Will MBSE become mainstream this decade? Three expert opinions
Last week I took part in the Automotive Day in Böblingen. It was well worth the visit: interesting presentations, a beautiful location and a great supporting program. But I also like to use such events to capture atmospheric images. For a while now, I've been trying to better understand how quickly MBSE will spread. So I put the following question to many participants:
"Will MBSE become mainstream this decade?"
In addition to many responses, I would like to highlight interesting points from three participants. I was lucky that Will StreitIBM Vice President, Product Management, had traveled from Austin, Texas. Danilo BeucheCEO of Pure Systems, also gave me some good ideas. And of course the opinion of Andreas Willert who organized the Automotive Day with his company.
Andreas Willert: It depends on the definition of "mainstream"
To decide the question, we must first define what mainstream actually is. The question: "Will MBSE mainstream in this decade?" answered Andreas Willert "Yes" without hesitation. But then we looked for a suitable definition. One possible one is:
A technology is mainstream when more than half of the teams for whom it would bring added value actually use it.
Once we had agreed on this definition, he agreed that this decade looks rather bad for MBSE.
ALM as a mainstream technology
To get a better feel for the "mainstream", we looked for other technologies that had already made it into the mainstream. We quickly came up with ALM to talk about. Here, too, we could discuss at length what is actually meant by ALM. But if we look at how software is developed nowadays, we see an enormous amount of evidence for reasonably sensible ALM. On the one hand, this is due to the low entry threshold thanks to inexpensive or free tools and platforms such as Jira, git, gitHub and the like. On the other hand, software developers have always had a need to try out the latest technologies.
Will Streit: Drastic increase in inquiries at IBM
Will Streit is Vice President, Product Management at IBM and therefore also responsible for tools such as DOORS Next and Rhapsody. When asked about the future of MBSE, he referred to the drastic increase in inquiries in this regard over the last two years. He emphasized that these come not only from customers, but also from many partners and also from groups not previously associated with IBM.
Of course, it would be interesting to quantify this statement: Is it an increase from 0.01% to 0.1% or from 1% to 10%? In principle, this statement confirms what Andreas Willert said: interest in MBSE is increasing noticeably. But there is still a long way to go before it becomes mainstream.
I did not receive any reliable information on the subject of Rhapsody. But more on the tools in a moment.
Danilo Beuche: Systems engineering is often enough
Danilo Beuche has already published an article on SE-Trends on the subject of Standards for product line engineering published. His company sells solutions for variant management, which can also (but not only) be used in combination with MBSE. But even from this perspective, I took a lot away from the conversation with Danilo.
One interesting thought was that there is indeed a great need for improvements in product development. However, not all challenges are so great that MBSE is the most sensible solution. In particular, for many companies with less complex products, the step to systems engineering is more expedient. Modeling, at least on a large scale, is neither necessary nor does it necessarily offer the best cost-benefit ratio.
Here again, a comparison with ALM is helpful. I hardly know of any company that works in software development without version management (git, Subversion) or a ticket system. The use of these technologies is not yet a holistic ALM, but can be sufficient for small developments. Systems engineering without the full MBSE vision is also sufficient for many teams.
Tools as a sticking point
For a subject as demanding as MBSE, proper tool support is essential, even if the tool alone is not enough. In this respect, I have been keeping a close eye on the Development of SysML 2.0. But for years there has been little movement in the modeling tool market, and that gives me pause for thought. The serious modeling tools and the years of the first release are Rhapsody (1996), Cameo/MagicDraw (1998) and Enterprise Architect (2000).
Is MBSE becoming mainstream? Yes, but not yet
My personal conclusion: MBSE is currently experiencing a strong boost. Larger companies with complex products in particular see the long-term potential in MBSE and are prepared to invest. But this is only a small proportion of the teams that would benefit from MBSE. And for many reasons, the introduction of MBSE is associated with risks that smaller companies shy away from.#
If we use the Gartner Hype Cycle I believe that MBSE is still at the "peak of exaggerated expectations". It will be many years before we reach the "plateau of productivity".
So: I am convinced that MBSE will become mainstream at some point. But not in this decade.
Photo by 愚木混株 cdd20 on Unsplash
