No abstract

Systems-of-Systems (SoS) are complex systems composed of managerially and operationally independent constituent systems (CS). Smart cities are examples of SoS. These types of systems impose challenges to the traditional software architecture design, such as describing heterogeneous CS that are constantly evolving and identifying emergent behaviors from the interactions of those CS. Executable models (ExM) have been envisioned as a possible solution to deal with the challenges raised by SoS architectural design. However, to the best of our knowledge, no systematic mapping study (SMS) exists that investigate current state of art in the research area of SoS architectual design using ExM. Therefore, the main goal of this study is to synthesize evidence of that research area, observing research trends and identifying possible research gaps yet to be explored. Results reveal that (i) the use of model transformation from a static SoS domain-specific language to a ExM is intensely explored and (ii) ExM were used mainly to evaluate and synthesize SoS architectures but also to observe emergent behaviors and to measure quality attributes.

Systems of Systems (SoS) are complex systems composed of managerially and operationally independent constituent systems (CS). Smart cities are examples of SoS. However, these types of systems impose challenges to traditional software architecture such as highly evolutionary architecture due to the evolution of individual CS and emergent behaviors that results from the interoperability of CS. Executable Models (ExM) are a class of models that can be executed and that can assist on architectural design of SoS. By using them, architects can predict the SoS structure and behavior by visualizing and simulating the SoS still at design-time. On the other hand, as any other emergent technologies, it suffers with absent scientific evidences of its benefits, mainly in industrial context. The main contribution of this master thesis project intends to be the presentation of evidences about the use of ExM to solve problems in the SoS software architecture design. For achieving such purpose, this project is structured in well-defined steps: (i) a systematic mapping study, (ii) elaboration of a conceptual map derived from the mapping study, (iii) a survey to obtain perceptions from software engineering professionals on the use of ExM in the context of engineering simple systems and (iv) a new survey expanding the previous study and to obtain perceptions from software engineering professionals on the use of ExM in the context of architectural design of SoS. Preliminary results reveal that ExM have been reported in diverse SoS domains, offering a dynamic and interactive view to the SoS. We also found that software engineering professionals see ExM as suitable solution for dealing with complex and critical systems as SoS.

Contemporary complex systems often exhibit dynamic structures and behaviors, several components/systems involved, and multiple interoperability links. Those systems have been exposed to fragilities of traditional software specification languages (e.g. UML and SySML), since such languages were designed to document single (not multiple interoperating) systems. Those limitations can potentially further compromise the quality of the final software product. In this context, Executable Models (ExM) technology, such as simulation models, models@runtime and executable UML, satisfy the aforementioned requirements by supporting engineers with visualization of the system structures (still at design-time) and the ability to exercise their behaviors and interactions. In our prior study, we presented the results of an exploratory study on the perceptions of those professionals (from both industry and academia) regarding the use of ExM to solve problems in their current practice. We exposed 58 professionals (researchers and practitioners) to situations to solve problems using a specific type of ExM (DEVS simulation models), based on survey research. Responses were quantitatively and qualitatively analyzed. In this article, we extended the obtained results by analyzing and compiling a list of strategies to improve ExM notations to better address the needs of software engineering professionals. Later, we assessed those strategies with software engineering researchers to confirm the importance of the proposed strategies. Results revealed that executable languages still require advances to bring them even closer to the current software engineering practice and towards a more significant adoption in the future. The proposed strategies focus on improvements on the robustness of the ExM notations, visual representation of the models, the usability of the models, and user support.

No abstract