Free at last... now what?

Now that I am finally finished with the dissertation, I find myself with considerably more free time. As my wife has recently pointed out, I have a rare opportunity now to pick up a hobby before I allow myself to get bogged down with the next project or any of life's other pressing demands on my time. So, in addition to spending more time with my kids, and possibly doing little light blogging from time to time, what else could I be doing that I would enjoy and still fell like I'm making good use of my time.

One of the things you learn while getting a PhD is just how little you know out of all that is possible to know. I feel that I've learned a great deal during my studies, but there is a great deal more that I want to know... that I need to know. So, I find that it is time for me to leave the sheltered halls of academia and to try and make my way in the real world. I'm ready to put my hard earned education to work, but I am fully prepared to have to learn more in order to do so.

Power and propulsion are two topics that I love to ponder, but I don't think I'll have the resources to start tinkering with actual hardware for some time. If I'm extremely lucky, I might be able to land a job with one of the new space startups, where I'd be able to do that kinda thing as my day job. I'd jump at the chance to work with the likes of SpaceX, Blue Origin, XCOR, Masten Space Systems, Armadillo Aerospace, Ad Astra Rocket, or Scaled Composites.

I've also been wanting to get more seriously into robotics and astronomy for long time now. I have some brief experiences with both of these, but I'd like to be able to advance my skill set to the next level if possible. I often find myself thinking about better ways to interface to tele-operated devices. If Jeff Hawkins is correct about the hierarchical organization of the brain, then it should be possible to build a robotic system which can learn some basic, low-level behavior in response to commonly encountered stimuli, but still be able to pass on unexpected or novel information to a higher cognitive level (possibly a user). Specifically, I'm thinking about robotic telescopes which will be able to scan the sky autonomously, but be able to direct their attention to more novel phenomena based on their own past experiences.

One project that I've been putting off for a long time now (because I couldn't justify spending the time on it with the dissertation hanging over my head) is to look into doing more sophisticated mission prototyping using the spaceflight simulator Orbiter. Although the development of the main program seems to have stalled, there is still an active community of third party developers generating fresh and interesting add-ons for the simulator.

I have been contemplating a building rapid prototyping tool for designing rockets, RLV's, and other spacecraft. The program would be similar to NASA's Rapid Aircraft Modeler (RAM) that I worked on a few years ago. One simply selects from of list of standard components, then resizes and repositions them to rapidly assemble a desired configuration. Then, I would have the program output the necessary files to fly the model in Orbiter. By making use of some of the advanced modeling capabilities provided by the Orbiter API, it should be possible to get a fairly reasonable approximation of the vehicle's performance for a given class of mission.

And one final thing which looks like it will have to be a hobby since I can't seem to get my supervisor interested in letting me do it for work. I've recently managed to get CUDA installed and running on my laptop. I'd like to teach myself how to program the GPU. There is a lot of potential computing power tied up in those little chips, but not all computations scale equally well on the GPU. I think this is probably something that I will not truly be able to appreciate until I learn it through trial and error. Still, there's at least one instance of someone doing discontinuous Galerkin simulations on the GPU with encouraging results. I'd like to see if I can manage to get at least part of my DG solver running on the GPU.