New software preps satellite for potential gyro failure

Scientists and engineers who work with the Far Ultraviolet
Spectroscopic Explorer (FUSE) have pulled off a second daring and
unprecedented rescue of the satellite observatory from serious
guidance problems.

This time, though, they didn’t actually wait for the guidance
problems to happen.

In response to hints of the potential for future new difficulties
with FUSE’s gyroscopes, which are used to check the satellite’s
pointing accuracy, researchers redesigned software for three
computers aboard FUSE and recently uploaded the new software to
the computers.

The staff of FUSE, operated for NASA by Johns Hopkins University,
compared the feat to a “brain transplant.” They currently maintain
detailed control of FUSE’s precise orientation through the gyros’
ability to sense even very small shifts in the satellite’s position.
If too many of the gyros stop working, however, the new software
will allow controllers to switch over to using the fine error
sensor, a camera aboard FUSE, in their place. In the new guidance
mode, detailed information on where FUSE is pointing will be
determined via the positions of key stars imaged in the fine error

Jeff Kruk, principal research scientist in physics and astronomy in
the Kreiger School of Arts and Sciences at Johns Hopkins and deputy
chief of observatory operations for FUSE, said the new “zero gyro”
mode has already been tested and proved to be even more effective at
keeping the satellite precisely pointed.

“We’ve had several periods of a week or so where we’ve taken the
gyros out of the loop and flown on the new software, and the
pointing stability is actually a little better with the fine error
sensor than it is with the gyros,” Kruk said.

But Kruk and Warren Moos, professor of astronomy at Johns Hopkins
and principal investigator for FUSE, cautioned that there’s still
work to be done in fine-tuning and error-proofing the new system.

“Things are going extremely well so far,” Moos said. “We haven’t
found any major problems, but we’re not out of the woods yet.”

FUSE, launched in 1999, has gathered important data about the
universe by analyzing light in the far ultraviolet portion of the
electromagnetic spectrum. The “brain transplant” in April was the
second improvised but extraordinary effort to rescue the orbiting
probe from approximately 500 miles below on Earth. In December
2001, the failure of the second of four guidance system components
known as reaction wheels sent FUSE into a pre-programmed “safe mode”
configuration. In less than two months of intense work, engineers
and scientists were able to bring the satellite back online using
parts known as magnetic torquer bars in place of the reaction

(For the complete story, see )

This year’s pre-emptive rescue and the testing of the associated
software have had little if any impact on FUSE’s scheduled
scientific observations, said Bill Blair, chief of observatory
operations for FUSE and a research professor of physics and
astronomy at Johns Hopkins.

“Since the upload, which took about a week, we’ve been back to
normal science operations,” he explained. “But there’s also been
a long, low-level tail of activity to just kind of optimize things
and track down small problems with the new software.”

The upgrades are a product of nearly two years of work by engineers and
scientists at Johns Hopkins, Orbital Sciences Corp., Honeywell Technical
Solutions Inc., NASA Goddard Space Flight Center and the Canadian Space
Agency. Researchers began to work on a new method for guiding FUSE when
one of FUSE’s six gyros, always anticipated to have a finite lifespan,
went dead unexpectedly early on May 31, 2001. Two gyros were built into
FUSE for each of the three axes of motion. If any axis were to lose both
gyros, controllers would no longer be able to point FUSE precisely.

“We were highly motivated when the first gyro went dead on May 31,” Moos
recalled with a wry laugh. “There have been very, very few attempts to fly
precision-pointed spacecraft without gyros, and learning how was a major
step forward.”

Among the obstacles faced by controllers was developing ways to make sure
information could be sent back and forth quickly enough between FUSE’s
three main computers. Moos compared the process to trying to stop a fall
from a tree–not only is there very little time to sense when an
appropriate branch might be within reach, but the time send a mental
command to reach out and grab that branch is also very short.

Controllers also had to develop a way to deal with the periods when the
guide stars used by the fine error sensor to fix FUSE’s position were
eclipsed by the Earth as FUSE orbited around it. Moos said their
solution depends in part on detailed models of how torque from the
Earth’s gravitational field will twist the satellite, and in part on
readings they could obtain from an instrument aboard FUSE known as a

Kruk added that the new software uploaded to FUSE in April contained
improvements to several housekeeping functions in the satellite, in
part to prepare it for reduced round-the-clock human monitoring as FUSE
enters an extension of its originally planned mission.

“We were able to build in more “smarts” to make FUSE capable of
gracefully handling almost anything that might come up,” Kruk

“Blair concluded, “With these repairs in place, and astronomers from
around the world lining up to use FUSE, the mission is on track for
at least several more years of operations.”

Johns Hopkins University news releases can be found on the World Wide
Web at