[meteorite-list] Ground Teams Struggle to Save Mars Orbiter from Itself (MRO)

From: Ron Baalke <baalke_at_meteoritecentral.com>
Date: Mon, 9 Nov 2009 16:57:39 -0800 (PST)
Message-ID: <200911100057.nAA0vdYF010782_at_zagami.jpl.nasa.gov>

http://www.spaceflightnow.com/news/n0911/07mro/

Ground teams struggle to save Mars orbiter from itself
BY STEPHEN CLARK
SPACEFLIGHT NOW
November 7, 2009

Although engineers are still weeks from uplinking new command logic to
eliminate an unlikely, but potentially fatal, scenario jeopardizing the
Mars Reconnaissance Orbiter, the mission's project manager said Friday
he is confident the $720 million mission will resume soon.

After four science-halting computer resets this year, officials in
charge of MRO decided to keep the probe in safe mode until they could
ensure the glitches would not threaten the mission.

Engineers are pursuing two paths of analysis to reach a solution to the
reset problem.

One group is trying to devise a fix to be uplinked for the spacecraft to
tell itself it is at Mars. Another is investigating the root cause of
the events.

In one far-fetched but plausible scenario, MRO could revert to its
pre-launch mode and attempt to make a hardline connection with ground
controllers more than 100 million miles away.

"If we had the same kind of resets that we've seen four of this year,
but you get more severe ones and you get them too close together, you
could have the vehicle forget that it's in mapping orbit around Mars and
instead think that it's still on the launch pad and only communicate
through an umbilical cable, which isn't long enough to get there
anymore," Erickson said.

Four major causes for the resets are being studied, according to Erickson.

The candidates include a momentary glitch removing power from a
component in the computer, a problem with reference voltage, radiation,
or a grounding issue.

"It's more than likely to be one of those four things or a flavor of
them," Erickson said.

According to Erickson, it will be several weeks to a month before MRO is
ready to gradually return to normal operations. The spacecraft has not
been conducting science observations since its last computer reset Aug. 26.

Since late August, MRO has been in safe mode with its solar array
tracking the sun for power and its antenna pointed at Earth to maintain
communications.

"It's the safest condition for the spacecraft, so we said just leave it
there until we get a better handle on this," said Doug McCuistion,
director NASA's Mars exploration program.

In early September, officials said they expected it to take a few weeks
to recover MRO and resume operations, but it has now spent nearly three
months in safe mode.

Erickson said engineers are testing the new algorithms on the ground
before uplinking them to MRO.

"We take our jobs of protecting this vehicle really seriously," Erickson
said. "It's a really important asset to the American people. When we
find something like this, we try to make sure that it can't happen
(again)."

The worry is that MRO could experience two computer resets, more severe
than any of the glitches so far, on its primary and redundant control
strings within one minute of each other.

"The first one has to wipe out all information on the side of the
spacecraft it's on now, and cause a side swap to the other side,"
Erickson said. "And then within a minute, we've got to have the same
thing happen, where it wipes out all the information about what mission
phase it's in."

It takes about a minute for the second string to repopulate the first
string with information on MRO's mission phase.

"So you could have resets that are 1 minute and 5 seconds apart and it's
not a problem," Erickson said.

The computer resets began in February, followed by another anomaly in
June and two in August. The increasing frequency of the events concerned
NASA managers.

The fix being designed by NASA involves changing data parameters in
MRO's computer. When the spacecraft reboots, it searches a table in its
nonvolatile memory in the command and data handling unit's computer
module interface card, or CMIC, to determine if the mission is in
pre-launch, launch, cruise, orbit insertion, or mapping mode.

"In all the places where it's going to look, we have inserted only the
possibility to be in mapping," Erickson said.

Engineers are leery of writing to the nonvolatile memory, which is
similar to flash, so officials are being cautious to ensure the fix will
not cause additional problems.

"They've been working hard on the testbed to try to understand the
interactions of the software and the CMIC," McCuistion said.

The new data parameter logic will likely be loaded into MRO's computer
before officials identify the most likely root cause of the resets. It
could take longer for engineers to close out the fault tree.

"They're having these resets due to a problem in the actual command and
data handling system," Erickson said. "It's a hardware problem. These
resets are happening so fast that they leave virtually no trace of
what's causing them to happen."

"It's happening in the nanoseconds to milliseconds range, so there's
nothing. Just some indication that the event was 'this trigger happened'
and that's it," Erickson said.

In a teleconference with the NASA Advisory Council last month,
McCuistion said the problem could stem from a parts aging issue.

Launched in 2005, MRO arrived at Mars in March 2006 and completed its
primary phase of science operations in November 2008.

Once the probe its given a clean bill of health, it will restart science
observations and play a larger role as a communications relay station
for NASA's Spirit and Opportunity rovers. MRO will also be a significant
part of the communications plan for the Mars Science Laboratory, or
Curiosity, rover when it arrives in 2012.

NASA says MRO has returned more scientific data than all previous Mars
missions combined. The orbiter's six instruments include a powerful
telescope-like high-resolution camera, atmospheric sensors, an imaging
spectrometer and a radar to probe the Martian subsurface.
Received on Mon 09 Nov 2009 07:57:39 PM PST


Help support this free mailing list:



StumbleUpon
del.icio.us
reddit
Yahoo MyWeb