The SET Level project optimizes test methods for developing autonomous vehicles
A total of 20 partners from the worlds of business and science are working together to make autonomous driving in the city a reality.
Focus on autonomous driving, simulation-based testing
Braunschweig/Paderborn. April 29, 2021. Before autonomous vehicles are let loose on our streets, we must be able to guarantee that they are safe. This requires numerous tests in a simulated environment since not all situations can be tested in real traffic.
In the SET Level project, dSPACE is working on an efficient simulation technology together with 19 partners from the worlds of business and science. The objective: to develop a type of simulation that can be used flexibly for different applications and stages in the vehicle development process and that can transfer a substantial share of road tests to a simulated environment.
“It is only possible to launch autonomous vehicles on the road with powerful, simulation-based tools and methods that can be used throughout the development process and within the context of the approval process,” says Prof. Frank Köster, founding director of the German Aerospace Center (DLR) Institute for AI Safety and one of the two coordinators of the SET Level project. “These tools and methods round off our comprehensive portfolio of test and inspection tools, such as test benches and test sites, which will make it possible to test autonomous vehicles to the extent necessary in the future.”
An autonomous vehicle must be able to operate reliably in any situation. Testing this on the street would take years. Therefore, being able to perform the tests efficiently in simulation is decisive for securing the approval. The SET Level project partners are developing a method that can be used to render critical traffic situations completely digital, thus making it possible to simulate them. This will significantly reduce the future effort involved in approving autonomous vehicles.
One unique aspect of the SET Level project is its close connection to open standards and simulation tools, which means the SET Level method can be used by many different companies and research institutions and even developed further after the project is over.
“The aim of our work is to achieve solutions and standardizations that can actually be used in practice. This is the only way to create an open, flexible, and expandable platform that supports simulation-based validation and testing for autonomous vehicles in complex traffic situations,” explains the second coordinator of SET Level, Dr. Stefan Rude, who is responsible for virtual validation in the autonomous vehicles division at BMW.
These practicable solutions will be presented to an international audience of experts using three simulated traffic scenarios on April 29, 2021, at the project’s virtual half-time event, where digital renderings of vehicles will be simulated as part of various traffic situations based on a wide range of different criteria in order to draw conclusions about the reliability of the sensors and algorithms. At the event, the researchers will focus on the test methods and simulation tools and how they interact with the software components.
The Federal Ministry of Economics and Energy (BMWi) is funding the SET Level project with a total budget of more than EUR 30 million. It is a continuation of the joint PEGASUS project, which was completed in May 2019. In this project, the German Aerospace Center and its partners developed quality standards and methods for validating highly automated vehicles so that they can be operated reliably on our streets in the future. While the PEGASUS project focused on researching highway-related issues, the SET Level project is directed toward the development of simulation platforms that can be used to test autonomous and connected driving functions in urban areas in an economical and flexible manner. SET Level expands on the simulation approaches laid out in PEGASUS on a large scale and extends its application to the entire traffic area.
dSPACE is able to contribute its expertise on analyzing requirements, simulation-based testing, developing models, and instantiating tool chains in the project, while also supporting the consortium with its proficiency in software-in-the-loop (SIL) and hardware-in-the-loop (HIL) testing. Furthermore, dSPACE has made its VEOS integration and simulation platform available for the project, along with ModelDesk for modeling scenarios and MotionDesk for 3-D animation.
The partners involved in the SET Level project: ADC Automotive Distance Control Systems GmbH (part of Continental AG), Audi AG, BMW AG, dSPACE GmbH, the DLR Institute of Transportation Systems, ETAS GmbH, Ford-Werke GmbH, Fraunhofer LBF, the FZI Research Center for Information Technology at KIT, IPG Automotive GmbH, MAN Truck & Bus AG, the OFFIS - Institute for Information Technology, Opel Automobile GmbH, PROSTEP AG, Robert Bosch GmbH, RWTH Aachen (Institute of Automotive Engineering), TU Braunschweig (Institute of Controlling Engineering), TU Darmstadt (Institute of Automotive Engineering), Volkswagen AG, ZF Friedrichshafen AG.
The project will run until August 2022. Funding is being provided as part of the BMWi’s New Vehicle and System Technology program.
Subscribe to our newsletters, or manage or delete your subscriptions