Background

Went to school for aerospace engineering, finished with a degree in computational physics. And while I’d like to say that this was the result of some deep self-discovery, it’s really not. But it did leave me with an appreciation for simulation.

The appeal is hard to explain, but I’ve come to think it’s this: you can’t fake your way through building one. You have to decide what matters, write it down, and hit run. If you got it wrong, it tells you. Usually not politely.

Somewhere along the way, that became the test for whether I actually understand something. Could I build it from scratch and have it behave the way the real thing does? If not, I probably just know about it.

That habit followed me out of school. I spent two and a half years in the Office of the CTO at a defense contractor, building data systems and dashboards. Different field, but the same rule applied. A dashboard is really a sensor, and every sensor is a decision about what matters.

Now I’m an engineer at Sanmina here in Austin, where I’ve been building an automated inspection cell around a collaborative robot, mostly on my own. It’s the first time the thing I’m modeling can push back. Turns out reality is even less polite than a simulation.

That cell is also what’s pulled me toward robot learning, which is a little ironic. A lot of the best models work great and can’t tell you why. Not what I planned, but it’s been a pretty good accident so far.