Software

Release 2023 R1: Vehicle – ADAS/AD

Updates and improvements to AVL’s simulation solution

The automotive industry is in a constant state of change and along with it your needs. To meet these needs, we at AVL are permanently working on the further development of our products and services to support you in fulfilling your daily tasks.

The latest release of our simulation solution for development and virtual validation of automated and autonomous driving functions includes improved OpenSCENARIO standard coverage by AVL Scenario Designer™ and high-performance Simulation Platform powered by Model.CONNECT™.

AVL Scenario Designer™ - Road Builder Add-On and Local Road Library

With the new Road Builder Add-On, users can quickly generate road segments in ASAM OpenDRIVE format. No tiresome manual editing needed – just choose from a list of available templates like curvy rural roads or highways with ramps. All templates are parametrizable allowing for a variation of road and scenario at the same time. This opens up limitless possibilities such as making scenarios suitable for different countries.

The Local Road Library is a set of OpenDRIVE road networks that come automatically with the Scenario Designer installation. Boost your scenario creation capabilities by choosing a suitable road network instead of manually creating new roads.

Vehicle – ADAS/AD Release R1 2023 - Figure 1: OpenDRIVE roads networks in Scenario Designer

Figure 1: OpenDRIVE roads networks in AVL Scenario Designer™

Esmini Interface for High-Performance ADAS/AD Controls Testing – Even on Your Notebook

When testing the control algorithms of automated driving systems, an object-based representation of the environment is sufficient. The focus is on decision making, path planning, and motion control. Esmini is a lightweight open-source tool that can execute scenarios in ASAM OpenSCENARIO standard with a very high performance. By integrating esmini into the Model.CONNECT co-simulation platform, you can easily setup your closed-loop test setup and define parameter variations. The built-in Job Management System (JMS) executes all the tests on your local machine. The optimal number of parallel simulations depends on your hardware.

The simulations can be massively scaled in our distributed cloud computation cluster allowing you to execute millions of test cases overnight.

Frequent testing is a crucial part of a successful agile development process for ADAS/AD systems. Using our Python API, you can easily integrate and fully automate the test execution in you CI/CD pipelines.

Figure 2: Performance benchmark of an Adaptive Cruise Control System (ACC) test setup executed on a normal notebook