[gblake]

I'll bet that driving over real terrain and obstacles, and getting enough endurance from your battery(s) (affected by battery storage, motor efficiency, payload power needs, and misc) will dominate your physics problems, if you have to traverse non-trivial distances or stay on-station 24x7. But they won't be your biggest problem (see last paragraph).

So,

• Build miniature prototypes out of wood, VEX/Tetrix parts, etc.
• Build a full-sized, full-weight prototype out of 2x4s and your chosen wheels/tracks (see Ike's post).
• Drive (tow, push, whatever) the prototypes outside.
• Use easily-revised, modular framing for the real thing. It will need to be adjusted as you iteratively design/integrate/test

To solve any other problems, if you burrow deeply enough into the archives here, you can probably find a workable solution.

Also, it's a good rule of thumb to estimate project size/complexity by the number of interfaces (mechanical, electrical, software) you have to design and/or implement, instead of counting the number of subsystems that exist. It's good to pay attention to both, but it's the interfaces that will eat you alive (in part by causing both your time spent searching for or choosing parts, and your resulting parts-count to explode).

Also, everything takes about 4 times as long as you think it will/should; and simple 10-minute tasks often balloon to take 2-4 hours by the time you are finished discussing them, setting-up for them, doing them, testing them, and cleaning up after them. The clock is your biggest enemy, not robot-physics.

Blake