The ZF ProAI, scheduled for series production in 2018, will help cars to “understand” their environment by processing and interpreting sensor data by means of Deep Learning Technologies.
Current series-production Electronic Control Units (ECUs) for automotive deployment typically embrace multiple computing units, of which each one is responsible for a dedicated function. Experts however are convinced that the computing power of such approaches won’t be enough for highly complex tasks like automated driving. This is why ZF chose Nvidia’s AI platform – according to ZF CEO Stefan Sommer, the system offers the number crunching capacity of a supercomputer. Being a scalable system, the Drive PX 2 is capable of processing the signals of multiple cameras, plus radar, lidar and ultrasound sensor signals in parallel. ZFs autonomous driving computing platform ZF ProAI will enable the vehicle to “understand” all relevant events around it in real-time. It can localize the vehicle on high-definition navigation maps and plan a safe trajectory.