It's my last semester at Purdue for my B.S. in C.S., and I'm taking the capstone course they require for all Computer Science undergrads. I wanted to do a project based on voxelization, for several reasons. I'm interested in game development in the long run, and voxels seem to be growing in popularity with terrain and scene generation but are also essential in most volumetric rendering methods used in medical imaging. I pitched the idea of an application that loads and renders a volumetric model, but that's about as far as I got. I've found the Visualization Library, which seems to implement most of the algorithms (Marching Cubes) etc.

Are there any unique or "fun" applications of voxelization/volumetric modeling that a lowly undergrad could accomplish?

Between having a bad flu, the university being closed for a full week after, and the professor "overseeing" the course being largely inaccessible for his own reasons, elicitation of project ideas from professors has been hampered (though anticipating some reply emails with a few faculty in visualization research). The criteria the project must satify includes relating to multiple C.S. disciplines (AI, computer graphics, compilers, etc.), must be "useful to society" in someway, being "innovative" (a fifth of the grade depends on this), and we have essentially less than 3 months to complete the project and report. In his words, it could either be 80% research and 20% implementation, or vice versa.

Are there any unique or "fun" applications of voxelization/volumetric modeling that a lowly undergrad could accomplish?

I don't have any huge ideas on what you could do other than the obvious, but voxel rendering would probably be a very interesting thing to work on even if it were pretty much face value.