DARPA's Grand Challenge goes urban!

Oh DARPA, how we love thee, let us count the ways. Your first Grand Challenge brought us giggles and guffaws as those first-gen automonous bots failed to even complete the course. Grand Challenge 2005 returned to the desert with oohs and ahhs with four vehicles actually completing the rugged, 132-mile unmanned course in less than 10-hours. Now, DARPA presents the third and best event yet, the Urban Challenge! This time around, DARPA will award prizes to the top three ($2M, $500k, and $250k respectively) autonomous ground vehicles that safely complete a 60-mile urban course on a simulated military supply mission (read: not likely to be blowing things up) in under 6-hours. The main event is schedule for November 3, 2007 with several qualifying events to be held in the run-up. And yeah, you can expect a full-on grudge-match between 2005's winning Stanford team and runners-up Carnegie Mellon (owners of the Crusher) who both announced participation in what was hopefully a sweaty, WWF-like expletive-ridden shout-off with folding chairs a-flyin'. [Warning: PDF link]

[Via MSNBC, Thanks William]

[bill murray @ May 2nd 2006 9:06AM]

is this the high tech rv that was driven in Stripes?

[Drew @ May 2nd 2006 9:14AM]

Just a few weeks ago PBS had on a special on Nova about last year's race; it was really pretty good. It's here at www.pbs.org/wgbh/nova/darpa/. Dang them Stanford kids and their winning ways...

[Reid Bode @ May 2nd 2006 9:18AM]

Hey, I saw that Nova show on the PBS HD channel. Good stuff.

Can we assume this'll be a realistic military delivery (w/ gunfire, terrorists, and roadside bombs)?

[William @ May 2nd 2006 10:11AM]

Reid,

As per the document, "vehicles must autonomously obey traffic laws while merging into moving traffic,
navigating traffic circles, negotiating busy intersections and avoiding obstacles."

So, that's a no. I looks a bit more mundane than a combat situation. That's slated for "DARPA Grand Challenge eXtreme!!"

------
http://www.gorobotics.net

[OMAC @ May 2nd 2006 10:12AM]

Yes this os the RV that was in Stripes. Great movie!
Boot chuggaluggalugga Boot chuggaluggalugga Boot chug boot chug boot!

[Josh @ May 2nd 2006 10:12AM]

Dessert!

[Penguin @ May 2nd 2006 10:57AM]

More specifically, that is the EM-50 Urban Assault Vehicle.

[Richard @ May 2nd 2006 11:59AM]

I think you have WWF (World Wildlife Fund) and WWE mixed up. It's been a while since WWF won their lawsuit and forced the wrestling idiots to change their name.

Or, maybe you ran a foul of some nature lovers ...

[James Fabin @ May 2nd 2006 2:39PM]

I'm sure CarTV.com will be there again to cover it. They have some awsome coverage from the last one - here are some links:

Part 1 of the Final:
http://www.cartv.com/content/research/channels/index.cfm/action/showvideo/vid/e_0158/vscat/DARPA/vcat/Event

Part 2 of the Final:
http://www.cartv.com/content/research/channels/index.cfm/action/showvideo/vid/e_0159/vscat/DARPA/vcat/Event

both of those have links to all their other coverage of it. It's pretty awsome for those who haven't seen this. I can't wait till the next one!

James

[Bill @ May 2nd 2006 2:42PM]

I'll only be impressed if the bots can navigate a New Jersey traffic circle.

[kenny @ May 2nd 2006 7:48PM]

you know it has to be the size of the Stripes van in order to actually work/navigate and probablly finish the course...due to the fact how the Hummer sized ones did

[roy andrea crabtree @ May 3rd 2006 2:13AM]

Per comments #3&4 above: welll, actually,given the number of faiklure mnodes that occurred in the currnetly just completed event, if DARPA actually tried to run the coure in a real world traffic scenario, it miht actually be roughly the equivalent of real combat ith casualties. If done in a simulated environment, then the casualties will only be simulacra...
----
I wish the best to the combatants, err, contestants; given limited resources (time, personnel, funds expertise [which must be developed)), failures are always spectacular. SO, throughout his career, was Thomas Alva Edison, among many others.
----
It would, howevr, be good, this time, if one of the entries had a fully populated virtual modl with enough of a peformance envelope to have some failure fallbacks: not just to colete the course, but enough to warrant serious ral life consideratons.
---
All in 19 months, of curse. Mere technical brilliance is not gonna do it this time: it willl take team work that exceeds the echnical brilliance to succeed. A sugestion to the temsto be involved: you will want an operationsal manageent specialist on our team, to flow your resources. You may already have the equivalent; if not, you will wnt to invite a few more members to join your team(s). Good luck!