Pretty driven: The difference between MBSE and MDSE
Modeling has long since found its way into Systems Engineering held. But there are different terms for this. In particular, you often come across MBSE (Model Based on Systems Engineering) and MDSE (Model Driven Systems Engineering)? I would like to clarify the differences below. And while we're at it: Of course, models can also be used to tackle other topics, such as architecture (MDA) or design (MDD). Incidentally, the article 18 things you can do with system modelsand the SYSMOD Intensity Model.
Driven or based?
On the differences between the various approaches Jordi Cabot wrote an article worth reading in his blog. I can only agree with him that "model-based" is softer than "model-driven." The former means that models are important, but not necessarily "first class" artifacts. In a model-driven approach, on the other hand, the models are central artifacts of development. As a concrete example, in the first case the model could be used as documentation for the software developers, but in the second case the code would be generated by the model. This distinction also means that MDSE is a subset of MBSE: to exaggerate, a state diagram inserted by copy & paste could already be described as MBSE.
Engineering, Development and Architecture
The next question is about the meaning of the letters that follow "Model Driven", i.e. MD(S)E, MDD or MDE: Engineering, Development and Architecture. Jordi has also taken a clear position on this in his article. I also see engineering as the overarching term, and for me, requirements management (and therefore the modeling of requirements) also falls into the category of engineering. For me, engineering covers the full breadth of systems engineering.
Development, on the other hand, is about deriving an implementation. Here, code generation is often mentioned as an easily recognizable feature. This is also the area in which UML is traditionally used.
By architecture, I mean the framework into which the parts of the system to be developed fall. The architecture model is therefore less detailed than the development model. Jordi has a slightly different view, as he only considers the OMG's understanding of MDA considered. Under this assumption, I agree with Jordi's analysis. However, I see the OMG's view as only one possible interpretation of MDA.
However, the result is the same: the various approaches mentioned here can be understood as nested subsets. This was nicely visualized in Jordi's blog using an image taken from the master thesis of David Ameller and can be seen in the cover picture.
Conclusion
A clear understanding of the terms is the basis for any discussion. We should not assume that the terms discussed here are firmly established. It will certainly take a few more years (decades?) before the same image appears in everyone's mind when the term "MBSE" is mentioned.
Image: David Ameller / Free Stock Textures
With regard to the definition of MBSE, I see things a little differently. In MBSE, models are first-class artifacts. The models are the masters of the engineering information and documents, for example, are views of the model. By „based“ I mean that the models are the basis of the engineering. By „driven“ I mean that the model artifacts are always the source for further (generated) artifacts. The boundary between MB and MD is certainly blurred.
I also find the definition of MBSE from INCOSE somewhat weak. It says that the purpose of modeling is „to support“ engineering. I would rather call it model-supported systems engineering (MSSE), although I shy away from using another term and abbreviation.
Hi Tim, your argument is understandable, but as you say yourself, there are many definitions here, some of which are contradictory. Basically, I found Jordi's approach quite charming, and wanted to give it a bit more visibility. I also found the nesting in the form of subsets helpful: even if the boundaries are blurred, we should agree, for example, that MDSE is a subset of MBSE.
As for the introduction of new terms, I'm glad you're shying away: I can't help but think of this XKCD: https://xkcd.com/927/ :-)