I sort of want to do a project for the Tri-Valley Science fair(http://tvsef.llnl.gov/}, because the winners of that contest go on to the Intel Science fair, which has huge prizes and looks very good on a college resume. The only problem is, the thing that really interests me is Computer Science but I can't think of any type of project that I can do using computer science. I was thinking some random ideas could be like AI or some sort of physics engine or something, but I don't know if I could make something like that. What other types of stuff could I do? I know last year someone did something really lame like they tried to see how cheap they could go with a computer or something for basic tasks, so they basically built one with really cheap parts and installed several linux distros and did an informal usability test or something, but that doesn't seem like a very scientific thing to do, more like something I would do for fun.

How about a project that sets up two light beams and records when those beams are broken, the raw data is sent to the computer and it can then calculate the time between the two events. As a demonstration this could result in how fast can you roll a ball? How long does ti take you to travel from one point to the other? or even as an experiment to calculate g the acceleration due to gravity.

I suspect that most of the equipment is available in a standard physics lab...

You've almost defeated yourself already with this statement. It is just information you don't yet have. Go Get The Information Make failure not an option. Don't leave yourself a way out.