Jump to content

Talk:2005 DARPA Grand Challenge: Difference between revisions

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Content deleted Content added
Shakey
No edit summary
Line 43: Line 43:


More than ten years ago a fast vision-guided autonomous [[Mercedes]] robot programmed by the team of [[Ernst Dickmanns]] (the pioneer of robot cars) already performed pretty amazing feats - the Grand Challenge doesn't look so impressive in comparison. I was surprised that I could not find a decent Wikipedia article about Dickmanns, and wrote a first draft. [[User:ERDI|ERDI]] 20:53, 5 July 2006 (UTC)
More than ten years ago a fast vision-guided autonomous [[Mercedes]] robot programmed by the team of [[Ernst Dickmanns]] (the pioneer of robot cars) already performed pretty amazing feats - the Grand Challenge doesn't look so impressive in comparison. I was surprised that I could not find a decent Wikipedia article about Dickmanns, and wrote a first draft. [[User:ERDI|ERDI]] 20:53, 5 July 2006 (UTC)

==Carnegie Melon cheated?==

I was told by someone familiar with the competition that Carnegie Melon cheated because of the pressure on their team. Something about hiring a helicopter to overfly the course, or something. I don't know the details, so I'm just asking if anyone else had heard of these rumours.

Revision as of 06:34, 9 January 2007

Correcting the Times

"The Stanford Racing Team was the first to cross the finish line, with a time of 7 hours and 54 minutes. Red Team Too finished soon after with a time of 7 hours and 59 minutes. The third team to cross the finish line was Read Team, with a surprise upset at a time of 7 hours 52 minutes. It is important to note, however, that due to the staggered start, the order in which they crossed the finish line is not indicative of their place. Note: These results are not official."

Where are these results from? If they're directly from the Grand Challenge.com flash interface, they're incorrect, as the the time hasn't been stopped, and is continuing to run.

I was taking the times from the flash interface, and you are right, they are wrong. I'm changing to times to match a news article (http://www.gizmag.com/go/4720/). --mdd4696 21:29, 8 October 2005 (UTC)[reply]


Stanley Hype

Local Patriotism

This site and others link to a rather one-sided Wired magazine article on the DARPA race winner Stanley (vehicle), and in the Stanford AI lab home page I found this announcement of another one: "Dec 29, 2005: Stanford built three of the top ten robots ever! According to a recent evaluation by Wired Magazine, three of Stanford's robots were among the top ten robots ever: Stanley (Number 1), Shakey (Number 5), and the Stanford Cart (Number 10). Wired Magazine polled numerous experts to determine the 50 Best Robots Ever. Check it out!"

The problem is, of course, that Wired magazine is based in San Francisco and has strong ties to Stanford. Local patriotism may be fun, but few if any unbiased roboticists would agree with that list.

So what kind of "experts" did they poll? Maybe cartoonists, since Number 2 is a fictional Japanese comic strip robot. The other cars that finished shortly after Stanley are not even mentioned...

Presumably the list is not meant to be taken seriously. Any serious list would be dominated by real Japanese robots (and would not even mention fictional ones), since Japan clearly dominates robotics research and has 40 percent of the world's robots, including many of the most expensive and sophisticated and famous ones. I don't think any mere car would rank among the top 5. But don't expect Wired magazine to publish such a list!

Nevertheless, Wikipedia articles such as this one might want to link to a bit more objective sources. De-Hyping Stan 16:54, 28 March 2006 (UTC)[reply]

Agree, though bias can be hard to spot. I would agree with the fact that Stanly is a top ten robot (though I think the Mars rovers beat it) - Ravedave 18:05, 28 March 2006 (UTC)[reply]
Well, even if we accept the (in my opinion questionable) premise that Stanley should be in the top 10 list, shouldn't the vehicles of, say, CMU appear there, too? One of them led the race and apparently just lost due to a technical glitch, finishing just a few minutes after Stanley. I guess Wired magazine would have included them if CMU was a local Californian university... De-Hyping Stan 21:42, 29 March 2006 (UTC)[reply]
Shakey wasn't even built at Stanford; Shakey was built at SRI International, a think tank near Stanford. I saw that Wired list; it was weak on Japanese research robots. Still, I was quite impressed with the Stanford grand challenge entry, which represented a significant breakthrough in computer vision. The CMU effort was basically an attempt to throw money at the problem, and wasn't really that innovative. --John Nagle 06:25, 18 August 2006 (UTC)[reply]

Chicken and Eagle

How much of a challenge was it anyway? The course was precisely layed out by numerous waypoints. No sophisticated planning procedure was necessary. Four separate teams managed to finish the course in time. This seems to suggest that the task was not that hard.

In fact, several days before the race the Stanley team leader announced that this time with very high probability there would be a winner, thus reflecting the general feelings of all the experts. Afterwards, however, he fueled the hype by claiming, rather inconsistently: "the impossible has been achieved!" Then he compared himself to Charles Lindbergh (no joke!), according to the Wired magazine article, which also says the "Wright Brothers / Kitty Hawk spot" belongs to Ernst Dickmanns (1986). Lindbergh! Like the chicken claiming "I'm an eagle!" One of the most ridiculous comparisons I've seen lately. Now waiting for the confetti parade :-) De-Hyping Stan 19:15, 30 March 2006 (UTC)[reply]

Just to clear some things up, first, all 4 teams that completed the race used relatively sophisticated obstacle detecting sensors and path planning software. Although the waypoints were given, the path contained obstacles and cliffs. Second, the Stanley team leader's announced belief that there would be a winner reflected his insider opinion. His statement that "the impossible has been achieved" refers to how hard a problem this was to solve, stumping researchers for a decade. The statements only contradict when taken out of context.

Mention Foreign Contributions?

The original DARPA web site was headed by the slogan "Harnessing American Ingenuity." But perhaps the Wikipedia article should also mention some of the foreign contributions? For example, the University of Parma (Italy) contributed to one of the cars that finished the course. And the winner was a German car with a German team leader. (Who promptly compared himself to Lindbergh - see section chicken and eagle above.) De-Hyping Stan 19:15, 30 March 2006 (UTC)[reply]

1995 S-Class Robot Car of Dickmanns

More than ten years ago a fast vision-guided autonomous Mercedes robot programmed by the team of Ernst Dickmanns (the pioneer of robot cars) already performed pretty amazing feats - the Grand Challenge doesn't look so impressive in comparison. I was surprised that I could not find a decent Wikipedia article about Dickmanns, and wrote a first draft. ERDI 20:53, 5 July 2006 (UTC)[reply]

Carnegie Melon cheated?

I was told by someone familiar with the competition that Carnegie Melon cheated because of the pressure on their team. Something about hiring a helicopter to overfly the course, or something. I don't know the details, so I'm just asking if anyone else had heard of these rumours.