Course 3, lesson 30 of 100, Ages 7+

Robots and self-driving cars

Sensing, thinking, moving

Like I’m 5

A self-driving car is like a careful driver with eyes all around. It looks, thinks about what might happen next, then steers and brakes.

The big idea

Self-driving cars use cameras, radar and sometimes laser sensors called lidar to build a 3D picture of everything around them: cars, bikes, people and road signs.

AI then predicts what each of those might do next, plans a safe path, and controls the steering, speed and brakes, many times every second. Driving in rain, crowds and unusual situations is still very hard, which is why testing takes years.

Examples

  • Lane keeping: Many cars already help steer to stay in lane on motorways.
  • Robotaxis: Some cities have taxis with no driver in limited areas.
  • Tricky moments: A ball rolling into the road means a child might follow, and the AI must predict that.

How it works

  1. Robots and self-driving cars use sensors like cameras and radar to see what’s around them.
  2. AI works out what each thing is, a person, a bike, a stop sign, and predicts where it will move next.
  3. Then it plans a safe path and controls the wheels or arms. It gets tested again and again for safety.

Check your understanding

What do self-driving cars use to see?
Options: Sensors like cameras and radar; Binoculars; Only a map.
Answer: Sensors like cameras and radar. Cameras, radar and other sensors give the car a view all around it.
Why does a self-driving car predict what others will do?
Options: So it can plan a safe path ahead of time; To win races; Because it's bored.
Answer: So it can plan a safe path ahead of time. Predicting people's moves lets the car slow down or steer safely in time.

Remember

Robots sense, predict, plan and act, again and again.

Talk about it

What would a robot need to know to cross a road safely?

Go deeper

Self-driving systems split the job into perception (detecting objects), prediction (forecasting movement), planning and control. Rare, unusual situations are the hardest part, which is why companies test in simulation and over millions of real miles, and why most cars today still need an attentive driver.