Fraunhofer Institute for Structural Durability and System Reliability LBFDevelopment of test procedures, systematics and methods
VVMethods - Analysis and test procedures for autonomous vehicles
The development of test procedures and systematics as well as the provision of methods to provide proof of safety for automated vehicles were the topics of the BMWK lead project “VVMethods”. The partners worked on the use case of an urban intersection with a focus on driving functions up to the complete automation of vehicles (SAE Level 4 and 5).
Figure 1: Process steps of functional test planning for autonomous vehicles.
Figure 2: Combination of probFMEA with component fault trees as a new, efficient method for assessing the safety of autonomous vehicles.
Test planning
The LBF played a leading role in the development of a systematic procedure for deriving functional test specifications for autonomous vehicles. Figure 1 shows the process steps of functional test planning for autonomous vehicles with their input and output variables.
Probabilistic FMEA
By combining the probabilistic failure mode and effects analysis (probFMEA) developed at LBF with IESE's component fault tree method and its implementation in the "safeTbox" software tool, an analysis method for evaluating functional safety was developed that can handle the volume of data and the complexity of the systems. Figure 2 shows how a new, efficient method for assessing the safety of autonomous vehicles was developed by combining probFMEA with component fault trees (CFT). By combining a bottom-up analysis (probFMEA) with the top-down approach of component fault trees, a better system coverage is achieved, among other things.
Camera-in-the-loop tests
Together with the FZI Research Center for Information Technology, the LBF realized a test bench to investigate the robustness of a camera-based object recognition system under vibration load. The multi-axis vibrations are calculated using a dynamic vehicle model, considering the tire-road interaction. A multi-axis vibration exciter and active vibration control enable the simulation of vertical and rotational movements of the camera mounting position up to approx. 200 Hz (see Figure 3). It is determined how the imposed accelerations, which were previously recorded, affect a test camera to evaluate the accuracy of object recognition.
Figure 3: Application of the recorded acceleration to a test camera to evaluate the accuracy of object detection.
Funding and partners
- Federal Ministry for Economic Affairs and Climate Action
- 22 partners from industry and science
- Consortium management: Robert Bosch GmbH and BMW AG
Privacy warning
With the click on the play button an external video from www.youtube.com is loaded and started. Your data is possible transferred and stored to third party. Do not start the video if you disagree. Find more about the youtube privacy statement under the following link: https://policies.google.com/privacyDesign & Test Planning, Presentation at the final event of the project, Jürgen Nuffer (german)