Automotive ADLS

Dajsuren, Yanja; Brand, Mark van den; Serebrenik, Alexander; Huisman, R.G.M.

doi:10.1145/2304696.2304710

Cited by 10 publications

(1 citation statement)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While more and more software systems are built into automobiles, the construction of these systems suffers from the usage of heterogeneous modelling languages and tooling. Automotive architectural description languages [39] have been proposed as a method to model all relevant information of automotive systems. Domain-specific tools such as MATLAB/Simulink are, however, often required in the development process nevertheless, since they offer specific analyses of the system under development.…”

Section: Case Study: Automotive Systems Engineeringmentioning

confidence: 99%

Flexible views for view-based model-driven development

Bürger¹

2013

Proceedings of the 18th International Doctoral Symposium on Components and Architecture

View full text Add to dashboard Cite

Modern software development faces growing size and complexity of the systems that are being developed. To cope with this complexity, several languages and modelling formalisms are used during the development of a system in order to describe it from various view points and at multiple levels of abstraction. These languages and modeling formalisms are specific to the domain of the system under development, to the paradigms that are followed, and to the developer tools being used. For example, in a component-based process, the software architecture may be defined using a component metamodel, while class diagrams are used for the object-oriented design of the system; performance and reliability models express the extra-functional properties; program code in a general-purpose programming language defines the execution semantics, and is used for the implementation of the system. Although all these artefacts follow different concepts and formalisms, they express the same system from various view points and can thus be understood as views on the system under development. The entirety of these views can be identified as a complete definition of the system. Although the usage of multiple formalisms offers the advantage that different developers can describe the system from different view points, and in languages and models that are specially designed for this purpose, it introduces the problem of fragmentation of information across heterogeneous artefacts in different formats, concepts, and languages. With increasing complexity of the systems, this makes navigation through the system description difficult for developers and other roles in the development process. Furthermore, since the concepts and formalisms are managed independently, they can share semantic overlaps, so that a piece of information about the system under i Abstract development can be expressed in several ways and contexts. For example, while a component model may be used to express the prescriptive software architecture of a system, the implementation in a general-purpose programming language also determines an implicit architecture of the system, which leads to redundancies in the system descriptions. The independent evolution of these artefacts can lead to inconsistencies, which cannot be identified automatically if the semantic correspondances between the artefacts are not modelled explicitly.View-based software development processes offer two fundamental principles of adressing this problem: While synthetic approaches integrate the information of heterogeneous models to form a complete system description, projective approaches introduce a common formalism in which all view points of the system can be represented, and generate the views from this central model. While the pure synthetic approach suffers from a quadratic increase of interdependencies between the view types, which become very numerous with a growing number of formalisms that have to be supported, the projective approach faces the problem of providing a common formalism that is ab...

show abstract

Section: Case Study: Automotive Systems Engineeringmentioning

confidence: 99%