The question how to setup the MIC structure within QM is rised quite often here. I would like to collect possible architectures here, together with their advantages and disadvantages. Any additions are very welcome!
| Architecture |
Advantages |
Disadvantages |
| One separate plant (master data plant, virtual plant, template plant) for all MICs, contains no operational data |
- Central maintenance and guidance can be easier enforced
- Authorization can be strictly separated from operational ones
- Positive effects on reporting
- Duplicates can be identified quickly
|
- Slightly more effort in customizing as it requires a separate plant
- Authorizations (e.g. for displaying the MICs) have to be provided to the other plants and therefore be included in their operational roles
|
| One central plant (corporate plant) that is used for all MICs, but also contains operational data from the corporate unit |
- Central maintenance and guidance can be easier enforced
- No effort for separate plant
- Positive effects on reporting
- Duplicates can be identified quickly
|
- Authorizations difficult to separate, specially for the non-corporate plants to access and maintain the MICs. Might result in unwanted access rights
|
| Several plants (sites), each holding their needed MICs beside their operational data |
- No additional efforts for authorizations, can be added to the local roles
- Each plant gets exactly what it needs, no big catalog to select MICs from
- Emphasizes the idea of independent responsibilities within each plant
|
- Reporting will be difficult, as identical MICs might have different names in the plants, or even worse: MICs with the same name might be totally different concerning their content
- Coordination effort to ensure data quality is very high
- If using batch management and where multiple plants make similar/same products and run the same tests significant problems can be encountered when producing certs, producing accurate batch data analysis, and tracking down quality issues. Usually voids the use of much of batch determination and complicates inter-plant transfer of batch manged material with regard to quality data.
|
This content has been composed for this initial version on the basis of the experience I gained so far from implementations, but also by some comments from other posters (e.g. FireFighter, Sujit,..).
Thanks Martin! This will actually dove tail nicely with a series of blogs and documents I'm working on. Hopefully I'll start putting them out shortly.
FF
Would be great to read some details soon. I also have a virtual "would be nice to have as document here" list in my mind, but as usual time is the lacking factor...
Regards
MH
Craig and Martin
Good Master Data maintenance consulting approach. Most of the time Consultant spend time on traditional thinking similar to Green field implemenation. Over the period of time business expands , roles & Responsibility well documented, then we can think of Global Master data ( Client Level ), Local Master data ( Plant specific ) in major rollout or Big consolidation / Harmonisation projects.
All in all , very good inputs for Consultant 🙂
Regards
Girish Deshpande
Thanks a bunch for this article, Martin. Extremely helpful! We are using option 1 in our company and I am able to relate to it ONLY after reading your blog.