Simulating Multiple Scenarios for Autonomous Driving

The Task

The steadily growing number of functions for automated driving requires a powerful test environment that can handle the exponentially increasing effort for the tests in early development phases. To achieve this with utmost efficiency, a simulation-based evaluation of vehicle software for functions for highly automated driving and self-driving cars is a necessity.

The Challenge

Comprehensive tests of functions for automated driving require:

  • The automated simulation of different traffic scenarios on a wide range of road networks and with different environmental conditions
  • A large number of tests to cover scenario variations and randomized testing
  • The simulation of millions of test kilometers
  • Consistent management of test parameters and a result analysis
  • The reuse of tests in all development phases

The Solution

Multiple instances of ASM Traffic are simulated in parallel on a VEOS cluster.

The seamless dSPACE tool chain opens up powerful and efficient possibilities for realistic simulations of autonomous driving. These include:

  • The graphical definition of road networks and traffic scenarios in ModelDesk
  • The simulation of many different traffic scenarios comprising a multitude of road users with ASM Traffic
  • Testing multiple traffic scenarios by means of PC clusters with a high number of VEOS® instances for parallel execution, which reduces the time required for testing to a minimum
  • Data management and traceability via dSPACE SYNECT®
  • The seamless reuse of tests for MIL, SIL and HIL simulation on platforms like PC, VEOS, and SCALEXIO 

More Information Success Stories

Subscribe newsletter

Subscribe to our newsletters, or manage or delete your subscriptions