What's the next step after failing an AQL?

F

FMBOCCONGELLE

Hi there! Long time voyeur first time posting.[FONT=&quot]

I distribute stainless steel water bottles and I have been doing AQLs on the imported product. I am importing anywhere from 160 - 3000 units of any one different type of sku. I Usually due a .1 AQL and often find that the audit fails the AQL. What is usually the next step in this process? Do I do another AQL? How do I build my confidence without doing a 100% inspection.

I look forward to any help. Thanks!

[/FONT]
 
S

Sturmkind

There are several contributors to your question and some clarification would help.
1) Is there an abnormal existence of 1 particular defect?
2) Has there been any changes in the decision clarity of 'borderline' defect cases?
3) Are the defects isolated to single production lots or mixed production lots?
4) Does the existing sampling plan allow for clearing all rejects from the inspected batch and then reinspection of the entire lot?
5) Defect severity may need re-evaluation. If the defect(s) have a minor influence on the next user, then a review of the AOQL may be in order.

Double sampling plans (Dodge-Romig, Wiley Classics Library) can be an effective tool if adequately understood and supervised.

More importantly though, has the reject data been communicated to the supplier with a request for containment and corrective action?
 
D

DrM2u

I am backing Sturm on this one. What I would suggest in addition is to evaluate if your AQL level is adequate or too stringent based on the product's application. If it is too stringent then loosen it up and re-do the inspection based on the new AQL.

If the current AQL level is adequate then, like Sturm said, you need to communicate your requirements and findings to your supplier(s) and ask them for corrective actions. I would also aks for certified shipments (100% inspected?) until the corrective actions are implemented in order to minimize your receiving inspection.
 

Jim Wynne

Leader
Admin
Hi there! Long time voyeur first time posting.[FONT=&quot]

I distribute stainless steel water bottles and I have been doing AQLs on the imported product. I am importing anywhere from 160 - 3000 units of any one different type of sku. I Usually due a .1 AQL and often find that the audit fails the AQL. What is usually the next step in this process? Do I do another AQL? How do I build my confidence without doing a 100% inspection.

I look forward to any help. Thanks!

[/FONT]

Whatever your sampling plan is, it seems to be working, so I don't know why you'd want to mess with it. As others have suggested, you need to do something with the results.
 

somashekar

Leader
Admin
A very important application of AQL sampling plans is to get agreements with your suppliers and they in turn apply the same or better AQL sampling plan at their outgoing product stage. This will make the producer and customer operate on a level field.
 
F

FMBOCCONGELLE

Wow thanks guys!

We have a QC program set in place at the suppliers and things are always improving due to a strict AQL. I am the last person to inspect the product before it goes off to our dealers. I'm concerned about 2 things, one being another AQL on product that has been inspected at the factory and it failing and secondly, how do I ensure to our customers that there is a 98% confidence level. I think I may be too picky.

The big problem is that the defects or blemishes are never consistent. The blemishes are mostly minor and if I do find a major issue I am sure to find more. This has been a problem with the labor in the factory.

I just don't know. I would hate to lose a client because they received 1 full carton of bottles that are all bad because it was the last of a run and someone thought better of it and gave it a pass. This has happened.

I am basically a small pick and pack business in WA and I now have contracted to manage QC and now have to implement a program. I know it begins at the factory and I shouldn't have to an AQL a second time.



Thanks for listening to me rant and all the help.
 
Last edited by a moderator:

somashekar

Leader
Admin
Wow thanks guys!

We have a QC program set in place at the suppliers and things are always improving due to a strict AQL. I am the last person to inspect the product before it goes off to our dealers. I'm concerned about 2 things, one being another AQL on product that has been inspected at the factory and it failing and secondly, how do I ensure to our customers that there is a 98% confidence level. I think I may be too picky.

The big problem is that the defects or blemishes are never consistent. The blemishes are mostly minor and if I do find a major issue I am sure to find more. This has been a problem with the labor in the factory.

I just don't know. I would hate to lose a client because they received 1 full carton of bottles that are all bad because it was the last of a run and someone thought better of it and gave it a pass. This has happened.

I am basically a small pick and pack business in WA and I now have contracted to manage QC and now have to implement a program. I know it begins at the factory and I shouldn't have to an AQL a second time.



Thanks for listening to me rant and all the help.
A sampling plan best represents the most true results on outputs coming from controlled processess. This is as you and me know, a statistical method.
You have the liberty to adopt different AQL for different characteristics.
I just don't know. I would hate to lose a client because they received 1 full carton of bottles that are all bad because it was the last of a run and someone thought better of it and gave it a pass. This has happened.
Your above said situation is not at all a controlled process. Having known it, you have to adopt different techniques to stop this and bring about a consistant output levels that your system is capable of. Your sampling plan will them be more true to you.
 
F

FMBOCCONGELLE

Thanks Somasheka. I think your right about having to break everything down into it's own process or issue and set an AQL for each.

best wishes.
 
Top Bottom