How Does a Robot Know What to Do?

Robots don't simply wake up in the morning and decide they'd quite like to assemble a car.

SCIENCE & TECHNOLOGY

9/12/20261 min read

a lego robot with two large eyes and a plant

Robots can build cars, explore other planets, move objects through warehouses and even help scientists investigate places that are dangerous for humans.

But how does a robot know what it's supposed to do?

Robots don't simply wake up in the morning and decide they'd quite like to assemble a car.

They follow instructions.

Those instructions can be written as computer programs. A program tells the robot how to respond to different information and what actions it should perform.

Many robots also have sensors.

Sensors act a little like technological versions of our senses. Cameras can detect visual information. Distance sensors can help measure how far away an object is. Other sensors can detect movement, pressure, temperature, sound or light.

The robot's computer processes that information and follows its programmed rules to decide what happens next.

Some modern robots also use artificial intelligence, allowing them to recognise patterns, make predictions or adapt their behaviour in ways that simple programmed machines cannot.

But even very clever robots depend on technology designed and trained by people.

Think like a programmer: Imagine giving a robot instructions for making a sandwich. How detailed would your instructions need to be?

“Make a sandwich” sounds simple to us.

To a robot, it could be the beginning of a very long instruction list.