IROS 2010 post-mortem

Well, I’ve just had solid 3 days of awesome.

I guess I’ll go to sleep, or something.

But before that, I suppose I should tell you about what actually happened, especially since I didn’t actually tell you blog-followers yet. I was very lucky, and got to attend IROS, the awesome-cool annual IEEE robotics conference. I wasn’t insanely lucky, however, because it wasn’t in NYC like Maker Faire was. Instead, it was in Taipei. Well, I guess I was insanely lucky.

So, first a list of what I learned:

• Sitting down for almost the entire duration of a world-spanning plane ride is hell on your feet.
• Dehydration is a terrible thing.
• Not having any purchasing power because you forgot your credit card pin is a terrible thing.
• I know almost nothing. It’s quite pathetic, really.
• Just because you’re attending a technical conference does not mean the internet is great at the conference venue.
• Just because you’re in Asia does not mean the internet is cheap and fast. Especially at some swanky hotel.
• Knowing that someone on Omegle is 12 hours off from you is one thing: actually moving 12 hours offset from most of the people you know is another.
• The rain there wasn’t quite as vicious as the worst NYC rain I’ve seen, but it definitely made up for it by raining the entire damn time.
• City design should account for the weather. This means making sure there are no slippery tiles of death, especially when it rains all the damn time.
• Willow Garage is the shit.
• ROS is the shit.
• Visual SLAM is the shit.
• Repraps are not intelligent robots
• You can include anime in your talk and get away with it.
• Google does not care if you can’t read chinese: it’ll serve up ads in chinese if you’re in Taiwan. Not a complaint, but it was still funny.
• Brilliant people can give terrible talks.

With all the single sentence taglines out of the way, time for some analysis.

Ahem

It is really, really, REALLY inspiring to see all this completed work. It makes me think that I can do something with myself. Hell, it has me thinking that grad school is really something I should do at some point in time. Not maybe, but something that I really should do. Failing that, R&D (and who the hell fails to go to grad school? I mean, the USA imports grad students, and the deficit is definitely due to people just not caring enough to try). I suppose the fact they had to squeeze all their work down into 20 minutes makes it seem a whole lot easier that it actually is has something to do with how inspiring it was…

In that vein, I need to check out some books. I’m attending a huge university with a huge library, there’s absolutely no reason why I’m not taking every advantage of my position. First stop: Probabilistic Robotics. I’m still shaky on my Kalman filters, not to mention particle or unscented particle or rosy-smelling car filters. Then, the best non-linear optimization book I can lay my hands on, and then… well, I’ll come up with something. Basically, I need to start I gigantic book orgy until I graduate, which is fine by me: re-reading old Methods of Rationality yields diminishing returns. </geekgasm>

Speaking of being inspired, I want to expand on one of my previous taglines. Namely, the fact that many of the speakers could not present to save their lives. True, not many of the presenters have English as a primary language, but a presentation is still a presentation, and knowing the speaker hasn’t been immersed in an english-speaking culture for 21 years doesn’t magically endow the listener with comprehension.

You would think that academics would be very good at transmitting information: they’ve been on the receiving end of explicit information transmission for most of their academic lives, and surely they’ve picked up (at least by osmosis) what works and what doesn’t. An academic should know how it feels to be on the receiving end of notes being read aloud, and yet many presenters just read their speaker notes. There was very little engagement of the audience: the professor that sponsored me on this trip was the only speaker I saw that came out from behind the podium while speaking (he’s very cool, and his accessible and engaging presentation only cemented that belief. P(coolProf|coolPresentation) = P(coolProf)P(coolProf and coolPresentation)/P(coolPresentation) where P(coolPresentation) > 0.99, yo, definite positive update).

Maybe an ignite presentation style would be good: we wouldn’t have the guys that fumble around with their slides (the usual ‘oops, overshot: lemme go back, oops again’), and squash the 30 minute presentations. We’d have to make it work with the people that finish their presentation right before they present, but it seems like it should be a trivial problem.

Meta-note on the conference: why in the world did they have to go and kill all those trees to make those huge conference books? Smartphone apps and interactive webpages not only fit inside a phone/laptop, they can do things like grep and do interactive planning. I’m sorely tempted to write a conference app framework, at least for android (finally dip my toe into phone development).

Another meta-note on the conference: it was really terrible that we managed to DOS the conference center routers. You couldn’t even connect to the router until Thursday, when people realized the rain wasn’t going to let up before the conference ended, and started to go explore the city anyways. I know the conference center doesn’t live and die by the quality of their IT infrastructure, but it’s an IEEE conference. Surely we could’ve whipped up some wireless links to relieve congestion on the main routers, or something (this thought makes me think that there’s some merit in becoming a conference/meetup hacker: someone that notices things wrong with an event, and hacks something up to remedy it before the event ends. Now, only if there were a job in it…).

On that note, I was able to actually get things down without always-accessible internet. Surprise, surprise. And then, those few times I got an uplink were always frenzied moments of trying to squeeze as much doing into as little time as possible. It adds some weight to the notion that I should not have accessible internet everywhere…

I was not impressed with the night market. It always seems like these things have 95% of merchants selling the same old clothing/shoes/jewelry/electronics (maybe the electronics booth is more of a NYC street fair thing), and the rest are actually doing something that you won’t find in every other street market in the world. Like, the bajillion claw-game parlor: I have no idea how they turn a profit with that many claw-games. We might have been in the wrong market: it was pretty tame compared to what  I thought it would be like, and it seemed like most of the shops were just extensions of the store fronts that are always there. So, my thoughts may very well be moot.

Finally, I noticed that I’m a very bad introvert: I seem to get energy from talking to people, especially helping people. Definitely not traditional introvert behavior. Not sure why that happens, when it started happening, or if I ever really was introverted.

Okay, I think that’s enough for a first mind dump. Time to actually unpack and start getting back into doing hw…