Software Engineering

Result & Added Value

INVENSITY helps a Tier 1 automotive supplier efficiently and automatically test and validate each software release through newly developed automated test cases. The debugger and testing tools were automated to drive and evaluate.

Initial situation

Whether it’s sunshine or heavy rain, it’s hard to imagine safe and comfortable driving without a functioning air conditioning system.

Intelligent software is required to set the air temperature and airflow comfortably for the driver. This software must be tested and validated with each new release. To do this, various test cases must be performed with a debugger and the CANoe tool, and the results subsequently evaluated.

Methodology

For this purpose, the existing requirements of the test cases and the individual software tools were first analyzed. Then INVENSITY automated the diagnostic and debugger functions. Thereby, e.g., stack, tasks, runnables, DET, or diagnostic functions could be controlled, measured, and evaluated. Finally, it was possible to import the test results into DOORS in an automated way.

By implementing the automation, INVENSITY realized the desired increase in effectiveness and the resulting cost reduction for the customer.

Results & Added Value

INVENSITY was able to increase the efficiency of testing and thus save test engineer’s time through HiL simulation for Tier 1 automotive supplier.

At a glance

Our team developed a SiL simulation for the development of virtual tire tracks in the camera image of an automobile. The SiL simulation enables a shortened development time of 6 weeks. Instead of blocking a test vehicle for 100 hours for the test phase, in the future the complete test phase can be simulated within 1 week. This efficiency gain brings corresponding cost savings.

The Story

A large number of newly registered vehicles have cameras that show the driver the immediate surroundings of the vehicle. In addition, the vehicle’s tire tracks are projected into the camera image. This enables the driver to track his steering movement. The projection of the correct tire tracks is of central importance. There was a discrepancy here at the start of the project: The simulation of the tire tracks felt subjectively wrong.

For the revision, however, it was not clear at the beginning which parameters had to be changed for this. First, all image frames of the camera were re-recorded and the previous overlays of the projections were removed. Then the full deflection of the steering wheel could be simulated, and the overlays inserted into the camera image. In this process, the parameters were successively adjusted until the projections matched the steering angle of the steering wheel and the subjective feeling. On this basis, the parameters were further adjusted until the entire steering angle of the steering wheel matched the projections and the feeling of the driver.

After a year of development, the software was successfully tested and deployed. These were successfully integrated in a SiL simulator.

By integrating the projections with the help of the SiL simulator, a test vehicle will no longer be necessary in the future. As a result, test drives of over 100 hours no longer have to be performed on the real vehicle, but instead cost-effectively with the SiL environment. The development time could also be reduced to 6 weeks.