Use of selected sets

[maarten_muetgeert2]

I've a question about the use of selected sets for charateristic MICs.

One of the characteristics is appearence. Per material we can have a different specification. For example appearence for material A can be 'clear colorless solution' and for material b 'clear, yellowish liquid'.

How should I use here the MICs and selected sets? Should I add the specification in the description of the MIC and use one general selected set with the options OK, Not OK? Or should I use one MIC 'appearance' and create 2 different selected sets?

Thanks.

[former_member42743 · Accepted solution]

This is an area I think SAP could stand to improve on a lot here.

You pretty much nailed the choices.

Your selected set is unique for the test.. not the material being inspected.

So if you have a test like appearance where you have different "OK" values for different materials you have to create a separate MIC for each combination with it's own selected set.

So you can't really have just one appearance test and somehow use specifications to influence what codes are OK and which are not.

On the other hand, your selected sets will only have specific codes related directly to that line of products. 

So, depending on your business and needs there are two approaches.

1) A General "Appearance" test set up with just a PASS/FAIL value.  The "pass" description is than either included in the text of the unlocked MIC in the plan:  "APPEARANCE - CLEAR & COLORLESS" or if you want to keep the MIC locked, (my preference), in one of the extra INFO fields available.  You can also link a method to the MIC which provides the specification.  In most case, experienced techs in the labs know what they are looking for anyway.

2) Multiple "Appearance" tests each with their own selected sets.  This is more master data intensive but is usually the most detailed and allows the best statistics and fewest errors.  Usually, for most companies this is a one time hump of master data to get over.  Once you've created 20-30 appearance tests, you can usually pretty much use the same ones over and over unless you suddenly develop a whole new product line or you take over another company or something.

I prefer option 2 where possible.. but that's because of the primary industries I serve, Pharma and Chemical.

FF

[maarten_muetgeert2 · Accepted solution]

So what you then would get is the following:

MIC     Description     Selected set     Codes

001     Appearence     Clear colorless solution     Conform / Not Conform

002     Appearence     Clear yellowish liquid         Conform / Not Conform

I could then use the internal description or search field to define the spec in the characteristic. Or would you do this in the description itself?

They're (customer) coming from an old LIMS system so they've no idea what should be the best way. This is also for a pharmaceutical company.

[former_member42743 · Accepted solution]

If I was doing this for a pharma, with no LIMS involved, I would probably recommend doing individual MIC's each with their own selected set of codes.  Especially if you want them to have more than one or two accepted codes or grades and you want to capture those minor differences for statistical purposes. 

So for a product a "clear, medium yellow" and a "clear, light yellow" and a "Clear, colorless" might all be acceptable values.  But they want to know the yellow vs. clear for whatever reason.  So part of the question is what would they want to see in the QMIS?  If you don't set up the selected sets you just see the pass/fail which gives you very little statistical value for QC.

FF