Thursday, May 19, 2011

Proving Vs Disproving

Some of the most fruitful work I did during my PhD studies was with a Postdoc. Working with SuperK was the classic East-meets-West in middle and total opposites attract (no, not in that way). Quiet, determined, calm, yet surprisingly funny versus loud, persistent, obnoxious, and a bumbling idiot at times. Can you guess who's who?

Anywho, SuperK and I worked on a bunch of topics, wrote a bunch of papers, and traveled to a bunch of conferences together. After SuperK moved on, I was pretty upset because my sparring partner in the lab and the one person who understood me academically was no longer there. We still keep in touch but SuperK's at a company now doing other cool things but it's all secretive.

When SuperK left, there were still a ton of experiments to finish and things to prove in the lab. The end goal of all of this was to prove the uber simple system and have our industrial partner pay the university (and us) an 18-wheeler full of $100 bills to license the technology. I tried to do as much as I could but I was saddled with my main PhD project and other projects and ultimately we didn't have the funds at the time. Right around the time I defended last fall, OldEuropeU hired NewPhD to finish working on that project. I agree to mentor NewPhD and things were going great....

... until yesterday.

Aside from building a demonstrator in the lab, we also needed to make sure the theory matches the results. Before building the final setup, we spent a few months to building different systems and making sure the more known methods indeed worked. Plus we had a few conference proceedings to write and we wanted to do a comprehensive paper with each system at the end.

Well, some things changed with the setup and things got sidelined in Europe. We needed to get a key component fixed in order to prove the final setup. In the meantime, I moved back to the US and am now trying to start new stuff here yet finish stuff from there. That left NewPhD to finish the system characterization while I worked on the theory. Before I left Europe, we did some rough calculations and I was pretty confident.

I should really learn to check my math. It turns out, the ideal system (for practical reasons) will not theoretically work for all cases, so it will not be robust enough for transferring to industry. And, to add some fuel to the fire, we guess wrongly with our intended results and were accepted for a talk at a conference and the paper is do ASAP. We were fairly vague in our description in the initial abstract but now our proceedings that was going to prove something actually disproves our initial ideas.

I'm pretty bummed about it this. The really sad part is this technology could be adapted to a bunch of different systems and widely applicable.


  1. How frustrating! Sometimes that's the way things go, but I'm sorry it went that way for you.

  2. It's ok. I still think there's a way around it. Just not smart enough to figure it out yet...