We have an automation / vision system that is monitoring the parts for defects that we teach it to look for. Issue is, these vision systems can only look at a particular part of the container....where we "teach" it to check so if the metal doesn't land in that exact spot it doesn't get kicked out. We do not have metal detectors on our line.
We are considered primary food contact packaging and that's why I am being so meticulous with my research. That post you referenced was an identical one I posted in May when I was first asked about the issue, but no one replied.
Since I have a Quality background, my answer is I want no metal defects to be present or allowed, but Engineering is looking for some kind of AQL for metal either in size, type, amount, etc. from me in case there are machine issues that lead to this happening. I am just trying to find out if other companies have a standard, AQL or something they abide by when it comes to metal.
I appreciate your help.
Thks for reply.
Unfortunately It seems very few people here are using the type of monitoring system you refer. I speculate that if this hazard is relevant to their production they are using metal detectors ?
I can recall a Codex example where a sieve's performance for rejecting metallic contamination is evaluated/validated by setting a maximum percentage of "passed metal" in its output.
However I doubt that an AQL approach is generally suitable since, despite the "tolerances" discussed in yr OP attachment, this hazard IMO tends usually to be practically viewed/handled as more of a zero-tolerant type.
Hence the rejection function in metal detectors for any "magnetic" contamination exceeding the MD's limit of detection (LOD). The latter is typically expected to be set at the highest sensitivity available/usable (ca. LOD). (In practice, for MDs, specific characteristics may "interfere" with ideal LOD, eg shape/configurational aspect of contamination).