Evaluating Modeling Solutions on Their Ability to Support the Partitioning of Automotive Embedded Systems

A pool of competing modeling solutions have been proposed to cope with the problems induced by the growing complexity of automotive embedded systems, i.e. the E/E (Electric/Electronic) systems of automobiles. As the principal features of these solutions are axed around modularization and high level of abstraction, it is necessary to investigate their ability to support the implementation. An objective evaluation will be helpful to define how each modeling technique should be enhanced for a better support of the implementation, whenever necessary. This paper defines a framework to evaluate the capacity of modeling techniques to support the implementation in the automotive engineering domain, particularly the partitioning. Following the state-of-the-art in the partitioning of automotive embedded systems, we present the evaluation framework. Then, we introduce the most common modeling solutions used in the automotive embedded systems design and we use the framework to evaluate and classify them.