Skip to Content

A continuation of a series of blogs:

Blog 1: http://scn.sap.com/community/erp/operations-qm/blog/2013/09/23/qmcharacteristicsspecificationsclassification-blog-1

QM/Characteristics/Specifications/Classification – Blog 2

GC’s and MIC’s

Ok.. lets talk about this linkage thing.  How does this work and what does it mean.  First off, lets understand the main reason this is done.  We already discussed how we have two different areas of SAP that kind of grew up separately and now need to talk to each other. But in order for that communication to happen, there has to be a common language.  With regard to the characteristics this primarily means the following three items have to be compatible:

    • – Data type (char or numeric)
    • – Unit of measure, (UOM),
    • – Decimal places

This ensures that the values transferred from the MIC are not truncated, misrepresented or other wise corrupted.

There are basically only two combinations  of characteristics that can actually be linked.

  • – A GC of Char data type that uses catalogs as the values, can be linked with a qualitative MIC
  • – A GC of numeric data type can be linked with a quantitative MIC.

You can not use GC’s with a date, time or currency data type.

You can not use GC’s that are set up for multiple values.

You can not use MIC’s with the “classed recording” control indicator clicked on.

You can not use MIC’s that only allow defects recording.

Those are the general rules.  Depending on how you use some MIC’s and what control indicators you select, you may run into trouble with trying to link them to a GC.   If you try to change control indicators on a linked MIC, you will get a pop-up box like the one below.  PLEASE click the help icon in the information box.  The help there will explain some of the reasons certain control indicators can’t be changed.

Screen Shot 2013-10-03 at 5.04.32 PM.png

Now… before we actually create and link any characteristics, from a consulting point of view we need a rationale as to exactly what is going to constitute a test.  There is no hard and fast rule for this.  I’d like to think there is, but the client is always correct.

What I try to push as a standard is that if any of the following items are not the same, it should be new GC/MIC. 

  1. Basic test method
  2. UOM
  3. Number of Decimal places (Sensitivity)

So lets take an example. A simple Assay.  A common test in many companies.  We can have the following IF there is only one way to run the test:

GC_Name

GC_Description

UOM

# Decimals

FE_PCT_0

Trace FE, %

%

0

FE_PCT_1

Trace FE, %

%

1

FE_PCT_2

Trace FE, %

%

2

FE_PCT_3

Trace FE, %

%

3

FE_PCT_4

Trace FE, %

%

4

Depending on HOW the assay is run we could have multiple assays.  For instance, FE can also me determined by Mass spec and by GC-MS. 

GC_Name

GC_Description

UOM

# Decimals

FE_MS_PCT_0

Trace FE by Mass Spec, %

%

0

FE_MS_PCT_1

Trace FE by Mass Spec, %

%

1

FE_MS_PCT_2

Trace FE by Mass Spec, %

%

2

FE_MS_PCT_3

Trace FE by Mass Spec, %

%

3

FE_MS_PCT_4

Trace FE by Mass Spec, %

%

4

FE_GCMS_PCT_0

Trace FE by GC Mass Spec, %

%

0

FE_GCMS_PCT_1

Trace FE by GC Mass Spec, %

%

1

FE_GCMS_PCT_2

Trace FE by GC Mass Spec, %

%

2

FE_GCMS_PCT_3

Trace FE by GC Mass Spec, %

%

3

FE_GCMS_PCT_4

Trace FE by GC Mass Spec, %

%

4

If there are other testing techniques, then the list of characteristics would be expanded as appropriate.  

Consulting Note 1: Determine your list of general characteristics first, before attempting to develop your MIC’s.  It is much easier to create your MIC’s with a reference to the GC then it is to establish this link later on.  Even if you don’t plan to use all your characteristics as GC’s or batch characteristics, plan and develop your test list as if everything will be a batch characteristic. 

The holy grail is a single list of tests utilized across the entire enterprise with no duplications.  Plants and products are not to get their own unique tests because they think they are ‘unique’ or are different.   Remember, different method names and even different preparation methods are allowed in different plants.  The key would be to determine if customers, or an outside regulatory agency, would, on an analytical basis, accept the test values as interchangeable.  If they would, they are the same test.  If not, they are different tests.

You might have some tests that you expect to only set up as MIC’s because they can’t be linked to a GC.  That’s fine, but go through the same exercise for all your tests and pretend as if they will all be linked to GC’s.  Once your master list is established, you just identify these as not needing GC’s.  But your descriptions, decimal points, UOM’s etc. will still all be needed to set up the MIC. 

Conversely, you might have GC’s that won’t need MIC’s.

There is no real hard and fast rule..  But you want to attempt to avoid duplication of what is basically the same test.  This is simply a guideline and based on my past experience.

Consulting Note 2: Some UOM’s are directly convertible.    For instance, %, PPM, PPB and PPT can all be converted by SAP. Laboratories across the enterprise should standardize on what UOM they wish to record against in the laboratory and internally as a business for a given test.  If the data needs to be reported in different UOM’s, this can be done directly in the COA profile.  I.e. you record in % but report to certain customers in PPM’s.  If you do this, you may also need to consider a custom text description for the characteristic within the COA profile. You are installing a computer system, maximize it’s usage!

Consulting Note 3:  Some tests can easily be calculated from others.  A good example are tests in the chemical industry called “Screens”. These tests are primarily for evaluating particle sizes.  It involves using a screen(s) to pass material thru.  The amount of material left on the screen is an “ON” test while what falls through the screen is a “THRU” value.  A product may be passed thru three or more screens.  A THRU value can be calculated as 100% minus the ON %.  And of course the ON % can be calculated as 100% minus the THRU % value.  Some customers want the “ON” value reported to them on the COA, while others want the “THRU” value reported.  Don’t make your lab techs record virtually the same test value multiple times in an inspection lot!!!

Plan to set up the system so that all laboratory personnel record one or the other.   In the above example get the technicians in the lab trained to do all THRU measurements.  Then all the “ON” tests are placed as calculated characteristics at the end of the inspection lot.  Don’t make them record both because some customers want the ON value while others want the THRU value.  And don’t have them record two THRU values and one ON value.   Pick one or the other for the techs to record.   All of course will all need to be set up as GC’s and MIC’s.  These are just a few examples.  I know there are other series of tests like this.

The idea is to get the labs to be consistent so that everyone records things the same way in all plants with NO duplication of recording.

This helps in training and in reducing lab errors.

Consulting Note 4: Try to establish the list of characteristics as if there were no customers.  Stick to what  you would have the lab testing your products for if you were just bringing the product to market with no customers.

There really should be NO customer specific tests. They are simply tests.  You might choose to run a test on a product for a customer, but during the setup and design of your characteristics it should not be designated as such unless the test is globally known as the “ACME WETTING TEST”.  (yes.. I totally made that up.)  These types of tests are usually application specific tests.  I.e. they are unique to the customers end use of the product, for example, time for a particular paint formulation to dry on a customers proprietary substrate material.

You will have tests, mandated by customers, that must be set up and run as part of a material’s standard inspection approval process.  Maybe because that material is only sold to one customer.  But don’t make all the tests on the product as customer specific tests.  Only a handful will truly be application or customer specific.  Set them up simply as another test.  If you scan down your final list of GC’s and MIC’s to create, and in the test descriptions, you constantly see customer names cropping up, you’re probably headed for a difficult implementation regardless of what final design you settle on.

Ok.. probably a good stopping point here.

I expect my next blog will deal with the general characteristics and their interaction with the MIC’s when they are linked.

Please add comments and feel free to add any “war stories”.  I love to hear of good and bad examples!!!  Especially bad!!

This blog is Copyrighted October 3, 2013 by HCS Enterprises, Inc. This blog may not be used by anyone for any commercial purposes without permission.  It may not be reproduced, in part or in its entirety, in any form, printed or digitally reproduced.  Permission is granted to individuals to maintain a copy for personal use and reference only.  It, or any part of it, is not to be used for training documents, training classes, project documentation, end-user handouts, cheat sheets, reference documents, slide presentations or any other use that may not as yet be envisioned.  Permission for use is provided to SAP under the terms and conditions agreed to by participating on the SCN website.

To report this post you need to login first.

4 Comments

You must be Logged on to comment or reply to a post.

  1. Nitin Jinagal

    Hi FF,

    That is really great work. I hope you would also share your knowledge about different Inspection types and their usage in detail. Many of us know most about them but it would be advantageous if the lesser used types could be explained and modified versions can be introduced. e.g I know insp type need to be maintained in order type for 03, but not for 04. Similarly, other unique traits of all types and knowing basic reasons behind would be so good to identify the best out of two in a particular scenario.

    Kind regards,

    Nitin

    (0) 
    1. Craig S Post author
      Thanks Nitin,  I will definitely keep your suggestion in mind for an inspection type series.  Maybe some of the other guru’s might think about doing a few blogs on this topic as well.  I’m sure there are some other folks here that have more experience than I do with some the lesser used inspection types.  FF
      (0) 

Leave a Reply