TÜV SÜD Mobility collaborates with NVIDIA and AVL to standardize validation of autonomous vehicles with NVIDIA DRIVE constellation™
Cooperation Will Leverage Large-Scale Hardware-in-the-Loop Simulation to Validate Autonomous Vehicles through TÜV SÜD Scenario-Based Testing
Last year, TÜV SÜD launched a cooperation with NVIDIA and AVL GmbH to develop a certification process for highly/fully automated vehicles. The goal of this collaboration is to validate and establish simulation as a method and approval tool for the digital homologation of the future. With his presentation at GPU Technology Conference (GTC) in Silicon Valley this week, Dr. Houssem Abdellatif, Head of Automated Driving and Driver Assistance Systems at TÜV SÜD Mobility, offers a detailed look at the development scenario of digital homologation.
Simulating traffic situations plays a crucial role in homologating highly/fully automated vehicles due to the sheer volume of situations that need to be tested. TÜV SÜD experts estimate there are up to 100 million situations per fully automated driving function to determine self-driving capability. The time and costs to perform traditional on-road tests to evaluate these critical scenarios are no longer practical. In future digital homologation, virtual procedures will augment established methods. Performing simulation testing at the necessary scale requires enormous computing capacity.
Dr. Abdellatif on partnership: “In the future we need simulation for the validation and approval of autonomous vehicles, since physical testing is limited and not scalable. Therefore, TÜV SÜD is looking for simulation tools that are trustworthy, robust, and scalable for the approval of autonomous vehicles. NVIDIA DRIVE Constellation™ provides a powerful and highly scalable solution.”
TÜV SÜD, NVIDIA and AVL are working together to validate and establish simulation as a tool for developing standards and are actively seeking solutions for the homologation of vehicles. TÜV SÜD assesses whether all specifications for homologation and authorization are complied with when using digital testing equipment. Their experts are responsible for the definition of safety requirements, exemplary critical driving scenarios, and the necessary evaluation criteria, along with the data analysis and outcome report that serve as a basis for digital homologation.
“Only by qualifying the simulation can the results be used reliably for type approval,” explains Christian Gnandt, head of virtual methods at TÜV SÜD. AVL combines the real and virtual worlds thereby builds the bridge to testing real vehicles by running them on a test bench for autonomous driving called AVL DRIVINGCUBE™.
Zvi Greenstein, general manager at NVIDIA, notes, “The standardization of autonomous vehicle validation led by TÜV SÜD is an important initiative and is supported by NVIDIA.” Digital homologation is also a hot topic at the GPU Technology Conference NVIDIA GTC. In their presentation at GTC, “A Rocky Road to ADAS/AD” Dr. Abdellatif and project partner AVL describe the steps leading to the authorization of fully automated vehicles. They take a discerning look at the topics defining critical scenarios, the connection between physical and virtual testing, and functional safety, with an eye to international regulations for homologation. “Together we face the challenge of how digital homologation can look and be implemented in the future,” says Dr. Abdellatif.
From the beginning, TÜV SÜD has been the partner for safety in the development of highly automated driving in diverse fields. In the German Federal Ministry for Economic Affairs and Energy’s PEGASUS project, for example, experts are working with sixteen partners from industry and science to formulate requirements for methods and tools intended to make highly automated driving functions safe. As a member of the German Federal Ministry of Transport’s special committee on Driver Assistance Systems, TÜV SÜD experts are helping to revise the relevant standards.
Together with the German Research Center for Artificial Intelligence (DFKI), TÜV SÜD is developing a ‘TÜV for algorithms’ and operating a testing ground for companies in collaboration with the testing-ground operator for autonomous driving in the federal state of Baden-Württemberg and the Karlsruhe Transport Association. It is also involved in many international development projects on automated driving – such as CETRAN (Center of Excellence for Testing & Research of AVs at Nanyang Technology University in Singapore) and is participating in the world’s first standard for authorization of fully automated vehicles in Singapore.
Press contact: Vincenzo Lucá