Sensors Mag

Robots Take to the Streets

November 9, 2007 By: Melanie Martella, Sensors


E-mail Melanie Martella

On November 3, eleven autonomous robotic vehicles took to an urban course to compete in the DARPA Urban Challenge. Tasked with obeying traffic laws, negotiating streets and intersections, and avoiding manned (and unmanned) traffic, the vehicles were asked to complete three missions, each approximately 20 miles long, within six hours. Six vehicles finished.

Evolution

In 2005, in an earlier challenge, autonomous robotic vehicles followed GPS breadcrumbs over a 150 mile desert course. Five teams finished, with the winner (Stanley from Stanford University) navigating the course successfully in a little over six hours. The Urban Challenge was a far more complex proposition. In the desert, the vehicles had to follow a set route and avoid immobile obstacles and had ten hours in which to do this. In the urban challenge, however, the environment was significantly more complex, the vehicles had to plan and alter their own routes, some of the obstacles were moving, and the vehicles had to move faster while simultaneously obeying traffic laws. I know humans who can't do that!

Science Daily's article "Carnegie Mellon Robotized SUV Wins DARPA Urban Challenge" is an excellent description of the event. More coverage is available at TG Daily and there you will find videos of the competitors. Go watch, because even though these robots are doing amazing things it's clear that they're still nowhere near ready for traffic in a non-controlled environment.

See What the Robot Sees

Tartan Racing, Carnegie Mellon's team, used software to analyze how its vehicle (Boss) was doing. To quote from the Science Daily article,
"One of the team's advantages was a software system it developed called TROCS, which produced graphic animations of Boss's sensor and data inputs during each run. Much as game-day video allows the Pittsburgh Steelers to review and analyze their play, TROCS enabled Tartan Racing to understand what Boss saw as it drove and how and why it responded to its environment. Troublesome behaviors could be quickly identified and fixed, while appropriate behaviors, which might occasionally look odd to an observer, were left untouched."

Put simply, aggregating the data in an easily grasped way allowed the team to understand how their robot was reacting to its environment and to optimize its responses.

Inspiration

There's a wonderful interview with Dr. Tether, the director of DARPA, interviewed by C|Net before the race ("DARPA sees inspiration as trophy of robot race"). Reading this interview gives you a sense of how far these creations have come; watching the videos confirms it. Maybe these robotic vehicles will inspire the same kind of interest and excitement in science and engineering as the lunar landing did in the 1960s. I sincerely hope so.


Add Comment




IIoT University


Deep Learning for Vision Using CNNs and Caffe: A Hands-on Tutorial – 9/22/16 – Cambridge, Mass


IDE






Sensors 2017 Call for Speakers


Sensors Midwest


Advertise


Subscribe



Twitter Feed

Find It Fix It Forum

Sensors invites you to join the Findit-Fixit Forum, where you can get answers to your sensing questions—concerning technologies, products, methods, applications, and services--and also offer help to your fellow engineers. The Forum covers all kinds of topics, from the basics to the extraordinary.

Join the discussion!


© Copyright 2016 Questex, LLC. All Rights Reserved. Sensorsmag. Privacy Policy | Terms of Use

If you are having technical difficulties or considerations, please contact the webmaster.