I’m relatively new to Capella, and currently working at the SA level of system that’s designed to be integrated into a vehicle.
The scenario: I have 3 human actors that interact with the system - “driver”, “manager”, “service technician”. They have common functions, e.g., they can all monitor the system warning messages and supress them if the wish, or drive the vehicle. However “manager” and “service technician” have unique functions. “Service technician” will have the highest level of access to the system to service it, change system parameters etc. “Manager” can set system settings etc. But all 3 actors use the same system HMI
Question: is there a better way than to have 3 actors represented separately, given their tiered level of access to the system, as I now have 3 separate component exchanges “System HMI” to the system that are all actually the same thing in reality?
I am also new to Capella and have this exact situation.
The answer is to define an Operational Role (the purple ones) of “System Supervisor” or some such name and then allocate the monitor system parameters HMI operational activities to that Role.
A role can then be allocated to multiple actors and multiple roles can be added to each actor.
Hi, Thanks for the quick response! I have done some role allocation at operational level - how do the roles then transition into system analysis level?
I’ve just run an experiment, and it seems that at the system analysis level the description of role is no longer required.
Once you do your automatic translation from OA to SA, you can create a system architecture diagram, this will allow you to add the system actors and their corresponding allocated functions (that have been pre-populated). When you do this you will see that all the actors that implemented an operational role will have the system functional equivalent of the contained operational activities available to them.
Further digging has shown via the properties editor that it is not creating multiple versions of the system function as they have the same Unique ID.
So the SA layer allows System Actors to share System functions, ie more than one actor can provide a system function.
It appears then as the actors are separate at SA, its normal for the component exchanges to be replicated. So in this case, I’ll have the 3 actors, with their functions implemented by 3 component exchanges that are “HMI”, even though in reality the system will likey have 1 HMI, not 3. Do I need to make use of the REC/RPL functionality for the component exchanges?
I Just had a thought. Does it even matter, i.e., in this case do I even need different actors at the system level? Ultimately they are all types of operator, and they will all be interacting via the same HMI. We just need to show that the system has the functionality to deal with different access levels from the operator actor. What value does differentiating the operator actors at system level have - on reflection - I’m not seeing the value in representing them separately…
For me it is a matter of completeness and clarity.
The actors justify the need for a system function and the need for the inclusion of HMI rather than just the need to report information as a pure data exchange.
Separating the different types of actors will also help when you are completing your scenarios that show who does what and when.
However, if the system is still described correctly by just having one entity/actor as System Operator that contains all the actions an operator can make, and the fact that all actions won’t necessarily be completed by a single user or at any given time is not important to you, then it is still a valid approach.
I think that is the beauty of Capella, you can use the various diagrams to your advantage to help explain the system to internal and external stakeholders, so if it works for you, correctly describes the system, and is a “validated diagram” in Capella, it is all good!