DSPACE: EFFICIENTLY PROVE THE SAFETY OF AD FUNCTIONS IN AUTOMATED ENDURANCE TESTS

At the Automotive Testing Expo (Novi, Michigan, October 24-26), dSPACE is demonstrating in a joint showcase with AAI how to minimize residual risks when developing functions for autonomous driving and how to provide evidence of safe system behaviour in known environments. To do this, dSPACE has combined AutoScenarios, an AI-based application from the AAI ReplicaR platform, with its Automotive Simulation Models (ASM) and AURELION, the software solution for sensor simulation and visualization.

The validation of functions for highly automated driving is elaborate and complex. In order to provide proof of safety in accordance with SOTIF (Safety of the Intended Functionality), several hundred thousand scenarios must be tested. With the joint showcase, dSPACE and AAI demonstrate how to maximize test coverage and thus safety proof, especially for unknown risks: Using an intuitive interface, AAI AutoScenarios provides the dSPACE simulations with huge amounts and variations of AI- and random-based realistic traffic situations with many road users, to which the test vehicle (SUT) needs to respond.

"With this solution, we are demonstrating how driving functions can be tested and validated easily, quickly, and efficiently in automatically generated endurance tests," explains Dr. Hagen Haupt, Director Simulation Models & Scenarios at dSPACE.