The robots are controlled by a neural network that decides how to move. There are 4 possible actions:
- Roll forward
- Turn Left
- Turn Right
- Stop

The inputs to the neural net describe what the robot can see.
- The robot can see food that is close by. It is categorized as in front, to the left, or to the right.
- The robot can see other robots that have a green light on. They can be seen from much further away than food.
- There are two inputs that describe what the motors are doing at the moment.
- There is one bias input that is always true.

The robots get points for being near food and the most successful robots get to go onto the next round. The order from most points to least is:
- There is food in front of the robot and it is stopped
- There is food in front of the robot and it is still moving
- There is food near the robot but not in front
- There is no food near the robot and its moving
- 0 points for sitting still not near food.