Waymo and Uber are expanding robotaxi access, but many people don't understand how autonomous vehicles (AVs) work. To build public trust, a clearer understanding of AV technology is needed. AVs use a four-part system: Perception, State Estimation, Planning and Prediction, and Control.
Perception uses cameras, radar, and LiDAR to observe the surroundings. Cameras capture detailed visuals, radar detects objects in darkness, and LiDAR creates 3D maps. Machine learning algorithms then categorize these objects.
State Estimation determines the vehicle's precise location using GPS, camera data, and LiDAR. This helps the AV understand its position relative to other road users.
Planning and Prediction uses this information to estimate environmental changes and plan the vehicle's path. This includes high-level route planning and low-level decisions like lane changes.
Control translates the planner's actions into physical movements, ensuring smooth acceleration, braking, and steering. This complex system requires precise control to avoid jerky movements or loss of control.
Autonomous vehicles are complex systems. A better public understanding of these systems can help build trust in the safety of the technology.