MBSE in Practice
Model-based Systems Engineering (MBSE) is heavily discussed. Sometimes in flourish words, sometimes in PowerPoint. Sometimes presented as Silver Bullet. Yes, heading for digital transformation implies the application of MBSE. But where to start and how to use MBSE and benefit?
Let me try to give some answers and thereby focus on three different areas:
- Organization
- Methods
- IT
Lever Organization
Recently Oleg Shilovitsky posted a very good article on silos. This post and the resulting discussion made me think. Digital transformation is a real challenge for an organization. But before “scrumitizing” every process, one may take into account that mechanical, electric/electronic and software engineering are really different. Better to apply modern enterprise architecture methodologies, like TOGAF, first. Figure out necessary capabilities and define priorities in heatmaps. Related artifacts can be modeled in SysML, which helps people to understand the artifacts, their relations, impacts of changes through visualization and is computer-processable in parallel.
Thus, for driving forward MBSE-Readiness of your organization, you should:
- Elaborate capability maps
- Derive heat maps
- Specify and optimize process artifact models
Such models are models of your company, ready for monitoring and simulating alternatives. Especially regarding digital transformation, with its dynamics and uncertainties, the value of these models is high.
Lever Methods
Volkswagen Group in its Strategy 2025 recently stated, that in future cars will be digital devices on wheels. While reading this, I asked myself, how they want to develop the software in collaboration with suppliers, how they want to do virtual validation and verification, how they want to assure that a car is not losing the legal approval after a software update and how they want to bring in a security patch into millions of cars with completely different configurations?
Traditional tests in transportation where performed on real roads, tracks, water. Now, with autonomous transportation additionally to the
- system models
- environments models
are needed. And tests have to be scenario based rather than test case based.
If a product and its architecture, but also the environment is specified e.g. in SysML and linked specialized models (function, behavior, shape etc.), one can do systems engineering in the same way, as it is done in software engineering. Completeness, requirements fulfillment, regulatory compliance etc. can be check automatically. And changes can be steered very granular, because at any time one can evaluate the existing relationships, simulate alternatives etc.
And finally stated, if in future a vehicle is nothing else like a digital device, why not apply more software engineering methods. MBSE allows to model related concepts easily. But of course, to understand a product like a car, an airplane, a vessel, a train as a digital device and consider REST, Microservices and DevOps approaches within its engineering is something that feels very strange for many people.
Lever IT
In principle, two different categories could be considered:
- Tools and Tool Chains
- Standards and Formats
Here, a lot of research work is still necessary. Not without reason around the globe many projects are running to provide related solutions. Additionally, standardization of necessary formats is driven forward.
For example, prostep ivip‘s project Smart Systems Engineering formed a solid basis for the industrial application of functional models via FMI (Funktional Mockup Interface). The German research project PEGASUS provides necessary methods for autonomous driving on highways. And shipbuilding recently started the OSP (Open Simulation Platform) for providing a neutral platform for model-based simulation.
Regarding MBSE, a skillful integration of
- process artifact models and
- system artifact models
is essential.
Realizing this, open the way for traceability. Traceability gives answers regarding questions starting from a project’s maturity towards providing the relevant set of information to prove regulatory compliance.