Thesis - MBSE applied to Start Ups

Hi there!

I am starting this topic as part of my Master Thesis in Systems Engineering, I am studying “The Applicability of MBSE in New Space Start-Ups”. My end goal is to perform a Case Study of MBSE on a subsystem of one of these startups and compare/analyze the outcome/benefits with the daily troubles of their current SE solutions.

In that context, I first need to decide which language, method and software to use. Although Arcadia and Capella are already very good contenders simply from their open source nature, I would like to ask you, Capella community, for any user feedback not only on the tool itself, but also on the ease of use and integration into your organizations.
How easy is the learning curve?
How difficult was it to convince colleagues?
Where there any push-backs during your switch to MBSE?

I would particularly be interested in anyone that has anecdotes of using the Arcadia method in the context of a Start Up.
What pushed you to use Capella vs other software?
What convinced you that MBSE was necessary?

I would be more than happy to start a conversation in the comments bellow on your experiences and possibly discuss further via PM if need be.

Thank you for your help!

Hi!

Great timing, our company does a lot of work with start-ups in the medical device space and I have just started exploring using Capella to incorporate MBSE into our projects. I’m probably too early in the process to give you valuable feedback, but here are some highlights of my reasoning and experiences so far.

  • Startups can be pretty cost conscious when it comes to tooling, especially tooling related to documentation. So having something open source and cross platform like Capella reduces the barrier to adoption.
  • Just starting with the tool the learning curve feels pretty high. Arcadia/Capella has a lot of terms that mean specific things within the tool and figuring out that structure and which views would bring the most value to our projects has been a multi-day project with a lot of reading, video watching, and experimentation.
  • In terms of convincing others, I think the ability to present stakeholders with diagrams of the system and sub-systems has enormous value. People seem to be able to reason better about complex things when they are shown in diagram form over paragraphs and tables of text. Even just using Capella as a diagram creation tool may be enough to convince others of its value.
  • The big struggle for me right now is figuring out a low burden way to link requirements to the model. Requirements definition and traceability are big elements documenting the design history when developing medical devices and integrating that into the model may not be the best approach for maintaining those.

Happy to talk more offline if you want to PM me.

Thanks!

Great feedback!
A couple of comments:

  • beyond what’s available online for learning Capella (Capella MBSE Tool - Training Material as a start, but there’s obviously much more), one may consider professional training for accelerating the learning curve
  • In terms of convincing others, I agree and adding that a possibility is also to find a way to extract information from the model in a productive way, bringing direct value to the project, using add-ons like Python4Capella or M2Doc
  • Publication4Capella may be something to look at when it comes to traceability with Requirements

Stephane

1 Like

Hi
For many reasons I can’t share more than what I mention below, but what I can say is some organizations that are no more start up for long time are discovering MBSE. This is new and only part of the management is convinced, and then the organization is not willing to spend big money on that. More the current business processes set the start of engineering at product design, without real function design before. In this situation, use an open source tool keep the invest to spent time, as PoC at least. On the other hand, Capella has the big advantage to have in the box a SE profile that you can play with. It has many business processes include, which have not most of competition, driving you to build a complete administration to define all the business processes and the object your need in your SE tool (I remember other place having SE tool maker coming with hundreds people for years to set up the customer variant). Arcadia objects and processes are easily useable for material countable products, serial or unique, but as well for many other situations, the point to keep in mind is have PLM strategy to connect to designed product on the administrative side, and then find the tool to do it, and then have a PLM strategy for the model, and its elements. In my current example, Arcadia gives opportunity to set modes; variants and component partition in a designed and better controlled configuration, to improve design reuse, but this is still a target.
Thierry

1 Like

Hi @C.Hem,

FYI we had presentations quite similar to what you are looking for at last year’s Capella Days. I think the speakers will be delighted to exchange. Especially since, last time I checked, they are still in the field.
You can also find out more by taking a look at the MBSE2022 programme.

Samuel

Hi @akehrer

Thanks for the feedback!
I agree that the cost barrier is non negligible, but hope that any cost is justified by the added benefit that MBSE provides.

My main concern currently is that for established companies, MBSE linked to PLM is incredibly powerful and the discussion with the stakeholders and/or C-suite seems to be more or less a business pitch that always falls along the same lines.
For startups, however, I expect that there needs to be short term benefits for the concept to even be considered, and these probably change from one startup to the next, which makes convincing and assessing the worth quite challenging IMO. I’m thinking of focusing on accessibility to FMEA through MBSE (I believe some add-on tools from MapleSoft or Safety Architect will help me here). In my experience, and as seen through conversations, FMEA is always on the critical path for A&D startups as they begin testing and their licensing process, yet somehow is often left behind as the first years are highly focused on design development. (Feel free to react if you disagree!)

I’m afraid this goal is a bit too ambitious as Case Study objective, or even as a demonstration of feasibility, but hope I’m wrong!

Where your stakeholders at all familiar with Model or Systems Based engineering practices? The ease of convincing may be a discourse between professional fields, not sure.

@SamuelRochet

Thanks for the suggestions!
I have been down a INCOSE webinar rabbit-hole these last few days and have not had the time to look at the Capella Days yet, though a quick scan though the topics has already peeked my interest on 4 presentations, so thanks!

As for the MBSE2022, I was not even aware of this program! I did not expect ESA to have a program such as this, and will definitely check it out and any possible past events.

Input much appreciated!

Hi @C.Hem

I think you are correct in your observation of the differences between established company C-suites and those at start-ups. The vertical integration of information that these tools can provide, linking the models to the products going out the door, gives all layers of management something to drive as well as the visibility they need for whatever goals they have. But for startups the goal is almost always getting to revenue before the funding runs out (i.e. ASAP), so having the level of visibility that an integrated PLM system would provide is secondary or tertiary to that primary goal.

For starups I think the MBSE idea is more useful as a tool for getting buy-in from stakeholders at the frontend and using that to define the boundaries of what the system is which informs requirements and testing. Generally, once there’s agreement on the model it will probably get set aside while the “Real Work” gets done, unless there’s someone on the team who wants to maintain it as a side project.

Regarding FMEAs, I could have a whole separate discussion about where I think they are appropriate within the scope of system safety engineering and how MBSE could be useful there as well. But I agree, safety engineering/risk analysis is often an afterthought (“we have to check that box so let’s do something”) to the overall system design effort. Standards compliance can also be included in that afterthought group unless there is someone from the industry on the team who can speak convincingly about it.

Honestly, even if your thesis is just a high level overview of MBSE and its applicability within startups or early stage product development that would be beneficial. It would be a starting point for more discussion and begin to bring awareness to groups that may not know about it. Folks in this space are not aware of MBSE as a discipline. They’ll draw diagrams of their systems for their own purposed, but not within the context of system engineering or system modeling.

1 Like

There is an discord channel for Systems Engineers that I have been attending lately. There will be a presentation for the community next Friday, the 14th oct 2022, of using MBSE in the context of a design company. This might be of interest for your thesis, because its not a usual heavy engineering organisation stereotype… anyway, here is the link if you are interested: Systems Engineering Professionals

1 Like