AIA: Camera Is Brain and Eyes of Robotic Vehicle
The U.S. Army hopes to automate one-third of its ground forces by 2015. Autonomous vehicles will carry supplies to troops or perform reconnaissance tasks like the unmanned aerial vehicles used by the Air Force. These vehicles must be able to travel through 200 miles of rough desert terrain and weather.
To foster development of autonomous vehicles, the Defense Advanced Research Projects Agency (DARPA) hosted a race, the Grand Challenge, for autonomous vehicles on March 13.
Fifteen vehicles were allowed to compete. Within the first few minutes and only tens of meters, eight vehicles broke down. Team Red's vehicle made it the furthest at 7.3 miles. But, it used a backup laser range finder that did not have the resolution needed to spot a small rock that broke the vehicle. The teams that made it the furthest used camera systems to augment other terrain sensing systems.
The Grand Challenge demonstrated the need for truly intelligent camera systems. That's where Team Armani and the VC2028 smart camera from Vision Components (Hudson, NH) come in. Team Armani is building an autonomous vehicle for DARPA's next Grand Challenge, and the smart camera will serve as the vehicle's eyes and brain. The camera can produce a 3D image of the environment even while the vehicle is moving. The VC2028's digital signal processor (DSP) can then analyze the 3D data to find the best path.
The VC2028 is a powerful smart camera, and it is compatible with standard lenses and stands. It does not require cooling, even in the desert. The camera's sturdy housing prevents moisture and shock damage. Most importantly, it carries a DSP that provides the equivalent of 1 gigahertz of computing power. DSP programs can be copied to other cameras quickly.
Team Armani is making significant progress toward completing their autonomous vehicle. In the first month of work, Team Armani has successfully tested the vehicle's autonomous path-following and power-balancing algorithms. It has been able to configure and program the VC2028's computer interface, overlay display and image processing routines. It has also developed algorithms for taking images, real-time object tracking, shutter control and focusing. The final challenge is to optimize and test all the programs for real-time speed.
Team Armani hopes to enter the newly founded International Robotic Racing Federation race in October 2004, and plans are being made for another DARPA Grand Challenge.
For more information on smart cameras call 603-598-2588 or visit www.vision-com.com.