Production quality and batch verification
The prototype passed. That was two years ago.
Almost every acoustic performance claim in the market rests on a single test of a single specimen, often years old. Since then the supplier of the core material has changed, the density has been value-engineered, the adhesive is different and the line runs faster. None of those changes were made to hurt acoustic performance and each one is individually defensible — but nobody has measured what they add up to.
That is the gap batch verification closes. Not a new type test, but periodic confirmation that what leaves the line still behaves like the thing that was certified. It is cheap relative to a type test and it is very cheap relative to discovering the drift through a customer complaint on a completed building.
How to set it up so it is worth having
- Fix the specimen definition once. Same dimensions, same mounting, same conditioning, every time. A batch test that varies its own method measures nothing.
- Sample on a rule, not on suspicion. Quarterly, or per production run, or per raw material change — decided in advance. Testing only when someone is worried produces a biased record.
- Keep the original type test as the reference. The question is not what the batch scored, it is how far it has moved from what you certified.
- Decide the action limit before the first result. What deviation triggers an investigation, and who owns that decision. Agreeing it afterwards, with a bad number on the table, never goes well.
A type test proves what the product could do on the day it was tested. A production verification programme proves what it still does two years and three supplier changes later.