Metal Detection Failsafe

We have just had our BRC audit in which the auditor had issues with the Fail-Safe in which when the air is turned off there is a 3 second delay before alarms sound (which he argues is too long even though we recall all product from the last successful metal detection test if the machine alarms sound).

I was wondering if anyone has had these issues before and if someone could give me any advice on dealing with this? 

Thanks!

Dear Bronaghmk1,

if your Quality System tells that you check everything from the alarm/failure to the previous check.. there is no auditor that can think that 3 seconds are too long.

All is described and you have a corrective action, so for me, no problem at all.

Roby

Dear Bronaghmk1,

 

if your Quality System tells that you check everything from the alarm/failure to the previous check.. there is no auditor that can think that 3 seconds are too long.

 

All is described and you have a corrective action, so for me, no problem at all.

 

 

Roby

so you think as a corrective action we would be passed if i was to include and highlight the checks we have and the corrective action already in place? Or would i be best to create a new SOP aimed directly at the Failsafe element and list all actions on this document?

Dear Bronaghmk1,

in my Quality System there are:

-Procedure for metal detenction

-Training for operators dedicated to metal detector

-Instruction on "how make the test" on the machines 

-Instruction on "how to manage failures"

in the last document i wrote what we do and how we test metal detectors... never got not conformities

Roby

"even though we recall all product from the last successful metal detection test if the machine alarms sound)."

Is that all documented?

Hello,

Is your metal detecting step is continuous process? What is the speed of your conveyor? 3 seconds is too long in my own opinion. Many products have already passed.

regards,

redfox

Hello,

 

Is your metal detecting step is continuous process? What is the speed of your conveyor? 3 seconds is too long in my own opinion. Many products have already passed.

 

regards,

redfox

yes it is a continuous process i am not sure of the exact speed but one pack is going through in those three seconds before the alarm sounds how does your failsafe work if you dont mind me asking? 

Hello,

Our SOP in MD: we verify sensitivity of the MD using ferrous,  non-ferrous and SUS, before the operation, every 30 mins thereafter and right after the end of the production. The product in canister weighing 350 grams is feed manually to the conveyor on the center-most side. 

If on the latest verification the MD fails, we will recheck of if possible rework the product from the latest check when MD is functioning well and to the latest when the MD fails. It covers all the product that passed through the 30-min period.

regards,

redfox

Hello,

 

Our SOP in MD: we verify sensitivity of the MD using ferrous,  non-ferrous and SUS, before the operation, every 30 mins thereafter and right after the end of the production. The product in canister weighing 350 grams is feed manually to the conveyor on the center-most side. 

 

If on the latest verification the MD fails, we will recheck of if possible rework the product from the latest check when MD is functioning well and to the latest when the MD fails. It covers all the product that passed through the 30-min period.

 

regards,

redfox

so.. it is not a problem if there is a 1 or 3 or 5 seconds delay :) if it is all documented and checked regullary by a trained worker and you check everything since the last check :)

Hello,

In our case, since the product is manually fed, we don't see any problem relating to fail-safe, because no product will be fed to conveyor until the verification is done. Product that passes the conveyor on last verification will wait until the latest verification is done. 

regards,

redfox

To be honest I would be concerned with a 3 second window. Depending on how fast product passes through you would find it difficult to pin down which pack / bottle actually failed. Our machines have an instant stop when metal is detected. When the belt stops and the alarm goes off he affected product is still under the head, with a 3 second gap that product could already be stacked on the pallet. Could you machine settings be altered by the manufacturer to make it an instant stop or would your company consider an upgrade? (our machines are Loma if I am allowed to say that) Also with that gap reworking all product since the last test would have the same issue, if something triggered the alarm how would you identify the exact pack with that gap? On our bottling line 3 seconds would mean we would have 15-20 bottles at the end that we didn't know which one triggered the alarm making investigating the non conformance hard as well. 

Hi,

There are no set rules on when a failsafe should stop production, it is down to opinion only and suggested times by the manufactures of the machines.

Has the auditor raised an non conformance for this finding, if so I would be inclined to appeal it should it mean you can get a higher grade...providing you can comply to the below fully.

The issue can be closed if you can show detailed control and sufficient training, control would be based in a procedure to cover all product since the last good (failsafe) check, should the failsafe engage (it is not sufficient to say 'once a failsafe engages we will have the line repaired before starting' and training for the person(s) involved in the operation of the machine who would be able to explain to the auditor that should a failsafe engage they would inform QA, eng dept and recheck every packed product from the last successful failsafe check, which could be the entire days production should the failsafe be tested at the start of the day and rechecked at the end of the day !!

Hope this makes sense

Hi, Bronaghmk1,

How is the operation of the "fail-safe"? Is "fail-safe" means if the air pressure is than required, will the ejector system continuously in reject mode and/or the machine stops? Or the line will continue to run and the  "fail-safe" mechanism only generate alarms and no effect on the ejector or machine? If it is the former, then there should be no issue as there will be no product passing through. If the latter, then there must be a document stating what needs to be done in case of non conformity. All of these should have been considered during the validation of controls (considering the worst case scenario). The risk of having 3 seconds "fail-safe" will now be then dependent on the operation.

It sounds like your "Fail Safe" makes more work than it saves.

If the metal detector stopped the line or rejected immediately, you would not have to check back to the last test. If the metal detector reacted in a way that isolated the metal, you would have a lot less work. 

Instead, your metal detector alarms AFTERWARD, requiring you to go back through all of that material. 

I don't think it's unreasonable, as an auditor, to question how often your staff actually goes back through all that material as how would you prove they did or didn't? Customer complaints? Records of inspection? Digital production and metal detector logs? 

Again, it sounds like this machine just makes your job harder. 

3-year old thread.

Thanks anyway. But you need to consider retesting back to the last "good" validation.

Hi All, 

Could you tell me how often failsafe check needs to be completed for BRC accredited site. Currently we are testing at the start and end of the run and then hourly check which contain consecutive check. 

Failsafe check  contains;

1. Bin Full,
2. Bin door lock,
3. infernal lock,
4. sensor check, air supply,
5. 8 bags full checks 
6. memory test
7. sensitive check - is this need to be checked
8. belt stop,
9. checkweigher test

Thank you for your help in advance.

In all my years, only ever did the checks with the imbedded wands or rods

IMO, maintenance should be doing 1-2-3-4-6-7-8 checks as per the manual frequency

The checkweigher frequency of verification should be based on your volume/speed of the production line, as long as your calibrations are done to manufacturers schedule

I know this is an old thread. The fail safe check is to check that you will be notified of a failure of air pressure which is usually critical to the reject mechanism, however a detection of metal during the failure of air pressure will still be prioritised by the machine.

Depends on your reject mechanism but ours will stop the MD conveyor and the following conveyor immediately if metal is detected but a product isn't detected falling into the reject box. 

I would argue with the auditor that the line would be stopped due to the other fail safe, but this depends on how quickly your line stops if metal is detected but does not get rejected.