Advanced Imaging

AdvancedImagingPro.com

   

Advanced Imaging Magazine

Updated: July 8th, 2008 05:26 PM CDT

Vision Systems Face a Grand Challenge

Autonomous Vehicle Systems
The Flying Fox used cameras from SICK to accumulate terrain data.
Team DAD
Team DAD’s vehicle used a laser-based system using laser emitters from OSRAM Opto Semiconductors and DSPs from Texas Instruments.
Courtesy Carnegie Mellon University
Carnegie Mellon University’s Red Team, which includes sponsors such as SGI, Boeing, Caterpillar, SAIC, Intel, and AM General, will compete in this year’s DARPA Grand Challenge with two entries based on AM General’s Humvee vehicle design — “Sandstorm” and “H1ghlander.”
Real Time Innovations
This graph shows how NDDS 4.0 architecture works. NDDS maintains isolation boundaries between application elements and allows them to communicate directly over any available transport. (Legend: A=NDDS provides each DataWriter or DataReader with a queue for messages.)
Advertisement

Hank Russell By Hank Russell
Managing Editor

The Mojave desert is a land of intense heat and rugged terrain. It is also a perfect place to mimic the conditions in Iraq.

The Defense Advanced Research Projects Agency (DARPA, Arlington, VA) annual challenge will award a $2 million contract to develop autonomous vehicles that will operate in these environs. The award goes to the winner of this year’s DARPA Grand Challenge on Oct. 9. (The winner was not announced as of press time.) The team whose vehicle could complete the course (approximately 150 miles long) in the fastest time under 10 hours will design autonomous vehicles for the military to deliver food and medical supplies to troops in Iraq.

Twenty finalists were invited to race. Three of the teams were Team DAD (Digital Audio Drive, Morgan Hills, CA), powered by Texas Instruments (TI, Stafford, TX); Red Team, led by Carnegie Mellon University (Pittsburgh) and powered by technology from Silicon Graphics Inc. (SGI, Mountain View, CA); and the Flying Fox team, built by Autonomous Vehicle Systems (San Diego) and using Network Data Distribution Service (NDDS) middleware from Real-Time Innovations Inc. (RTI, Santa Clara, CA).

A Laser Solution

This year, Team DAD used a laser-based terrain mapping system instead of the stereo vision-based system they used last year. “We abandoned it (the stereo vision system) because of the problems that everyone faces related to vision systems,” says Bruce Hall, Team DAD’s team director and spokesperson. “It doesn’t work at night. It doesn’t work through fog or in glare.”

The laser-based system was housed in a circular drum three feet in diameter and 12 inches high. Each circular pod sitting inside the drum houses eight 905-nanometer laser emitters from OSRAM Opto Semiconductors (San Jose, CA); as the drum spins at 600 rpm to obtain a 360-degree field of view, eight TI digital signal processors (DSPs) fired the lasers at 10-nanosecond bursts. Each laser pulse was fired 2,000 times as the drum made a full revolution.

1 2 3 4 5 6 next


Subscribe to our RSS Feeds