Inside DARPA's Robot Race

Belfegor writes "The PBS series Nova has a great feature on their website, regarding the coverage of the DARPA-sponsored 'Robot Race' in which driverless vehicles 'competed' in a 130-mile race across the Mojave Desert. The full show is available on the website, and besides that they have plenty more information about the robotics behind the challenge, and also some pretty cool out-takes from the show."

Seen it

[Moby Cock]

PBS broadcast that show last night. While I realise that is is a little 2001 to actually watch a program when it is braodcast, I did. And I really enjoyed it. I am hardly current on the status of autonomous robotics and I was pleasantly surprised by how far along the technology is. 130 miles through the dessert using only GPS and local sensors is a pretty amazing feat, and that course was tough. It features mountain switchbacks, tunnels and other hazards. If you even have a passing interest in robotics I recommend watching the show.

torrent

[Anonymous Coward]

http://www.mininova.org/tor/266446 [mininova.org]

Re:Great show but...

[CXI]

They'll get patent recognition if they, you know, filed any patents. These teams can do whatever they want with any innovations they make. Many of them, especially the school based teams, operate under grants from other agencies which might have limitations on who owns or can patent what. However, each team makes the choice about where their funding comes from and what strings are attached to it.

Tell PBS Thanks!

[IanDanforth]

I really enjoyed this, especially the fact that it was the full show online for free.

Let PBS know [pbs.org] what you thought about the format, show, or anything else.

-Ian

Sebastien Thrun's book

[jdduke]

If anyone is really interested in the technical and mathematical side of this stuff, I definitely recommend Probabilistic Robotics [amazon.co.uk] by (among others) Sebastian Thrun, director of the Stanford Artificial Intelligence Lab and leader of the winning team in this race.

Seminar about Stanley

[Anonymous Coward]

Anything you gleaned from the NOVA documentary is bullshit compared to watching the guy explain it himself [cmu.edu]

Mostly paid employees and purchased parts

[Animats]

Most of the successful teams had significant numbers of paid employees. Stanford had about sixty people back at Volkswagen working on the hardware. CMU had a huge headcount; they had more than fifty people on site at the Speedway, including people on the payrolls of Lockheed, Caterpillar, and other vendors. Oshkosh Truck was all paid employees. Didn't talk to the Grey Team much, but they were paid by some Insurance company.

The big breakthrough was Stanford's texture vision system. I was very impressed with that. Computer vision in unstructured environments has a terrible track record, yet they made it work. Everything else was basically integration of off the shelf gear.

One accomplishment not oftened mentioned is that, by year two, many of the components that weren't available in year one were available off the shelf. In year one, getting an integrated GPS/INS/compass/odometer system was very tough. Applanix had one that cost $70K, took up a 4U rack, and required air conditioning. (CMU used it.) By year two, you could get something comparable from any of three vendors for about $20-$30K, ruggedized and able to run on 12VDC. All the successful teams had one, usually from Trimble or Novatel. Once you have one of those, just staying on course is straightforward. Then it's all about obstacle avoidance.

Re:Mostly paid employees and purchased parts

[mmde]

A minor correction... Stanford actually had 60 people total on the team. There were 9 people from VW working on the vehicle. You can see a list of all of the team members at our website [stanfordracing.org].

Re:Fascinating program

[SpyPlane]

Do a google search on Sabastian Thrun, he was the team lead for Stanford, and formally at CMU (what a non-coincidence). Most of the software they used on Stanly (Stanford's bot) was either written by Sebastian in his former research or taken from experience gained on CMU's team the previous year. The ladar mapping he used, I know I saw on some former page of his that had all the gory algorithm details. It might just take a little bit of searching. He also has a c library out there somewhere that does a lot of this stuff, but I can't seem to find it now.

One paper that's of interest might be here: http://www.cs.cmu.edu/afs/cs.cmu.edu/user/thrun/pu blic_html/papers/thrun.ces-tr.html [cmu.edu] (sorry, no linky, writing in a hurry)

And that paper is mentioned in the readme of the BFL (Bayesian Filtering Library) found here:
http://people.mech.kuleuven.be/~kgadeyne/software/ bfl-trunk/ [kuleuven.be]

Lastly, at one point all of us competitors were required to give our design documents to DARPA, and they put them up on their webpage here:
http://www.darpa.mil/grandchallenge05/techpapers.h tml [darpa.mil]

BTW, I wasn't on Stanford's team, but I was on another finalist team.

Re:Sensors

[SpyPlane]

Honestly, not really. It was so damn dry out there that they water would spray the dust off and dry off in no time. I'd say rarely though did we ever see the water system turn on. Really, only in our mud testing did we ever get major buildup. Those LADAR's were pretty resilient sensors. The sun shining in them was much worse than any dust buildup.

Re:I'm a geek, so I watched this twice last night.

[thequux]

Sorry to break up the party, but the second race was MUCH easier then the first. For the first 7 or 8 miles, each vehicle was in a dry lakebed. Comapre this to the ravines and washes that were in the first 7 or 8 miles of the last course.

Why did they make it easier? My personal theory is the act of congress that calls for 2/3 of the armed forces to be autonomous vehicles by 2008 (or something of the sort; I'm probably wrong about the date).

By making people win (not to denigrate their achievement... debugging an autonomous vehicle is no mean feat!), DARPA has robbed the rest of the teams of a fair shot.

TerraHawk, in particular, was designed for a much mure brutal course. It was not the fastest vehicle, but in the terrain that we expected, it wouldn't need to be. (actually, the 5 m/s speed cap was in software... we limited it for safety reasons, as well as the trash can we murdered at the Site visit when we tried to raise the speed a bit :-)

We had problems because, In an effort to deal with that kind of course, we were replacing components right up until the end. (and we did our first test of the new vehicle the day before we went up to Fontana. The old vehicle was well tested, but it had issues with the pneumatics, and we were willing to risk failure in order to get a better chance of success.)

Now that I'm partially off topic, I might as well go the rest of the way.

I am most probably the only person in the world who has worked on no less than THREE teams (PVRW, Team Tormenta, and Terra Engineering). But, I also noticed that is was the big money teams that got through the NQE and on to the main race. Interestingly enough, after the race, when DARPA refused to announce the winner immediately, conspiracy theorists were arguing (with reason) that DARPA was trying to find some technicality to let the Red team win; 8 hours later, they realized that there was no such technicality, and their favored team lost.

Now, before that's marked as flaimbait, keep in mind that this is coming from someone who hasd been involved with this for 2 years, and who noticed the beaurocracy involved.

As a final illustration of this beaurocracy, at the gate to the team garages, ther was a seperate entrance fro people. For about 25 feet on the public side of the gate, there was a portable barrier set up so separate the footpath and the path for vehicles. On the other side, however, there was absolutely nothing. Now the guard was rather strict about the "humans on the human path, and 'bots on the 'bot path" rule, een when someone was coming out the the team area to go to something right next to the gate.

Comaring this to the creativity shown by the teams, you really had to wonder: who was really more organized?