The great challenge of 2004 Darpa was amazing failure. The Advanced Defense Research Projects Agency has provided a $ 1 million prize for the team that can design an independent ground vehicle capable of completing a rough road through flat desert terrain at times, sometimes. like IEEE SICTRUMIt was reported at that time, it was “the most variety of vehicles collected in one place since filming Mad Max 2: The Road Warrior.“No participant made it on the finish line. Some did not get out of the car park.
Comedy attempts, although any laughter comes at the expense of many engineers who spent countless hours and millions of dollars to reach this point.
Therefore, it is striking that in the great challenge in DarPa, the second, after a year and a half, has crossed five finishing line vehicles. Stanley, his developer Stanford race teamHe won the first place to claim a wallet worth two million dollars. This is the Volkswagen Taleg modification [shown at top] I completed a circuit of 212 km in 6 hours, 54 minutes. The sandy storm in Carnegie Mellon and H1GHLANERER ranked second and third, respectively, at times of 7:05 and 7:14.
Kat-5, sponsored by Gray Insurance Co. Of Metairie, Los Angeles, fourth place with 7:30. The car was named after Hurricane Katrina, who had just struck the Gulf Coast a month and a half ago. Terramax also ended from Oshkosh Truck the circuit, although its 12:51 time exceeded the 10 -hour time limit that DarPa set.
How did the big challenge from a total busts move to five strong finishes in such a short period of time? It is definitely a testament to what can be accomplished when engineers rise to the challenge. But the result of this race was preceded by a much longer research path, in addition to a little luck, which eventually led to victory.
Before Stanley, there was Mineeva
Let’s bloc until 1998, when the computer world Sebastian Thrun He was working in Carnegie Mellon and a completely different robot experience: a museum’s tour guide. For two weeks in summer, MineevaWhich looked like a Dalik From “Doctor Who”, move on an exhibition in Smithsonian National Museum of American History. Its main task was to rotate and dispense with the information nuggets about offers.
https://www.youtube.com/watch?Mineeva was a robot for the tour of the Tura Sebastian Museum.
In an interview at the time, Thrun admitted that Mineeva was there for entertainment. But Mineeva was not just a people. It was also the experience of machine learning. She had to learn where the maneuver could be safe without a visitor or an invaluable artifact. Visitor, unpopular; Display box, non -jamming case; Open floor, non -open floor. She had to interact with humans who crossed in front of him in unexpected ways. She had to learn to “see.”
Fast-FO-FIVES: Thrun moved to Stanford in July 2003. Inspired by the first big challenge, organized the Stanford race team with the aim of introducing an automatic car in the second competition.
In the excessive simplification of the main task in Stanley, the independent robot had to differentiate the road and not to successfully move the road. The Stanford team decided to focus its efforts on developing programs and using the largest possible number of ready -made devices, including the laser to survey the immediate terrain and a simple video camera to wipe the horizon. Programs overlap with both entrances, adapt them to changing road conditions during flying, and determine the speed of safe driving. (For more technical details about Stanley, check out TeamAbout 100,000 lines of the code did it and much more.
Stanford did not enter the great challenge of 2004 and was not expected to win the 2005 race. Meanwhile, Carnegie Mellon had two entries – in 1986, Humve and 1999 Hummer modified – was the clear preferred. In the 2004 race, the sandstorm of CMU was further, and completed 12 km. For the second race, CMU brought an improved sandstorm as well as a new vehicle, H1GHLANER.
In 2004, many other competitors reassemble their ranks to try again, and new in the battle entered. In total, 195 teams were applied to compete at the 2005 event. The teams included students, academics, industry and amateur experts.
After the site’s visits in the spring, 43 teams arrived at the qualified event, which was held September 27 until October 5 in the California motorcycle race, in Fontana. Each vehicle took four runs during the training course, moving through checkpoints and avoiding obstacles. A total of 23 teams were chosen to try the main path through the Mohafi Desert. Competition was expensive – the Red CMU team spent more than $ 3 million in its first year – the names of the shepherds were sprayed through vehicles such as car slogans on racing cars.
In the early hours of October 8, the final candidates for the Grand race gathered. Each team had an overlapping time to help avoid crowding along the way. About two hours before the team’s start, DARPA gave them a pressed disk containing about 3000 GPS coordinates that represent the course. Once the team reached Go, it was the case: the car had to drive itself without any human intervention. NOVA PBS produced an excellent loop on 2004 and 2005 Great Challenges I am highly recommended if you want to feel excitement, expectation, disappointment and victory.
In the 2005 big challenge, H1GHLANERER at Carnegie Mellon University was one of five independent cars to finish the race.Damian Duvarghan/AP
H1GHLAnder finished the pole, ranked first in the qualifying rounds, followed by Stanley and Sandstorm. H1GHLANERER is progressing early and quickly has great progress. This is the place where luck came, or rather its absence.
After about two hours of the race, the H1GHLANER is slowing down and the trader began to the bottom of the hill. Although she ultimately resumed moving forward, she did not restore her maximum speed, even in long and straight sections of the cycle. Stanley caught slower but heavy to H1ghlander in the 163 km sign (101.5 miles), passed it, and did not leave the initiative.
What happened in H1GHLANER remained a mystery, even after a widespread post -analysis. It was not until 12 years after the race – and again with a little luck – CMU discovered the problem: pressing a small electronic filter between the engine control unit and fuel injection caused energy loss and even turning off. The team members speculated that a few weeks ago had affected the candidate competition. (To learn more on how to finally discover CMU, see range The story of the editor -in -chief Ivan Akraman for the year 2017.)
The legacy of the big challenge to Darba
Regardless of those who won the big challenge, many success stories came out of the competition. After a year and a half of the race, Thrun has already made great progress in combating adaptive seizure and helping to maintain corridors, which are now easily available in many commercial vehicles. Then he worked on a view on Google Street and its initial self -driving cars. The CMU Red team worked with NASA to develop Rovers to explore the moon or remote planets. Closer to the home, they have helped develop their self -harvest for the agricultural sector.
Sepastian Trion, the Stanford team leader, holds a $ 2 million check, which is the winning award for the big challenge of 2005.Damian Duvarghan/AP
Of course, there was also a lot of noise, which was tending to overcome the military origins of the race – remember “D” in DARPA “defense”. Again in 2000, a Defense License Bill It was stipulated that a third of the US drone combat cars by 2015, and DarPa depicts the great challenge to stimulate the development of these independent vehicles. The US military was still fighting in the Middle East, and DarPa promoters believed that self -driving vehicles would help reduce injuries, especially those caused by improvised explosive devices.
DarPa sponsored more competitions, such as the 2007 urban challenge, as vehicles moved in a mitment and an environment in the suburbs; Robots Challenge for the year 2012 for disaster response robots; And the underground challenge for 2022 – for you – I guessed it – can wander under the ground. Despite competitions, continuous military conflicts, and huge government contracts, actual progress in independent military vehicles and robots did not take off to the required limit. As of 2023, automatic ground vehicles made up only 3 percent of the global armored car market.
Today, there are very few self -ruling wild vehicles in the US military; Instead, services have advanced forward with semi -protected systems, with the help of the operator, such as drones that are controlled by remote and ships dispersal. One Grand Challenge Finisher, who continued to work for the American army was Oshkosh Truck, the Wisconsin -based shepherd for Terramax. The company showed a system of loading a mobile transport platform in uninhabited vehicles to the American army.
Many contemporary reports on the big challenge expected that self -driving cars will take us closer to the future of “Jetsons”, with a self -driving car for you. But after two decades of Stanley, the offering of civilian independent cars was limited to specific applications, such as Waymo Robotaxis that transport people all over San Francisco or Grubhub Starships that struggle to serve food through my university campus at South Carolina University.
I will watch to see how technology develops outside big cities. Self -driving vehicles will be great for long distances on empty rural roads, but parts of rural America are still struggling to get enough mobile coverage. Will the small cities and areas surround them get a frequency range to accommodate independent vehicles? As much as I would like to think that almost self -driving cars here, I don’t expect to find one under my garage any time soon.
The story of two stanleys
It was not long before the 2005 race, Stanley was ready to retire. He recalled his experience in the Mineeva test at the National Museum of American History, Thrun believed that the museum would make a nice house. Loving it to the museum in 2006, and since 2008 I have set up Permanently in the museum collections, along with other great samples in robots and cars. In fact, not even the first Stanley in the group.
Stanley is now residing in the collections of the National Museum of the American History of the Smithsonian Foundation, which also includes another Stanley – this 1910 Stanley Renapot. Behring Center/National Museum for American History/Smithsonian Foundation
This distinction belongs to 1910 Stanley RunaboutEarly steam -powered car that was not clear after the internal combustion engine was the way to go. Despite the clear defects – Steam engines were tending to the explosion – “Stanley Steamers” was known for their wonderful letter. Farid Marriott He set the ground speed while driving Stanley in 1906. It reached 205.5 km per hour, which was much faster than the average Stanley speed in the twenty -first century of 30.7 km/h. In order to be fair, Marriott Stanley was racing on a flat and straight path instead of terrain on the rough roads transmitted by Stanley.
Throughout the century that separates two stanleys, it is easy to track progress. It is clear that both of them can be identified as landmark cars, but I doubt that as soon as the science fiction in the early twentieth century was subjected to severe pressure to imagine the set of technologies that would drive a self -driving car in the twenty -first century. How will the vehicles of the twenty -second century be? Will they even have four tires, or will they run on something new?
Part of a Continuous chainLooking at the historical artifacts that adopt the potential of boundless technology.
A brief version of this article appears in the release of printing in February 2025 as “slow and fixed to the race.”
From your site articles
Related articles about the web