How do driverless cars work?

Cars that don’t need people?

Companies like Tesla, Audi and Google have been working on cars that can drive themselves. These are called “Driverless” cars.

One day, we'll get here... but not for now.

Driverless cars can see and understand the world around them, avoiding obstacles to get to where they need to go.

How do they do it?

Inside each driverless car is a computer.

These computers tell the cars how to drive, using maps. These maps are made up of information the computer has been given by people, and new information is added to it while the car is moving!

Driverless cars use a series of cameras and sensors to gather the information their computers need to make the maps useful. The sensors can tell how far away other cars are, see where the road stops and where lanes are, and how to park. The cameras look out for traffic lights, road signs and obstacles... including people.

The computer combines all this information and adds it to the map. The cars use their maps and computers to make decisions about how to drive safely.

Driverless cars learn how to drive by practicing

While driving, the driverless car uses its computer to make decisions. After each decision the computer makes, it is told if it was right or wrong by a set of rules. These rules are based on instructions the computer was given by people – like “drive through green lights” and “always stop for people”.

If the computer makes the right decision, like stopping at a red light, it will add that action to the map. If the computer made the wrong decision, like hitting the garage door when parking, it won’t be added to the map. This way, the car learns not to make mistakes, and how to drive safely.

Driverless cars are very good at following rules!

After a lot of practice and training, the cars can drive themselves! One day you might have a driverless car too!