Improving the Model-Based Systems Engineering Process

Abstract

Modern embedded software systems are becoming more and more complex. Engineering embedded systems raise specific challenges that are rarely present in other software engineering disciplines due to the systems’ steady interactions with their environment. Research and industry often describe embedded systems as component and connector models (C&C). C&C models describe the logical architecture by focusing on software features and their logical communications. In C&C models, hierarchical decomposed components encapsulate features, and connectors model the data flow between components via typed ports. As extra-functional properties, for example, safety and security, are also key features of embedded systems, C&C models are mostly enriched with them. However, the process to develop, understand, validate, and maintain large C&C models for complex embedded software is onerous, time consuming, and cost intensive. Hence, the aim of this chapter is to support the automotive software engineer with: (i) automatic consistency checks of large C&C models, (ii) automatic verification of C&C models against design decisions, (iii) tracing and navigating between design and implementation models, (iv) finding structural inconsistencies during model evolution, (v) presenting a flexible approach to define different extra-functional properties for C&C models, and (vi) providing a framework to formalize constraints on C&C models for extra-functional properties for automatic consistency checks.