Where to Go From Here
There are many possible directions. Structurally, the options are all open. The Lego Bug is well and good for a robot to navigate a circle maze, but needed improvement before it could go to a competition. When building, keep in mind that if you are using a fan, it is quite heavy, as are sensors, and if you intend to put them on top of the robot, make the robot able to handle it. Keep in mind where you will need to go rather than trying to make two completely different structures. This was our mistake and not only strapped our resources but lost us a great deal of time.
Also, get to know the machine shop and handy board basics as soon as possible. Network with other schools going to the competition (most people are extremely friendly, willing to share ideas, and sometimes will even loan you an extra sensor if you have a catastrophe) as soon as possible. The ability to see other ideas at work will be a great deal of help and assist in your own creativity. UMASS has a perceptual robotics lab. We made contact with graduate student Patrick Deegan and professor Rod Grupen.
Keep things simple. Most robots had the same basic structure: they went into the maze, navigated through the halls, entered each room, and swiveled the sensor it used to find the candle. With the programming, we have already come up with the basic programs needed to get started, so start right in on the good stuff. Navigation is tricky - your best bet will be to do a dead-reckoning type program with help from the range sensors.