Systems of Systems

Hi - I am trying to work out the best way to transition from the Operational view to a Systems View which supports potential Constituent Systems (perhaps legacy) within the System of Interest. As far as I can tell the automated transition generates a single “System” - one assumes the SoI - with the ability to only add actors outside the System. I cannot add the equivalent of Components (eg Component Systems) until I get to the Logical level. This seems to be a limitation because I would like to be able to assign SFs to different components of the SoI. Perhaps I am misunderstanding the meaning of the System level of analysis?

You are certainly not wrong but Capella’s view is a little different here I think.
In Capella at [SA] you only have the SYSTEM (and actors around) - no components internally inside the system - as you already noticed.
In your case you should go down to [LA] or even [PA] where your SoI are represented by components. If you want to analyse these “components” in detail you can make a SystemToSubsystem-Transition [LA/PA]=>SA(AddOn, see download area) of each of your SoI. You will get a new Capella-SA (= project) for all your SoI.

Richard is right, in Arcadia and Capella, each perspective (OA, SA, LA, PA, [P]BS) has its own role. Notably, OA and SA are expected to capture the need that the solution should satisfy (and only the need), while LA, PA and BS describe the designed solution.
So SA, here, focuses on a “blackbox description” of the system, where you just want to capture the capabilities expected from the system, the use cases describing how it is to be used (via scenarios functional chains, states and modes…), and the “services” that it has to fulfill in these use cases (as System Functions and functional exchanges). This is not far from the ‘Service-oriented Views’ in architecture frameworks such as NAF or DoDAF, for example.
From a user need point of view, the way the system is decomposed (components) is of little interest, and there may exist different alternatives of decomposition for a same need (which are design decisions, not need description). This is why SA is clearly distinct from LA, and why there is no component inside the system in SA.

Jean-Luc and Richard
Thanks for your prompt response - I think I understand the nature of the SA view better now ( I suppose the overloading of the word “System” does not help in this regard - it made me think of MODAF SVs
The add-on seems very interesting in this regard.

Copyright © Eclipse Capella, the Eclipse Capella logo, Eclipse and the Eclipse logo are Trademarks of The Eclipse Foundation.