Innovative Production Machines and Systems Conference

We conducted a number of experiments to test the ability of the system to learn. These experiments were conducted in the environments shown above. These idealised grid based environments contain a number of obstacles. Each grid square represents a state. The robot can take one of eight actions in each state allowing it to move between states i.e. (North (N), South (S), East (E), West (W), NE, NW, SE, SW). The mobile robot can learn to move through its environment towards a recharging station located in the corner of the environment. Moves outside the environment or through an obstacle are considered illegal and not permitted; the aim of the robot is to find the shortest legal path back to the recharging station.