Incoming inspection: what to verify at the gate and what to accept on paper

The cost of a miss grows with every operation downstream

Value is added continuously after receipt, and none of it can be recovered when the underlying material turns out to be wrong. A casting rejected on arrival costs a note and a collection. The same casting rejected after machining, heat treatment and finishing has consumed hours of capacity that cannot be resold, and if it reaches assembly it may take good components down with it. That gradient, rather than any general principle, is what should decide where checking effort goes: items feeding long or irreversible routings earn scrutiny that a fastener destined for a subassembly does not.

What a certificate proves, and what it does not

A conformity document tells you the supplier asserts the material meets the specification. A test report with actual figures tells you rather more, particularly if the figures relate to the lot in front of you rather than to a periodic sample. Neither tells you the right material arrived in the right box, which is a packing question and one of the commonest failures in practice. Verification on receipt therefore often has less to do with re-measuring properties than with confirming identity: grade marking, heat number, part number, revision level. That check is quick, needs no laboratory, and catches the mix-ups that certificates by definition cannot.

Checks that only work at the gate

Some characteristics become impossible or absurdly expensive to verify once material has moved on. Bar stock loses its identity in a saw. Adhesives and coatings carry shelf life that starts running before receipt. Electronic components arrive in packaging that must stay sealed to remain valid. Anything with a temperature history, a cure clock or a moisture sensitivity has a window that closes on arrival. These are the items where a receiving check earns its cost even when the supplier is reliable, because the alternative is not a later check but no check at all.

Where suspect deliveries go while somebody decides

Receiving areas are the easiest place in a plant to lose control of material, because everything is transient and space is short. A delivery awaiting a decision must be physically separated and labelled, not merely flagged in a system that the forklift driver does not consult. The failure mode is familiar: a pallet is booked in to make the receipt count work, storage takes it because the rack is empty, and by the time the rejection is confirmed some of it is on the line. Book stock in only after the decision, or book it into a location that cannot be picked.

Rejecting at the gate is a commercial act as well as a technical one

A rejection at receipt stops a delivery, and somebody in production is expecting that material this week. Anyone with authority to reject needs a route to a decision that considers both facts, which usually means a quick conversation between quality, planning and purchasing rather than a form that circulates for days. Record the outcome in the same place regardless of which way it goes, including the occasions where suspect material is used under concession, because those are precisely the events that later explain a run of assembly problems nobody could account for.

Frequently asked questions

Is incoming inspection worth doing if the supplier has a certified quality system?

Certification describes how a supplier manages its processes; it says nothing about the particular pallet on your dock, and identity errors in packing and labelling sit outside what any process control catches. A short identity and damage check on receipt is cheap and finds the mistakes that matter most. Detailed re-measurement of characteristics the supplier already controls is usually poor value, and is better replaced by asking for their results for the actual lot.

How should checking be scaled for a supplier we have used for years?

Let evidence rather than familiarity drive it. A supplier with a long record of clean deliveries earns a lighter regime, but the reduction should be a recorded decision tied to their performance, reversible when something changes at their end. Watch for the triggers that reset the clock: a new site, new tooling, a change of subcontractor, a long gap in ordering, or a change of ownership. Comfortable relationships are exactly where escapes go unnoticed longest.

What do we do when production needs material that has failed the receiving check?

Treat it as a nonconformance with a documented disposition rather than an informal exception. Someone with authority reviews whether the deviation actually affects fitness for the intended use, decides whether the customer needs to be told, and records the reasoning and the batch numbers involved. The material stays identified as concession stock through the build. Skipping that record is what makes a later field problem impossible to bound, because nobody can say which finished units contain it.