USA-193: reboot problems

U.S. NRO spy satellite may be total loss
Wed Mar 7, 2007 10:17 AM IST
By Andrea Shalal-Esa

WASHINGTON (Reuters) - U.S. officials are likely to declare a Lockheed
Martin Corp. spy satellite a total loss after efforts to restore its
ability to communicate failed repeatedly over the past three months,
two defense officials told Reuters on Tuesday.

The experimental L-21 classified satellite, built for the National
Reconnaissance Office (NRO) at a cost of hundreds of millions of
dollars, was launched successfully on Dec. 14 but has been out of
touch since reaching its low-earth orbit.

Limited data received from the satellite indicated that its on-board
computer tried rebooting several times, but those efforts failed, said
one official, who is knowledgeable about the program and spoke on
condition of anonymity.

The satellite carried sophisticated cameras to take high-resolution
pictures and test equipment intended for use on the broader Future
Imagery Architecture (FIA) program, in which both Boeing Co. and
Lockheed are involved.

Its failure raises questions about the schedule for the already-much-
delayed FIA program, which was due to launch a first satellite in two
to three years, analysts said.

One of the defense officials acknowledged the satellite's failure was
"not helpful."

"It's part of an overarching architecture. When you're trying to move
forward on several dimensions, it can't help accomplish those goals,"
the official said.

The other official said he expected schedule adjustments, but no major
delays, as a result of the NRO satellite failure.

"It might impact the schedule for introduction of new technologies,"
he added.

Another government official said he was unaware of any changes to the
FIA program as a result of the satellite issue.

Lockheed, prime contractor for the experimental NRO satellite,
declined to comment. The NRO, which designs, builds and operates
reconnaissance satellites for the U.S. military and intelligence
communities, also had no comment.

One of the defense officials said the issue with the satellite
involved the computer that runs it, not the new sensors that it was
meant to test.

"The failure has nothing to do with anything new. It happened with a
set of components ... that historically is known to be good," said the

Jonathan McDowell, an astronomer with the Harvard- Smithsonian Center
For Astrophysics, said the satellite's software problems raised
questions about the adequacy of testing and oversight by the
contractors and the Air Force.

"The question is why was this software failure not caught in ground
test before launch," McDowell said, noting that oversight was
particularly challenging in classified programs.

He said the satellite's software woes were reminiscent of those
experienced by the Mars rover named Spirit, which was out of
communication for more than two weeks after it landed on Mars in
January 2004 because its flash disk kept filling up, prompting the
computer system to crash repeatedly.

Engineers finally solved the problem by sending a command to the
computer to clear the disk, enabling a successful rebooting of the
system, he said.