A robot is located on the cell with a red arrow, facing in the direction the arrow points. The robot is programmed to move forward until it reaches a wall (the perimeter of the puzzle or a black cell) or a cell it has already visited. At this point, the robot will attempt to turn 90 degrees to the right and continue moving forward again; if a wall or a cell that has already been visited it to its right, the robot will attempt to turn left instead. The robot will stop when it can no longer move to a cell it hasn't already visited. Shade in some of the cells black such that the robot will visit every white cell in the grid, and stop on a dot. Cells containing dots or arrows may not be black; additionally, the robot must exit a cell with an arrow (including the cell it starts on) in the direction the arrow indicates. (See here for an illustrated example.)
