False Positive Metal Detection Reject
Hi,
What is the correct process to take when you have a product that is packaged with metalized film but every now and then the metal detector will reject one, but you know it is due to the packaging . Are we still required to stop, and run all product back through from the last good check? Even when no metal is found?
Hi,
What is the correct process to take when you have a product that is packaged with metalized film but every now and then the metal detector will reject one, but you know it is due to the packaging . Are we still required to stop, and run all product back through from the last good check? Even when no metal is found?
If the machine rejects more than necessary why would you rework additional material? We have had high moisture/salt items that would give extra false rejects for metal detectors, but if they could be confirmed as false rejects would just be considered an unfortunate side effect of the formulation. Nothing more than lowered efficiency.
Defaulting to rework is for if it fails to reject when a known FM sample is present.
Correct, our process is to run it through 3 more times once this happens to verify a false positive. We got dinged on that today. Just trying to figure out what would be an appropriate method to implement since running the particular item back 3 times through the CCP is "not the way to handle this", per auditor
I very rarely side with auditors---but in this particular case I do
You should be following your deviation procedure--which should include:
Holding all product since last good check
Opening said packet and running the CONTENTS to verify that the actual food is not rejected
Running just the packaging in question to verify it IS rejected
Then you can release the product that is on HOLD
Assuming it's the packaging is nullifying having a MD at all
I would also be adding a non conformance back to the packaging manufacturer as their flaws are affecting your throughput
I would ask a different question... Ok I'd ask a whole heap of questions!
1. How do you know it's your packaging? Now that opens a whole can of worms... I've had metal detection "false rejects" which were later found not to be false. In fact in one site we're considering using it as an early warning system of something trending out of control. Just because on a physical inspection you didn't find anything doesn't mean it was false.
Metal detectors are really good at detecting metal but not necessarily at telling you what form it's in. For example, metal detectors can detect "metal dust" or swarf from wearing machinery (and do so before an x-ray sometimes if the total quantity in the pack is above the detection limit even if they're all very tiny pieces.)
So I'd be REALLY sure that it's definitely your packaging. What you could do for this is a natural experiment. For example, look at false rejects over time. Have they changed? If you have more than one line using different equipment, are the levels the same? Also it's good to sense check if you're getting any consumer complaints for metal? If this flags up any concerns at all, I'd be VERY wary in assuming it's just the film.
2. Why have you not changed your settings or equipment? If your packaging is setting off the detector then that's not a great test. So if you're absolutely sure it is the packaging, let's look at a few things:
a) You should be designing a test which minimises false rejections. It looks like you've not done so here. Could you decrease the sensitivity slightly while still being an effective test for the kind of machine damage you may get from time to time? This may seem counterintuitive but decreasing the sensitivity may increase the food safety. Right now, your staff will assume every activation is false. If you can really minimise false rejections, they're more likely to take a reject seriously.
b) There's other kit out there. Alternatives could include x-ray (expensive but effective) or if this is into a bag you can get throat metal detectors which detect just before the package is made and so is only checking what's going into the pack not the packaging. If you can't afford to change equipment now, can you plan capex for this for next year or the year after?
3. Why are you going back to the last good check? From your question, as I understand it, this is a false reject not a failure of a test pack to reject. Is that correct? If so, there is no need in UK standards at least to recheck since the last good check. That's if the test failed. A false reject should be rechecked individually and broken down until the contaminant is found and ideally shouldn't be put back into product flow (although some standards permit it if rechecked without detecting anything three times in three different orientations.)
4. Are you REALLY sure about the rest of the line set up? The reason I ask is I'm sure our British operators are no different to any others in that they are renowned fiddlers and tweakers! Even settings behind high level passwords end up with someone finding out, normally within a day (I am not kidding) and making a change to the programme. I really would rule this out. What if your "false rejects" were actually mistiming and rejecting the wrong pack?
5. Why are you living with this? I suppose that's been the whole point of my reply. But if I audited you it would alarm me from a cultural point of view. Because operators would almost always ignore rejects (and risk putting them back through without sufficient checking.) It also is an indicator that investment for food safety isn't priority...
Correct, our process is to run it through 3 more times once this happens to verify a false positive. We got dinged on that today. Just trying to figure out what would be an appropriate method to implement since running the particular item back 3 times through the CCP is "not the way to handle this", per auditor
The method you are using to determine IF it is a false positive or not would appear to be the problem. Your current method is basically a "best three out of four" where legitimate rejects can escape.
You'll need to assume the rejects are a problem and examine them for cause. Scampi and GMO detail some methods for more thorough review.
The method you are using to determine IF it is a false positive or not would appear to be the problem. Your current method is basically a "best three out of four" where legitimate rejects can escape.
You'll need to assume the rejects are a problem and examine them for cause. Scampi and GMO detail some methods for more thorough review.
Some schemes in the UK used to allow this, I have to admit to not being totally up to speed with retailer standards to date so I'm not sure if they've changed. It's not really "best three out of four" but rather pass through three more times in different orientations and if any one of them fail, it's a fail". But some retailers, like M&S and JS I think have banned that practice for a while.
To be honest I think all should. While I'm as keen as anyone to prevent wasted food, if you're getting loads of false rejects you're not addressing the root cause by giving people an alternative. You're just living with the problem.
Correct, our process is to run it through 3 more times once this happens to verify a false positive. We got dinged on that today. Just trying to figure out what would be an appropriate method to implement since running the particular item back 3 times through the CCP is "not the way to handle this", per auditor
1, If it's just your normal product being incorrectly rejected, then nothing needs to be done in terms of the technology itself, as long as that incorrect rejection is something you think you can accept.
2, But at the same time, your standard test piece used to verify MD, it almost absolutely must be rejected.
3, Generally speaking, the false rejection rate of normal products should be kept below one in a thousand.
If it's too high, it will cause a lot of trouble, such as operator crashes, and it will also ignore the real metal.
4, For standard test pieces, this should be controlled to at least one part in ten thousand.
When it's above about the 1 in 1,000 level, you lose confidence in the reliability of the MD.
5, These requirements, in principle, are not directly related to metal films.
For metal film, it is a MD technical issue, you need to get a professional technician to commission it carefully, MD is actually very complicated, its actually much more difficult than X-ray.
We are the tech support for MD, X-ray check weight etc, of course it doesn't seem to be possible to post direct ads here, but I'll do my best to provide support for any questions.