ADLINK’s autonomous driving computing platform enables the autonomous vehicle in multiple stages:
1. Perception
Perception enables sensors to not only detect but also acquire, identify, and track objects in their environment. During the perception stage, the image processing controller,
RQX-59G, will collect data from camera sensors and send it to the computing platform for sensor fusion and AI planning.
2. Planning
During the planning stage, the computing platform,
AVA-3510 utilizes perception data to calculate the path. This involves route planning, behavior planning, and trajectory generation modules. Route planning establishes the desired waypoints, behavior planning selects actions to maintain the path, and trajectory generation creates the precise path. These planning processes enable autonomous vehicles to navigate and make informed decisions.
3. Action
After the computing platform has finished planning, it will send the information to the vehicle's control systems, instructing it to function as the AI had intended. The functions include accelerating, braking, and turning the steering wheel.