A quantitative study on automatic validation of the diagnostic services of Electronic Control Units

As a result of the strong competition on the world car market the development cycles are decreasing while the complexity of todaypsilas E/E vehicle architectures is growing. This requires the introduction of new methods to speed up the development process and guarantee correct functionality of the deployed Electronic Control Units (ECUs). Especially, in the context of the diagnostic functionality provided by the ECUs the correctness of the services is significant, as the information provided by the built-in diagnostic subsystem shall guide the mechanic at the service station. This way high warranty costs due to unnecessary component replacements can be avoided and effective problem solving guaranteed. In this paper we introduce CANoe Option DiVa (Diagnostic Integration and Validation Assistant) as a new tool for diagnostic validation that closes an open gap in the existing tool chain used at General Motors Europe. We thereby establish fully automatic test case generation for the first time in the context of the Opel Insignia development. Based on the machine readable diagnostic specifications for Opel Insignia ECUs, DiVa generates test code to be executed within the CANoe test environment. In order to elaborate on the economic benefits of this new validation method we compare the efforts for automatic testing with manual testing and provide details on the resulting speed-up factor on the basis of a series production project.