It's been a while since I've posted on this site (just under two years in fact). In that time I've made a lot of progress on several projects, and just in my knowledge of robotics.
Over the past year I have gone through a total redevelopment of my robot hardware. I've moved from using PSoCs in my robots to using Atmel AVRs. I love AVRs. AVR-GCC is a great, and free, compiler - but the greatest thing about AVRs is their tolerance to power supply fluctuation caused by motors. I've also discovered the LM2940CT-5.0 power regulator, which is a great and fairly inexpensive regulator. The 2940 is easier to set up and I seem to get better results than with the 5V switching regulators from Maxim that I had been messing around with (until they discontinued them). For my latest robot, GMR (which is still under consruction), I purchased some of Lynxmotion's new Servo Erector Set parts to make the neck pan and tilt - I was very impressed with the parts.
This semester I've been taking a robotics course through UAlbany. We are using Sony AIBOs, with Tekkotsu. In theory this is a really nice package - a behavior based, c++ programmable robot. Unfortunately much of the documentation is poor or out of date, and some of the constructs in Tekkotsu are incredibly non-standard and hard to understand. All-in-all though I like the idea, and once you get used to (and figure out how to) use Tekkotsu, it's quite nice.
Finally, I'm working on my third (and probably final) robot for the senior division of the Trinity Fire Fighting competition. I believe I have really worked out all of the kinks of previous years (poor drive train, sensor placement, power failures). The underlying software design is really over the top for the problem at hand , but it is planned in such a way that it can be extended to a future Expert division robot. The competition is about a month away - so far I have the robot navigating the maze with near perfection. The room search functions are currently non as reliable, but I have high hopes for Bayesian analysis method of filtering the UVTron. I've got some information on all of my projects on my website