Audi expects to begin the series production of the central driver assistance controller (zFAS)—the core of future systems for piloted driving under development by Audi, within two years. Audi developed zFAS in collaboration with TTTech, Mobileye, nVidia and Delphi. (Earlier post, earlier post.) Delphi has been awarded the series production contract.
A wide range of sensor information comes together in the zFAS. The controller uses this to quickly compute a complete model of the vehicle surroundings and makes this information available to the various assistance systems. It is thus the central interface for all piloted driving functions.
At the moment, most driver assistance systems are managed by spatially separated controllers. Audi expects that it will be the first automobile manufacturer to implement this function in a central domain architecture. Audi has taken a holistic approach to consolidate the portfolio of functions, the sensors necessary, the electronics hardware and the software architecture into a central system. From the very beginning, the primary focus was on the safety concept.
High‑performance electronic components are a prerequisite for high‑powered computing in a compact package. The zFAS board is equipped with both the EyeQ3 mobile processor from Mobileye and the new Tegra K1 from nVidia. The tremendous computing power provided by this solution corresponds to the complete electronics architecture of a well-equipped mid-size car. With the high degree of integration, the new board is barely the size of a tablet PC. Its modular concept makes the zFAS flexible scalable.
Audi is also working with leading suppliers such as Bosch, Continental, Valeo and Delphi on the sensors and actuating elements, such as braking and steering systems. The objective is to develop common standards and offer customers modern driver assistance systems for greater safety, comfort and convenience on the road to fully automatic driving.
In the near future, Audi connect will enable the piloted cars from Audi to also learn continuously as they drive. The data computed by the zFAS board will be transferred via the cellular phone network—via LTE, where available—to an IT backend in the cloud. This will process these data using algorithms for machine learning and artificial intelligence before sending the data back to the car. The zFAS board will in this way continuously extend its capabilities to master even complex situations increasingly better.