In response to the suppositions that rust is the cause of the problem, our robot was (apparently) stored outdoors, in the rain, between ship date and the Silicon Valley Regional, in 2003. When we opened our crate, which was noticeably damp, there were two inches of standing water (so I was told, I cannot be absolutely certain about this fact), and all of the (non stainless) steel on our robot had rusted. Wetzel posted
this image shortly after the discovery. You can't see either of the two drill motors in that picture, so
here are a couple of good pictures. As you can see, they were caked with oxidization. We were all deathly afraid that our drivetrain was completely ruined (the geartrains were rusted, as well), but when we carefully started the system, there were absolutely no problems. The drill motors worked (as far as we could tell) perfectly.
This isn't to say that rust is absolutely not the problem in your situation, but it's some experimental data.