets sit in the quarantine lane, tagged, photographed, and matched to a measurement sheet that will be in the supplier's inbox before lunch.
Nothing dramatic happened that morning, which is the point. Auto parts incoming inspection is built to be uneventful: a fixed sequence that ends in either a pass stamp or an evidence package.
What Gets Measured, and What Only Gets Looked At
Not every property deserves a gauge. An auto parts incoming inspection routine that measures everything is slow and measures nothing particularly well, so we sort characteristics into three consequence tiers before any instrument comes out of its case.
| Tier | What it covers | Method | Coverage |
|---|---|---|---|
| Critical | Braking, steering, suspension load paths | Calibrated gauge, first article | 100% of first article, then AQL 0.65 |
| Major | Fit, finish, thread, bore, runout, plating | Gauge plus visual | AQL 1.5, level II |
| Minor | Carton print, label position, pallet wrap | Visual audit | Whole lot |
Critical characteristics are measured on every lot with an instrument traceable to a national standard. A brake rotor sitting 0.08 mm outside lateral runout tolerance will produce a pedal pulsation the driver feels within 3,000 km, long after the container has cleared customs.
Major characteristics cover fit and finish: thread pitch, bore diameter, surface roughness, plating thickness, mounting-hole spacing. These are gauged on a sample rather than on every piece, because a lot that has drifted usually shows it in the first twenty parts examined.
Minor characteristics are visual and documentary, covering carton print quality, label placement, pallet wrap tension, and desiccant presence. They seldom stop a shipment on their own, but they generate chargebacks at the destination warehouse, so they are logged every time.
The Four-Stage Sequence
Our auto parts incoming inspection sequence has not changed in four years. Only the sample sizes move, and they move with the supplier's recent history.
- Stage 1 — Identity. Part number, drawing revision, and production date against the purchase order.
- Stage 2 — First article. One piece per cavity per lot on the surface plate, full critical dimension set.
- Stage 3 — AQL sample. Random draw from across the lot, measured and logged against the drawing.
- Stage 4 — Packaging and loading audit. Drop test, stack height, carton compression, container securing.
Stage one is identity, and it is the cheapest stage to run. We confirm that the number on the carton matches the purchase order, the cross-reference list, and the drawing revision. Cartons from two production dates are never mixed on one pallet, even when the part number is genuinely identical.
Stage two is first-article measurement. A single piece per cavity per lot goes onto the surface plate for the full critical dimension set, which takes about forty minutes. If the first article fails, the lot stops there and no sampling time is spent on goods that are already suspect.
Stage three is the AQL sample itself. We work to ISO 2859-1 general inspection level II, normally at an AQL of 0.65 for critical and 1.5 for major defects. Tightened inspection switches on automatically after two consecutive rejected lots from the same supplier.
Stage four is the packaging and loading audit, and it is the stage importers most often skip. A dimensionally perfect brake pad in a carton that collapses at 1.8 m stack height arrives as scrap, and the loading photograph is often the only evidence that survives a six-week voyage.
Where Importers Actually Lose Money
Three failure patterns account for most of the claims we see, and none of them are exotic manufacturing defects.
The first is lot mixing. Two suppliers, or one supplier across two shifts, produce parts that each meet drawing but do not interchange cleanly. Buyers tend to discover this at the vehicle rather than at the dock, which is the most expensive place to find out.
The second is cross-reference error. A part that is dimensionally perfect for the wrong application passes every gauge in the building, which is exactly why the identity check in stage one exists as a separate step.
The third is documentation drift. Material certificates quoting an older heat number, or coatings certified to a superseded specification, will sail through a visual inspection and then fail a customer audit six months later.
The Standards Behind the Tolerances
Tolerances are not invented per order. They come from the drawing, and the drawing normally cites a published standard that both parties can point at when something goes wrong.
| Standard | What it governs |
|---|---|
| ISO 2859-1 | Sampling procedures and AQL tables |
| IATF 16949 | Automotive quality management system requirements |
| ECE R90 | Replacement brake lining and disc approval |
| SAE J2521 | Brake noise and judder measurement |
| ISO 2768-1 | General linear and angular tolerances |
| ASTM B117 | Salt spray testing of coated parts |
Where a drawing is silent, we default to ISO 2768-1 for general tolerances and ASTM B117 for salt-spray duration on coated components. Both are conservative defaults that a European or North American auditor will recognise without a translation note.
What the Report Looks Like When Something Fails
A rejected lot produces four artefacts, and they leave the building together within one working day.
- A measurement sheet listing each dimension, the instrument used, its calibration due date, and the result against limit.
- A photograph set taken with a scale reference in frame, so a size claim cannot be argued later.
- A quarantine tag that physically separates the pallet and records disposition: return, rework, or scrap.
- A corrective action request asking for root cause, tracked to closure before the next order is released.
The measurement sheet is deliberately verbose. It records the value that failed and the two neighbours that passed, because the shape of a deviation tells an engineer more than a single number does.
The corrective action request asks for a root cause rather than an apology. In practice, roughly half of returned requests come back describing a process change that was never validated, which is itself useful information.
Frequently Asked Questions
Do you inspect 100% of parts in a lot?
Only for critical safety characteristics on a first article, or when a supplier is on tightened inspection after two rejected lots. Full 100% gauging of every piece is not economic for fastener and filter families, where AQL sampling gives better information per hour spent.
Who pays for a failed inspection?
The supplier, provided the purchase order states the inspection standard and the AQL levels in advance. Inspection costs become a dispute only when the criteria were never written down, which is why we attach a one-page inspection appendix to every PO.
How long does incoming inspection add to a shipment?
Between one and three working days for a full 40HQ container, depending on the number of SKUs and whether any lot needs rework. A single-SKU lot with complete documentation typically clears in under eight hours.
Can a buyer attend the inspection remotely?
Yes. We run a live video session from the warehouse bench, and the buyer can call the measurement order in real time. Recorded footage is archived with the lot number for twelve months.
Reference Sources
ISO 2859-1:1999 — Sampling procedures for inspection by attributes, Part 1.
IATF 16949:2016 — Automotive quality management system standard.
UNECE Regulation No. 90 — Replacement brake lining assemblies and drum brake linings.
SAE J2521 — Disc and drum brake dynamometer noise and judder procedure.
ISO 2768-1:1989 — General tolerances for linear and angular dimensions.
ASTM B117-19 — Standard practice for operating salt spray apparatus.