Keynote Lectures

Future Directions for MBSE with SysML v2
Sanford Friedenthal, SAF Consulting, LLC, United States, United States

Responsible Language Design
Vadim Zaytsev, University of Twente, Netherlands, Netherlands

On Views, Diagrams, Programs, Animations, and Other Models
Henderik A. Proper, TU Wien, Vienna, Austria, Austria, Austria

Abstract
Model-based systems engineering (MBSE) has evolved over the last several years in response to the need to deal with growing system complexity and increased enterprise agility. MBSE is an approach to systems engineering where information about the system is captured in a system model. This approach can provide a more complete, consistent, and traceable system design than a more traditional document-based approach where information about the system is captured in a variety of documents, informal diagrams, and spreadsheets.
SysML v1 was adopted in 2006 and has been a key enabler of MBSE. Since that time, much has been learned about applying MBSE with SysML. The next generation of SysML (v2) is being developed by the SysML v2 Submission Team (SST) to provide capabilities that address the limitations of SysML v1 and enable the evolving practice of MBSE. This presentation summarizes the future directions for MBSE and how SysML v2 can support these needs.

Abstract
Software language design has become commoditised: there are many techniques, tools and workbenches which enable anyone to analyse a new domain and distil it into a language. Domain jargon transforms into concrete syntax, domain concepts form an abstract syntax, familiar visualisations become visual layout, domain knowledge morphs into semantic rules of correctness, conformance and execution behaviour. The burden of writing parsers manually or even having to optimise the desired syntax for the sake of language processor's performance, are seen as legacy of the past, and start to look irrelevant or insignificant nowadays. However, one burden of a language designer that remains, is the responsibility for the consequences of design choices. In the field of design, anything is possible, but everything has consequences. In the field of language design, we shape the ways future language users will think, and should take responsibility for that.
In this talk, we will revisit some of the existing languages for programming and modelling in different domains, to gain insight into the rationales behind some of their design decisions, as well as the consequences of those decisions which we need to deal with today. Learning from the past will prepare us for the future.

Abstract
In the context of model-driven systems engineering, we come across notions, such as views, aspects, perspectives, diagrams, etc. Next to that, the question if a programme is a model, or if an animation is a model, also regularly surfaces. The aim of this keynote is to take a fundamental look at these notions, while essentially positioning these notions as being specific kinds of models, albeit for fundamentally different purposes.
We start by zooming in on the notions of model and modeling language. In doing so, we will also visit the importance of a model’s purpose and its potential Return on Modeling Effort (RoME) in particular, as well as the role of the conceptualisation — in the mind of the modeler(s) — of the domain of interest that is (to be) captured by the model. This will then also take us to the notion of conceptual model, which is a class of models that has grown to play an important role in the field of information systems engineering. Identifying conceptual models as a distinct class of models, does automatically suggests there to be a class of models that are not conceptual. These “non-conceptual” models have an important role to play as well in the sense that by allowing for “conceptual compromises”, these models may gain other experiential and/or computational benefits (RoME), such as being executable, animatable, tangible, etc. .
Based on this discussion, we then position notions such as views (based on different perspectives or aspects), diagrams, programs, and animations, as essentially being specific kinds of models, covering different purposes, different audiences, while also spanning between conceptual and non-conceptual models.