End-to-End Trained AI Model
Our perception detects and tracks all traffic participants. Additionally, we predict their future behavior to safely navigate complex scenarios.
Our system understands the road layout and road topology without additional map data. This enables scalable driving even in unseen environments.
Flexible Architecture
The SafeAD Perception model enables flexible sensor configurations. Ranging from camera-only for L2+ use cases to camera-radar-lidar setups for L4 driving.