Virtual Vehicle

VIRTUAL VEHICLE, ZF and NVIDIA present the “Dream Car”

VIRTUAL VEHICLE and ZF present the “Dream Car” – a self-learning, AI-based Level-4 automated vehicle at the AI & Deep Learning Conference | GTC 2018, San Jose (CA), Mar 27, 2018

Learning while sleeping is the groundbreaking idea of the “Dream Car”, a self-learning, AI-based Level-4 automated vehicle - without driving around, the car constantly learns and adapts itself based on data acquired from other cars driving around somewhere else in the world. The key is AI and ZF’s ProAI which has been developed together with Nvidia within just one year. In close cooperation with ZF, VIRTUAL VEHICLE developed and programmed software for data analysis and compilation, built the “Dream Car” prototype and implemented the system for initial real test-drive scenarios. The current development status was presented today at the NVIDIA’s GPU Technology Conference in San Jose / CA.

San Jose, CA -  Mar. 27, 2018 -  ZF, VIRTUAL VEHICLE, and Nvidia have bundled their forces to develop an AI-based L4 vehicle for urban scenarios within only six months – the so-called “Dream Car”. The vehicle is basically equipped with cameras, laser scanners, and radar sensors enabling a redundant and reliable 360 degrees environmental awareness. As these sophisticated sensors require a lot of computational power, the ZF ProAI platform plays a key role. Today, ZF and VIRTUAL VEHICLE presented the current development status at the 2018 GTC in the heart of Silicon Valley.

ZF ProAI: Artificial Intelligence next level

AI is one of the cornerstones for self-driving cars in a complex, unpredictable, and hazardous world. It allows for a reliable object detection especially in different, even harsh environments. The benefit of the ZF ProAI is that both the hardware and the software are modular and can be scaled according to the application and the desired level of automation.
Arnold Schlegel, engineer of Advanced Development at ZF Friedrichshafen AG and project manager of the "Dream Car" explains: “We used the robot operating system (ROS) as development framework, which provides a modular development, flexible integration platform, and data exchange within the vehicle. Moreover it enables the possibility to evaluate and use solutions from the large robotics community, to achieve the goal to bring self-driving cars faster on the road.”

Bringing the “Dream Car” from virtual to real

Daniel Watzenig, Head of EE & Software at the VIRTUAL VEHICLE Research Center and Full Professor at the Graz University of Technology, describes the intense team effort to let the “Dream Car” become real: “Together with the engineers of ZF, we have implemented numerous functions that enable highly and fully automated driving up to levels 3 and 4. Without the vehicle having to be in motion, its AI recognizes this data and interprets it as if it were travelling exactly on this route. Steering angles, braking and acceleration correspond exactly to the transmitted trip. The vehicle can thus ‘learn’ how to interpret a traffic situation without actually having to drive in traffic.”