ThyssenKrupp Presta Hungary discusses the main challenges in applying a MBD approach

Automotive IQ sat down with András Balogh, Head of Platform and Software Development at ThyssenKrupp Presta Hungary, and talked about key issues in applying a MBD approach, ahead of the conference where Mr. Balogh also will be presenting .

Please tell us a bit about your current projects involving MBD at thyssenkrupp

We are building power assisted steering systems for passanger vehicles. This one one hand involves AUTOSAR, the standard architecture modeling language for in-vehicle SW, and for our sophisticated steering control algorithms we are using Matlab/Simulink as the core modeling environment.

How long have you been applying MBD and what improvements have you seen because of it (for example based on statistics, case studies measuring improvement, etc.)?

We are using MBD from the start of our development (back in 1999). We started with Matlab/Simulink, and after that we also introduced AUTOSAR in the architecture design. With the later, we experienced a significant boost in our productivity, because of the clear architecture and the automated generation of glue code.

What have been the main challenges in applying a MBD approach and how did you go about solving them?

Some tools were not really user friendly at the first time, and there was a strong limitation in scalability (in terms of model size). We did a lot of work to find the right tools and sometimes also to implement our own ones to achieve the right solution.

One of the main challenges in this industry seems to be the lack of a standardized approach to MBD. Do you think this is necessary and if yes, will we see something resembling it in the near future?

It would be really nice to have a single standard for several application fields, but at the moment I do not see any strong candidates for them. There are standards, however for a single field (like AUTOSAR for automotive), but they are also not covering all aspects of the problem domain.

I am not convinced that the single size fits all approach would be possible in this area.

What advice would you give to a company who is just starting to implement MBD for their systems, in order have a smooth transition?

I think MBD can be mastered only with lot of trials and fails. I would propose to start in small, experiment with many approaches and learn from the fails of others. I do not think that any consultant would have a silver bullet solution that works out of the box for all companies, so the key is to customize and combine existing solutions.

What are you presenting at the conference this year and why is this significant to the industry?

I plan to talk about our challenges in introducing, maintaining and scaling an MBD solution as a Tier-1 supplier. As we have many projects and several product lines in parallel, this is especially difficult if we also want to use all possible synergies between projects, and reuse whatever is available.

The main take-away of my presentation will be the open approach we introduced in order to streamline our development using MBD tools, and how we introduced MBD, continuous integration, and continuous testing in a large SW development organization.