Deciding what is in a sampling lot, choosing an inspection level… Most quality managers would say this is ‘Inspection 101’ and that their company does a good job at this.
Let's look at how you can make better choices when it comes to setting a quality inspection level and sampling lots...
Could Your Quality Inspections Be Improved By Adjusting Quality Inspection Levels And Restructuring Sampling Lots?
I have noticed that in many cases choosing what is in a sampling lot and choosing an appropriate inspection level actually can be improved. Most companies tend to settle for a ‘default’ and then apply it mechanically, without thinking. And, as a result, they either:
- Take unnecessary risks
- Spend too much time on visual inspections
Let’s use two fairly straightforward examples to illustrate these two potential problems.
Example 1: Sampling Fairly Similar Products
Generally, the products are different – a photo frame and a wood platter can be checked on the same day in the same factory.
These 2 products are made by different processes. Defects that might appear on one might not appear on the other, so it makes sense to check them in 2 separate lots:
(In the above example, let’s assume you inspected the photo frame many times and its quality is relatively stable. Therefore, you need only apply level I for that product, and level II for the platter.)
Now, let’s say you have to check 3 types of baskets, and they are each very similar. If you apply your usual logic, it will look like this (3 sampling lots):
It means the inspector will have to pick 285 samples for a visual check. But aren’t these products essentially the same, coming out of the same production process?
It would be more efficient to bundle them all in the same sampling lot. If that were the case, only 200 samples would have to be checked instead. This would save precious time:
How to divide the 200 samples among the 3 SKUs? I would say to do so according to the relative proportions of their order quantities.
Example 2: Sampling More Complex Product Selections
Sometimes it is not that straightforward. Let’s say you buy 2 styles of shirts (coming out of the same line, in the same material and the same colour) and a tie from the same factory.
The safest option would be to do 3 sampling lots:
However, it is a waste of resources! The shirts are very similar and there is no good reason for splitting them into two different lots.
The easiest option, at the other extreme, would be to bundle all three products into one lot.
This is a poor choice though, because 10% of the ties might be defective, and overall you might have to accept the entire lot simply because the shirts had very few defects. That’s a common risk when several products that are quite different are bundled together in this manner.
In my opinion, this is the most suitable option: The two shirts in one sampling lot, and the tie in another lot.
Oh, and what if the tie’s production is mostly automated, and an earlier inspection confirmed the quality of the first few pieces produced? That sampling lot can be checked in level I instead. Again, this saves time and resources.
If there is little historical information about the production of shirts, and it is mostly manual, it makes more sense to apply the deeper level II for them (the “default” level).
Making it easy to switch from one setting to another
How to ensure you optimize the settings behind your team's QC inspections, to avoid some of the pitfalls described above?
There are 2 approaches here:
- You may set some general rules - for example, a certain list of products would only be checked in level I, or even S-3 (if the risk is considered low).
- If general rules don't make sense, or if your team doesn't have time to think about this, you may set a manual review & adjustment process.
In both cases, a quality inspection software solution, such as SynControl, can support your needs and makes it a 1-click process.
Even quality managers who think that they have their current inspections running as effectively as possible may be able to save time and / or lower risks of poor quality pieces creeping through by taking another look at how their sampling lots are organised. Is this something that you've had success in doing?
Share your experiences or questions about this topic, or quality inspection software in general, in the comment section below and we’ll do our best to answer.