If you're a new programmer, or at least new to LabVIEW, I would recommend totally ignoring the default autonomous code. Our programing team looked at it and then decided,
this doesn't make any sense, lets write our own.
We mounted our sensors in a triangle that we placed in the front of the robot, so that one was about 2.5'' more forward then the rest. The code was actually very simple. We had a loop feeding 0.5 for both joystick values into our drive.vi until "left sensor" OR "right sensor" OR "middle sensor" was tripped.
Then it moved in to another loop that feed 0.8 for both joystick values in to our drive.vi; however when the left sensor was tripped the left value dropped to 0.5, and visa versa. We did this by arraying the boolens we got from the sensors, turning then in to a number, and then feeding the number in to a case statement that contained the constants we feed in to the drive.vi as the joystick values. This loop would stop when any combination of sensors that wasn't just the left or just the right was tripped.
This followed the line GREAT and stopped when it got to the end. It's simple (don't worry it's much easier to see then to say), easy to debug, and above all works really well.