Skip to Content

Prelude to a Robot Race

On Saturday, 11 autonomous vehicles designed by the best roboticists in the world will compete for a $2 million prize in the Urban Challenge.
November 2, 2007

This week, robotic cars overran the desert town of Victorville, CA. Thirty-five autonomous vehicles, capable of piloting themselves without any human aid, came to compete in the Urban Challenge, a robot race sponsored by the U.S. Defense Advanced Research Projects Agency (DARPA). The robots’ goal is to autonomously–and safely–navigate city streets, avoid obstacles in the roads, merge with traffic, park in a crowded lot, and perform other tasks on the roads of the former George Air Force Base. The robotic car that completes the course with the most points for speed, accuracy, and style will win a $2 million prize. Two runners-up will receive $1 million and $500,000.

Boss: The vehicle from Carnegie Mellon University, named Boss, qualified for the final Urban Challenge event on Saturday. Boss uses novel software that takes data from sensors and maps out a 3-D world as it drives.

(See robotic cars from Stanford, MIT, and Carnegie Mellon driving autonomously.)

The Urban Challenge is the follow-up race to DARPA’s Grand Challenge races, which occurred in 2004 and 2005. The previous races tested a vehicle’s ability to drive on an empty desert road. The Urban Challenge is much more complicated, which is a testament to the rapid advances of robotics and autonomous-vehicle technology over the past few years. DARPA hopes to use technology developed for these competitions to build combat and convoy vehicles that don’t need human drivers. In addition, many researchers in the field believe that elements of these technologies could be incorporated into consumer cars in the next five years to help drivers avoid accidents and to assist those who are impaired.

On Wednesday, the robots finished the National Qualifying Event, a set of pretrials designed to winnow the field to, at most, 20 cars. Yesterday, DARPA announced that only 11 vehicles were safe enough to qualify for the final event on Saturday. (See a slide show and video of six of the finalists.) The final course will be challenging, as live drivers and robots will be sharing the road. At a press conference on Thursday, Tony Tether, the director of DARPA, said that one of the obstacles that the robots must face could be a traffic jam. Ultimately, he said, the winner will be the car that completes the courses with the best time but still qualifies for a California driver’s license.

Multimedia

  • See images of the robotic cars at this year's Urban Challenge.

  • Mike Montemerlo of Stanford explains why driving is a "software problem."

  • John Leonard of MIT talks about the supercomputer that runs his team's robotic car, and also frets about qualifying for the final race. (It did.)

  • William "Red" Whittaker of Carnegie Mellon describes the 3-D modeling software used by his team’s car.

  • Brian Schimpf of Cornell explains the decision to make a robotic vehicle that looks like a traditional car.

  • Daniel Lee of the University of Pennsylvania talks about converting a Toyota Prius into a robot.

  • Charles Reinholtz of Virginia Tech argues that technologies exhibited at the Urban Challenge will be incorporated into future cars.

Saturday’s event will start at 8:00 A.M. Pacific Standard Time and will be webcast live at the Urban Challenge website.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.