To understand state machines, let’s make a state machine based on a classic childhood game: Freeze Tag.

Think about how you play Freeze Tag. You’re running to escape the person who is “It”, and if they tag you, you’re frozen. If a non-frozen teammate tags you, you’re back in the game!

“Frozen” and “running” are states describe a set of actions you should be performing. When you are in your “running” state, you are moving like the wind to stay out of reach of your opponent. When you are caught and “frozen”, you have to stand in place and beg for a friend to come and help.

What moves us between these states. What events make you go from “running” to “frozen” and back again?

Well, to go from “running” to “frozen”, the person who is “It” has to touch you. For you to get from “frozen” to “running”, a teammate has to touch you. We can show this with a state machine diagram:

State machines are useful because they help us to organize and modify the behavior of complex systems, such as the robot you’re building right now. If you use a state machine to control the robot, you can see what the robot should be doing, and the events which should trigger these actions. This means that state machines not only help us to build the robot’s brain, state machines also help us to diagnose any problems (aka debug your code!). If an event happens but doesn’t trigger the right state, you know something is wrong.