DARPA Challenge On NOVA

By on March 28, 2006

I know what I’ll be doing tonight.

NOVA: The Great Robot Race
The mettle of the world’s most sophisticated thinking machines is tested in a contest that pits artificial intelligence against unpredictable terrain.

Link. And if you miss it tonight, you can see it online tomorrow.

Another point of interest from the NOVA link; Google is providing “major funding for NOVA”. The AdSense & AdWords business must be good.

Gadgetopia

Comments

  1. I finally got around to watching this online last night (yup, missed the Tuesday broadcast, but I was sure thankful for the online version.) I’m not sure if it was just the way NOVA framed the story or what, but it seemed to me that the Stanford team’s entry was by far the more intelligent machine.

    The Carnegie-Mellon Highlander entry probably would’ve won if not for the glitch that kept it from running at full speed, but it seemed that they gave their machines an unfair advantage by using maps & satellite photos to plug in additional waypoints along the route. It looked like they had a team of 20 or so analyzing the route and filling in the spaces between the DARPA-provided waypoints. My 12 year old said it looked like they were cheating, and I have to agree.

    Maybe Stanford did the same, but the way it was portrayed on NOVA their system completed the course more by the rules than the Red Team. And if I were deploying a system like that for real-world work, I would rather have one that could do the job with a minimum of instruction at the front end. Especially if it’s going to be used in a in a war zone where the latest satellite images may not be recent enough.

  2. I agree with Dave’s 12 year old. Stanford appeared to only be concerned with proving their concept and Carnegie-Mellon appeared only to be concerned with winning. Somebody needs to sell the motorcycle guy a freakin’ clue too.

Comments are closed. If you have something you really want to say, email editors@gadgetopia.com and we‘ll get it added for you.