Developed jointly with Nvidia, ZF’s ProAI is a production-ready "supercomputer" touted as a key enabler to connect existing technologies and add the machine learning needed to achieve Level 4 autonomous driving. It allows automated vehicles to better understand their environment by using deep learning to process sensor and camera data.
ProAI uses the Nvidia DRIVE PX 2 AI computing platform and is capable of processing inputs from multiple cameras, plus LiDAR, radar, and ultrasonic sensors, and provides the computing power necessary to comprehend a 360-degree view and plan appropriate vehicle action. It is capable of over-the-air updates to add new features and capabilities throughout a vehicle’s life, and it is configured for V2X applications to communicate with other vehicles and the surrounding infrastructure.
Learning and networking capacity can help make an entire vehicle fleet safer and more efficient by means of "swarm intelligence," allowing cars or vehicles connected to one another to learn as a group instead of having to adapt to challenges individually. Announced ProAI auto applications include the first with Baidu for autonomous valet parking, with Chery for volume-production Level 3 vehicle for the Chinese market, and the e.GO Mover Level 4 electric minibus.
The open architecture’s scalability enables hardware components, connected sensor sets, evaluation software, and functional modules to be adapted across the mobility industry including for commercial vehicles, off-highway and agriculture, and mass transit. Even the electrified ZF Innovation Forklift, shown at this year’s Hannover Messe, features ProAI.
Click to view AVT’s 2019 ACES Award Winners best in class of