An AUTOSAR-Specific Static Testing Strategy for Educational Automotive Software Engineering

AUTOSAR (AUTomotive Open System Architecture) is the leading architectural base in software development for the automotive field in Europe. However, because of its inherent complexity due to the existence of a multitude of different tools, specifications and architectural guidelines, it is complicated to get started with. As such, students in the field of automotive software engineering need guidance and means to assess their programming output in a fast and comprehensible way with regards to standard compliance. Based on a previously proposed educational cloud platform for automotive software engineering, in this paper we present an approach to assess the quality of AUTOSAR-based software developments. We position this approach as a step in the development process where implementations by learners are eventually run during test drives to evaluate their real world performance. In such testing infrastructures there is always a shortage of resources regarding test execution, since real test drives involve a rather complex preparation and schedule. But also in cases where software or hardware-in-the-loop simulations are employed, execution needs to be coordinated. Accordingly, the methods described in this paper have a twofold focus. On the one hand we want to catch methodical errors as soon as possible before more scarce resources (such as simulators or real test vehicles) are allotted. For this, we propose static testing methods specifically designed to assess AUTOSAR-based implementations. Consequently, we want to avoid sure-to-fail implementations to be actually run. On the other hand, the output of said methods should make it easier for learners and developers alike to assess and compare the Quality of their results.