AV2.0

AV2.0 is one of two main approaches self-driving cars use to learn the road. The traditional approach (HD map-based driving) surveys roads in advance to build precision maps, and the car only moves within that map. Like a student who memorized the exact test material, it's strong on routes it has memorized, but gets confused by alleys not on the map or construction sites that suddenly appeared.

AV2.0 is the opposite. It doesn't use lidar (a sensor that scans surroundings in 3D with lasers) or pre-built precision maps. Instead, AI judges road conditions in real time based on what cameras see. Like a student who understood the underlying principles and can solve a problem they've never seen before, the car reads the situation and responds on its own even in an alley it's never visited. Companies using this approach argue that this means they don't need to send survey vehicles ahead when entering a new city, allowing them to scale faster.

AV2.0 is the name British self-driving startup Wayve gave to its own approach. The core idea is AI that learns from experience the way humans learn to drive, making judgment calls itself instead of relying on hand-coded rules.

News coverage explains it this way: "Wayve doesn't use lidar or pre-surveyed HD maps. The company calls this approach AV2.0." Contrary to a common misunderstanding, AV2.0 isn't the name of a specific company's product — it refers to the self-driving approach itself, where AI makes judgments from experience instead of relying on maps.