Closing time

DISCLAIMER: the opinionated political satire expressed here is in no way affiliated or aligned with the views of Cornell University.
DISCLAIMER: the technical investigations of this team should in no way be perceived as anything but a dedicated and rigorous research effort. Please do not reproduce or otherwise make available any portion of the material found within unless permission is obtained from the author first.

I don't really want to write this post, because when I put in the final period, it'll really all be over. And yet here we are, and it really is all over- no more Monday bullsh*t sessions (systems meetings), no more rampant cursing, no more white deer, no more psychopathic walks, and no more cars without drivers. I feel almost like I've lost a very close friend, and I'm not quite willing to believe he's gone.

I know I've said this already (and I hate to belabor a point), but what we have done is nothing short of amazing. I think people are finally starting to recognize it, too. We drove the Tahoe back to Cornell last week (under cover of darkness... one last sketchy thing) and parked it outside one of the main Engineering buildings. I figured I'd sit there bored, but instead I spent the entire day explaining to people that yes, really, this car could drive itself. We set up a little console and played logs for people, showing some nice passes, intersections, and that awful double loop of Ford Taurses.

It's nice to be able to speak about it fancifully, as if it only happened in a dream. It's far weirder to do a double take and realize that there were no camera tricks, that only two and a half weeks ago I was actually sitting in that California wasteland putting all my innumerable calculations to the test. And then if there were a cake of weirdness, so to speak, it would certainly be taken by the fact that all those calculations actually did what they were supposed to do.

So what happens next? Well, in the fallout and aftermath (NO MORE MATH, PLEASE!) of the Urban Challenge I somehow have a few billion papers to write, and we're all celebrities (of a very weird sort) on campus. Everybody wants to hear about the journey, and I do love to tell it. Some stories I keep to myself, though, for my own private collection of laughs. For those of us that lived it, there are just some things that can't quite be put to words: spirit naps, operational crashes, the wide stance, Parking McGhee, the racing stripe, Hammox, the DARPA Urban Helpdesk, the participants' conference, and the sign we used to keep on the door to our lab. At this point, it really is starting to feel like the end to a bad 80's movie, where we'd show pictures of everybody with little snippets about what they'll be doing with themselves in the years to come. To make a long story short I'll just say that the team is scattered to the winds. Still, I have a feeling that as long as there is an ethernet cable between us, there is a decent possibility that our paths may cross again sometime in the future. I, for one, have no plans of retiring my "Mr. Fusion" nickname.

To my team I will say this: that you fellows are the best and brightest I have or will ever know. When I was shamelessly dragging my advisor onto the project way back in the beginning I told him that I'd stake my professional reputation on you fellows eighty times over, and I'm glad everybody got the chance to see why. I hope those close to you are proud and show it, and I wish you all the best of luck in the years to come. I've been asked to close with a particular verse taken from Timothy 2: "I have fought a good fight, I have finished my course, I have kept the faith."

#26 forever.

fin.

My snowball's chance

DISCLAIMER: the opinionated political satire expressed here is in no way affiliated or aligned with the views of Cornell University.

DISCLAIMER: the technical investigations of this team should in no way be perceived as anything but a dedicated and rigorous research effort.  Please do not reproduce or otherwise make available any portion of the material found within unless permission is obtained from the author first.

Well, here I am, writing this penultimate blog post on my "laptop" (a notebook, a piece of paper, and a pen).  It feels strange, abandoning my role of "Mr. Fusion" for the less interesting position of "bystander" in a little less than 24 hours.  On second thought, maybe it feels strange because it feels so normal, sitting here on this curb and reflecting on the adventure that is rapidly coming to a close.  The conversations are already starting to drift: segways, robotics, frisbees, Mexico, and already the words "urban" and "challenge" are popping up less and less.

For those that didn't follow yesterday, I will give a quick recap.  The starts went well, with a few GPS hiccups for CMU and a quick restart for IVS.  As the competition got underway we were joined by Jamie and Grant of the Mythbusters, who did their best to guess what the heck was going on in our robots throughout the day.  There were a few near misses: CMU vs. Terramax, CarOLO vs. MIT (twice, believe it or not), MIT vs. us (which still remains the only crash in which no drivers took part), and a number of cars had some run-ins with other less animate objects like curbs, gates, and houses.  It's like I've been saying all along: there were countless bajillion permutations of yesterday that would've ended in disaster, but only one in which things went right.  Still, and fortunately for us, yesterday was one of those precious few days where more things went right than wrong.  We ended up finishing the course- all 58 frickin' miles of autonomous driving (or at least, so said our odometer).  Some things still went wrong: the camera spewed bad timestamps, pose had a few hiccups, and something went funky on our throttle actuator.  The last of those was by far the worst, as it reduced our top speed to about 6 mph, which, as it turns out, is not so good for running a race.  That took us way out of contending range for any of the trophies (we probably had the slowest overall time), but we still finished.  I'll take that any day- it shows exactly what hard work and dedication are good for.  That's a lesson I think everybody should take home: that you really can do anything if you put your mind to it.  And even if there are still some people out there who were frustrated by yesterday or even by NQE, at least I hope you will one day be proud of what you've done.  For this 3.5 year odyssey of mine, I sure can say it was worth it.

Very little now remains of the competition site: a few of those blasted concrete barriers that everybody hated so much, and the occasional truck that carts more of them away.  All the Tauruses are gone, and there are no more sirens to be heard.  I'll give one last shout out to our neighbors and friends, the Honeywell / IVS team and Team CajunBot, for proving that sometimes even your "enemies" at competition can turn out to be good friends.  In the end, I think we can all agree on that, because if nothing else, we all ended up on this crazy adventure together.  At the awards ceremony one of my teammates remarked that this felt a bit like the end to a cheesy 80's movie, and it almost does.  I feel like the only thing that's missing is for everybody to jump up and freeze-frame, but since that can't happen, I suppose we'll just have to commit all this to memory and move on to the next journey.

And, since the Honeywell / IVS guy keep bugging me, I'll just give a quick and shameless plug to Delphi radars: their team may flood your pit with car exhaust, but they sure make a darn good radar.  I'll make sure to pass that message along to my kids, assuming I'm not infertile from all the laser and radar radiation I've picked up over the past year.

Anyway, I think that just about wraps up everything I have to say about the 2007 DARPA Urban Challenge.  I'll give just one more quick nod to those at NQE, those in the finals, and especially those who finished: CMU, Stanford, VT, Ben Franklin, MIT, and of course us.  I think I'll have one more hand to play (i.e. I'll be posting here once more) with closing reflections, but it will have to wait until I get back to Ithaca.  In the meantime I'll just sum up everybody's feelings with one last story.  Dr. Norm Whitaker (project manager for the Urban Challenge) leaned over to us yesterday after we finished.  He asked, "so, what're you guys going to do next?"  Our reply was simple: "we dunno, you tell us."

The story of my so-called life

DISCLAIMER: the opinionated political satire expressed here is in no way affiliated or aligned with the views of Cornell University.

DISCLAIMER: the technical investigations of this team should in no way be perceived as anything but a dedicated and rigorous research effort. Please do not reproduce or otherwise make available any portion of the material found within unless permission is obtained from the author first.

For the past few days I feel like I've been living in a dream. For one, I think I've seen more moving cars without drivers than with. I've become accustomed to that (admittedly even grown to like it), so it feels weird that everything is rushing rapidly to whatever end might be in store. I've gotten a lot of people asking to meet me, some to talk about the tracking and pose estimation we've implemented (I think the running total is four interested companies). Ironically, the vast majority of the people I've met have all asked to meet me because I write this little blog segment, which I suppose means that the goofy crap that happens to us also happens to everybody else. As they say, misery loves company, no?

As polished as some of this stuff looks, I've definitely learned that we have a long way to go before we turn the wheels over to the bitters and the byters. I guess the thing that's difficult for passers-by to realize is that these cars lack the ability to learn: they only know exactly what they've been programmed. If one of these cars changes lanes or makes a turn smoothly, it's because some dedicated zealot stayed up late one night tuning the crazy thing until it looked just so. That's what makes it just as much a labor of love as anything else- the perfect example of art imitating life.

So that's what casual passers-by should take from all this. What have I (we) learned? If there were one single lesson to wrap everything up into a concise little package it would be this: that no matter how absurd some of these scenarios may be, the bad ones always seem to find a way to happen. Those who were new to the team learned that a few days ago with the IMU. And then there was today, just in case anybody was starting to forget.

DARPA ran practice starts today as a sort of dress rehearsal for tomorrow. I'm glad they did, because the Fates definitely conspired against everybody and against all odds this morning. All the cars lined up in the start chutes, in much closer proximity than they'd ever been. That may have been unrelated, but that was definitely only the beginning. The first car went out on the course, only to get stuck and start backing up. The second car refused to leave its start chute. The third car went out and drove in the wrong lane before being paused.

That was only the beginning. OmniSTAR, who provides the high-precision differential correction signal that everybody uses for positioning, apparently decided to do a firmware update this morning, of all times and all days. A few minutes later we realized one of our lasers had stopped spinning (permanently, I am sad to say), which brings the total expense incurred from broken sensors to about $126000. Ouch.

And so here we are, the day before competition. The Tahoe's front bumper lies quietly in the corner, part of something once beautiful now grotesquely dislocated. Nobody's worried- this is exactly how these things are supposed to go. I explain it like this: there's about eight bajillion permutations in which something goes wrong, and only one in which everything goes right. Besides, if everything went smoothly, the story would be boring.