Scalable ECU platform to accomplish level 3 autonomous driving
TTTech’s contribution to this system is a scalable ECU platform that combines a flexible architecture and safety kit that can be adapted to the needs of customers like Audi – but since it is available to automotive OEMS and suppliers, also other carmakers can be expected to put the platform into the center of their car’s electronic guts. TTTech has already developed a similar platform based on Renesas microprocessors, and eeNews Europe learned that Asian carmakers are taking closer looks to the system.
The various multicore processors in the zFAS (Mobileye and Nvidia for image processing, Altera / Intel for sensor fusion, number crunching and miscellaneous tasks, plus Infineon’s Aurix for steering and braking) are tied together through a safe deterministic Ethernet connection. TTTech’s software provides a complete runtime and development environment and complies to the highest safety level (ISO 26262 ASIL D). TTTech holds the responsibilities for ECU safety from prototyping, production of A and B samples and series production and ECU flashing, explains TTTech member of the executive board Stefan Poledna. “Our environment supports Autosar as well as alternative operating systems such as Wind River or Linux derivates”, Poledna said. Likewise, it is closely entangled with development tools like those from Vector Informatik and other third parties, Poledna assured.
With autonomous driving currently the hottest topic in the automotive industry, Poledna commented on the technological issues of this proposition. “I see three major challenges,” he said. “Two of them are already discussed on a broad basis across the industry. The first is sensor technology and environment perception. The second is trajectory planning, an area currently taken care of by many software vendors. But there is a third challenge – one that so far has not appeared on the radar of many market participants: Fail operational vehicle behavior. If an error occurs in the electronic control system, the vehicle must be able to continue driving until a safe stop is possible.” This requirement poses much higher challenges than the technology used in today’s vehicles. Safety-critical functions must meet the highest safety levels. The problem: “Today’s Artificial Intelligence Algorithms and platforms as well as image processing devices are not designed for ASIL D.”
This is the market segment where TTTech positions itself, Poledna added.
Related articles:
Why Audi’s zFAS is a blueprint for next-gen domain architectures
Audi rainmaker sketches new computing architecture for cars
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
