How Waymo self-driving cars work

Waymo is a company that develops self-driving cars and offers a ride-hailing service called Waymo One. Waymo One allows users to book a ride in a fully autonomous vehicle through an app, without a human driver in the front seat.

Waymo One is currently available in some cities in the US, such as Phoenix, San Francisco, Los Angeles, and Austin. Waymo One aims to provide a safe, convenient, and sustainable way of transportation.

Waymo’s self-driving technology consists of four main components: mapping, perception, prediction, and planning.

• Mapping creates a detailed 3D representation of the roads, signs, signals, and landmarks that the Waymo Driver uses to locate itself and navigate.
• Perception uses a suite of sensors, such as lidar, cameras, and radar, to detect and identify objects around the vehicle, such as other cars, pedestrians, cyclists, and animals.
• Prediction uses machine learning and simulation to anticipate the possible actions and intentions of other road users, based on their current and past behaviors.
• Planning determines the optimal route and maneuvers for the Waymo Driver to take, based on the information from the previous components, as well as the traffic rules and safety protocols.

The Waymo Driver also has an onboard computer that processes the data from the sensors and executes the commands from the planning component. The computer is powered by server-grade CPUs and GPUs, and can handle complex driving scenarios in real time.