|

Fiber-Optic Gyros Help Robotic Vehicles Race
Across Desert
Created in response to a Congressional and Department
of Defense mandate, the Defense Advanced Research Projects
Agency (DARPA) Grand Challenge is a field test proposed
to expedite research and development of autonomous ground
vehicles to assist on the battlefield.
The objective is to successfully race two driverless
Hummers against other teams across 175 miles of Mojave
Desert terrain in 10 hours or less.
Out of an initial field of 118 teams from around the
U.S. and Canada, two teams (dubbed Red Team and Red
Team Too) from Carnegie Mellon's Robotics Institute
have autonomous vehicles (see Figure 1) among the 40
semi-finalists competing in the DARPA Grand Challenge
national qualifying event in September 2005. From there,
only 20 teams go on to compete for the grand prize on
race day, October 8, 2005.
 |
| Figure 1. Fiber-Optic Gyros Help Stabilize
Navigation Systems on the Sandstorm and H1ghlander
autonomous vehicles. |
One of the vehicles, a radically modified 1986 Humvee
called Sandstorm, is the returning Grand Challenge recordholder.
Upgrades for the 2005 season include a new engine and
drive-by-wire, fast computing, and a million lines of
solid software. Sensors include lasers, radar, and single-camera
road following. The second vehicle, H1ghlander, is a
1999 H1 Hummer that incorporates new technology, viewing
terrain with seven laser scanners, three cameras, and
two radar sensors. Several of these are pointed and
stabilized by a three-axis gimbal using fiber-optic
gyros (FOGs).
Both Carnegie Mellon vehicles use DSP-3000 FOGs (see
Figure 2) on a stabilized sensor platform to sense pitch,
roll, and yaw angular rates relative to the vehicle's
position. Data from the FOGs is used to stabilize the
vehicle's navigation and guidance system regardless
of uneven terrain and vehicle motion.
 |
| Figure 2. The Fiber-Optic Gyros have a
patented, all-fiber design incorporating digital
signal processing. |
These FOGs have a patented all-fiber design (fabricated
from proprietary E.Core polarization-maintaining fiber)
and Digital Signal Processing (DSP). DSP electronics
improve performance in such critical areas as scale
factor and bias versus temperature, scale factor linearity,
turn-on to turn-on repeatability, and maximum input
rate. The DSP design also virtually eliminates temperature-sensitive
drift and rotation errors. The FOGs can be used in diverse
commercial and defenserelated applications.
Learn more about Carnegie Mellon Robotics Institute's
(Pittsburgh, PA) Sandstorm and H1ghlander autonomous
vehicles online at http://info.ims.ca/5215-334.
For more information on KVH Industries, Inc.'s (Middletown,
RI) DSP-3000 fiber-optic gyros and related patents,
contact Chris Watson, corporate communications manager,
at (401) 847- 3327 or cwatson@kvh.com,
and visit http://info.ims.ca/5215-335.
|