Automated cars: How safe is safe enough?

Blickfeld tackles this question together with nine other partners in the “VIVALDI” research project funded by the Federal Ministry of Education and Research. Countless test driving kilometers are required to make automated vehicles as safe as possible. This takes place both on the road and virtually in simulations. VIVALDI aims to develop virtual test methods to increase reliability.

Here’s the project at a glance:

What is the project about?

The VIVALDI project aims to advance virtual driving tests for automated driving and prepare them for worldwide standardization. For this purpose, virtual test methods and environments for sensor systems consisting of camera, radar and LiDAR are developed.

How is Blickfeld involved?

Blickfeld’s task is to design scenarios for which LiDAR provides crucial information. Everyday situations, as well as the so-called edge cases, are simulated, which are unlikely to occur but where an automated vehicle must nevertheless make the right decision to act. For the simulation, it is also crucial to characterize the entire environment in detail. Here, Blickfeld determines which parameters are decisive for successfully testing the LiDAR sensors. Furthermore, Blickfeld provides the LiDAR sensor technology for the test vehicle.

Which test methods are being developed?

The project includes simulation, i.e., the digital replication of the vehicle sensor technology, modeling, i.e., the replication of all function-giving influences, such as sensors, propagation of sensor signals, and vehicle environment, and as the final step towards series production and market release, the validation of all test results.

To validate the virtual tests, they are also compared with real tests. Therefore, a test vehicle is deployed for the project.

Why are simulations so crucial for automated driving?

Safety is the most important factor in automated driving. No self-driving vehicle will be licensed if its safety cannot be proven. The interaction of sensors and software for automated driving is very complex and therefore requires sufficient testing. This involves hundreds of thousands of test kilometers. Only relying on the real-world tests makes this process painstakingly slow. Therefore, a mixture of real tests and simulation is essential, with simulation shouldering the majority of the required test kilometers.

What different conditions are simulated?

The sensor systems are tested virtually in different environments, such as the city, country or highway, in diverse weather conditions, and with different road users. These tests are planned in several stages, i.e. first simple scenarios are simulated and then critical real traffic scenarios are tested.

Who is involved besides Blickfeld?

The project is managed by the Technical University of Ilmenau. Other project partners besides Blickfeld are Mercedes Benz, Continental, AVL Deutschland, IPG Automotive, the German Aerospace Center, the Karlsruhe Institute of Technology, the Technical University of Darmstadt, and the University of Applied Science in Kempten.

The project is part of a larger German-Japanese research plan with the overall goal of making highly automated vehicles safe.

What is the amount of funding?

The project volume is nearly 4.7 million euros, out of which the Federal Ministry of Education and Research is funding 3.7 million euros and the research partners themselves are covering the rest.

Share

Share on facebook
Facebook
Share on twitter
Twitter
Share on linkedin
LinkedIn
Share on xing
XING
Share on whatsapp
WhatsApp
Share on email
Email

Next events

Vision

October 5
- 7, 2021
Stuttgart

ITS World Congress

October 11
- 15, 2021
Hamburg

DVN

November 15
- 16, 2021
Frankfurt