[ChrisH]

The first thing is to decide what you want the field to look like at the end of the match. Do this for both autonomous and teleop. It can actually take several days to figure this out, depending on the subtlety of the game.

Then start working on how you are going to make the field look that way. Now you can start thinking about how will you manipulate the field objects to achieve your goals. Start with all sorts of ideas and narrow them down. Figuring out how much power is required to accomplish a certain task is one way to eliminate the really unpractical ideas.

Use the Kitbot or an old robot to prototype your best ideas. Experiment with how well your chosen design works in real life. Do you grab that Trackball firmly and easily or do you spend your time chasing it around the field trying to catch it?

Once you have a prototype that works well, go design your real robot. The build goes much faster when you know what your robot will look like before you start.

The BeachBots typically don't start building our real robot until sometime during week 3. In '05 we didn't cut metal until 4 weeks after Kickoff and didn't make a final decision on our drive concept until week 5. (We were carrying two concepts and the frame was designed to accomodate both)

By the time we start building we have a running prototype that our drivers can practice with and our programmers can use to debug the code, but it probably would not pass inspection. It almost certainly violates the "must be fabricated this year" rules. That is OK, it will probably never leave the shop, at least not until the off-season. As "real" parts are built we make 2 and incorporate improvements into the prototype bot. This way, when the real one ships it has an older brother left behind. Then our drivers go and wear out the tires on the practice robot.