Interview with Tim Weilkiens about SysML
Tim Weilkiens should be familiar to anyone who works with SysML. He is the author of several books on modeling and is co-developer of the OMG standard SysML. He is also a board member of oose, a consulting company based in Hamburg. With SYSMOD, he has created a toolbox for system modeling that can be used well with SysML.
At ReConf 2016, I had a conversation with Tim Weilkiens:
You played a key role in the SysML standard and also represent it at the OMG. How is SysML perceived by the SE community today? How far has SysML established itself at MBSE?
Well, you don't necessarily need SysML to practice system development, regardless of whether you are modeling or not. The impetus for SysML came from customers who wanted to use UML for system modeling. That was about thirteen years ago. There was a lot of hype at the beginning, but then it died down. In the meantime, interest has increased again. But the modeling language is of secondary importance, as it is more about how it is used.
SysML itself is still very much in flux. For example, with SysML 1.3, which was published in 2012, there were some major changes. I also have the feeling that many aspects of SysML are not yet well understood by all users, even so seemingly simple things like associations.
I can't really recognize a community around SysML. Of course there are activities at the GfSE and other groups, but it's all still very much in its infancy.
SysML itself is hardly ever questioned when it comes to MBSE. Thanks to the good tool support, this is generally a given and users can concentrate on the important topics such as methodology.
I didn't really invent SYSMOD. I simply compiled what is practiced and works. So I am the editor of SYSMOD, but not the inventor.
You presented the SYSMOD method in your book "Systems Engineering with SysML/UML" back in 2006. Now, almost 10 years later, you have published your own book on SYSMOD. Can you briefly outline in your own words what SYSMOD is? Who should use SYSMOD? How is the acceptance of SYSMOD?
I didn't really invent SYSMOD. I simply compiled what was practiced and worked. In this respect, SYSMOD is not a method or process work for me, but a collection of proven techniques in modeling with SysML.
I think you could compare SYSMOD with design patterns, i.e. a collection of large and small tactics and strategies for solving typical problems. Does this analogy fit?
Yes, that is a good comparison. The toolbox is alive and should be continuously expanded and adapted.
I had written the SysML book in German at the time. The first edition was later translated into English, but the following editions were not. That was a shame because new insights and ideas were not incorporated, at least not in the English version. That's why I also published the new book in Self-published by Leanpub published. It is available as an ebook and it is possible to update the book there regularly. Readers who have already bought the book will also receive these updates. I plan to update the book once or twice a year.
I don't know to what extent SYSMOD is used. But in most cases, the procedure is adapted. The term SYSMOD is rarely used, but that is not important.
SE-ZERT, i.e. the SE certification of INCOSE (SEP) in the German language, has been available in Germany for a relatively short time. This is also offered by oose. What has been the response from decision-makers and engineers?
We are getting a lot more feedback than I had expected. In particular, there is a lot of interest in the preparatory training. After all, the training requires a considerable amount of time. But organizations not only find the certificate important, they also pay a lot of attention to ensuring that the certified knowledge is available and can be applied in the project.
Organizations also take great care to ensure that the certified knowledge can be applied in the project.
What are your plans for the future?
One topic that has been on my mind for some time is variant management. Although this is possible with SysML, it is not easy. I am currently writing a book on this topic, which I am also publishing with Leanpub in my publishing house MBSE4U: Variant Modeling with SysML.
