Opportunity Follows Spirit to Mars

As NASA’s Opportunity rover streaked toward Mars for its 9:05 landing this evening, its crippled twin, Spirit, began transmitting data to Earth again -- albeit very slowly -- after nearly 48 hours of relative silence.

Spirit, which landed on Mars on Jan. 3, sent back about 30 minutes’ worth of engineering data. That was enough to tell controllers at Pasadena’s Jet Propulsion Laboratory that all its major subsystems were intact and operating, but not enough to tell them what was wrong with it and why it stopped communicating.

Engineers do not yet know how long it would take to repair Spirit, or even if it can be repaired.

“I expect that we will not be restoring functionality to Spirit for many days, perhaps even for a couple of weeks, even in the best of circumstances,” said project manager Peter C. Theisinger.

“We have a serious problem,” he said, but the vehicle was in no immediate danger. He said it will be stable “for an extensive period of time that will allow us to work on it.”

Theisinger expressed confidence that the rover would be repaired, but noted that there may be residual effects from the incident. “The chances that it will be perfect again are not good,” he said. But he added: “The chances that it will not work at all are also low.”

Meanwhile, half of the team of 300 scientists and engineers that has been monitoring Spirit on Mars is splitting off to take charge of Opportunity. The second rover will be touching down in Meridiani Planum, halfway around the world from Spirit’s Gusev Crater landing site. Opportunity will join its sister craft and three orbiters in the search for signs of past surface water on the Red Planet.

Controllers can do little today, however, except wait for Opportunity to reach Mars. All its actions are programmed in its onboard memory. But controllers are working feverishly to understand what went wrong with Spirit to prevent the same thing from happening to its twin.

The first step toward understanding Spirit’s malfunction began Friday at 5:02 a.m., when JPL controllers used the Deep Space Network antennas outside Madrid to send the rover a command to phone home. Spirit responded at 5:26 a.m., sending back a 20-minute message containing information on its system status.

But the data was transmitted by the rover’s low-gain antenna -- used when the craft detects a problem and goes into “fault” mode -- and the data rate was only 120 bits per second. When the craft is using its high-gain antenna, by contrast, the data transfer rate is greater than 11,000 bits per second.

Later, there was a 10-minute session, again at the slow rate, with additional status data.

Those transmissions indicated that Spirit’s onboard computer had been rebooting itself repeatedly since Wednesday morning -- at least 60 times by Friday morning. Each time the computer woke up from the reboot, Theisinger said, it would discover a condition that would cause it to wait anywhere from 15 minutes to 60 minutes, then reboot again.

Apparently the craft is seeing a different reason to reboot each time, he added. That fact is very confusing to controllers trying to trace the problem.

The data also indicated that Spirit was not going into its sleep mode at night. The craft is programmed to halt most of its activities during the long, cold Martian nights to conserve its batteries. By continuing to operate during planned sleep periods, the rover risks depleting them.

That is not a significant problem, Theisinger said, because the batteries could be recharged even after they were fully depleted. He said it would be better for the team if the rover did not reach such a state, “but we can cope with it if it does.”

At the end of the day on Mars, the team sent Spirit a command telling it to go to sleep for the night so that the batteries could recharge. But engineers do not yet know if the craft received the command or acted upon it.

Finally, the data also indicated that the software problems began while the rover was moving a mirror attached to its infrared spectrometer, a device the scientists call mini-TES. That operation was not completed, Theisinger said, so it was possible that the problems were triggered by some kind of hardware fault -- either in the motor or in the sensors that tell the computer the mirror’s position.

In past missions, engineers have been able to work around hardware problems, and the team hoped they would be able to do so in this case if that was, indeed, the root of the problem.

The team did get one good piece of news Friday, a dramatic photograph of the landing site produced by the Mars Global Surveyor 240 miles above the surface.

The highly detailed image shows not only the lander, but also the parachute, the rocket assembly used to slow the lander’s descent and the skid marks produced by the heat shield when it crashed and slid down the slope of a nearby volcano.

“That’s all the hardware we sent,” said Michael Malin of Malin Space Science Systems, the principal investigator for the Global Surveyor’s camera systems.

The image even shows the imprints left by the lander’s air bags when it bounced to a stop.

Malin said that he has been in touch with the British team that built the Beagle 2 lander, which was scheduled to reach the Martian surface on Dec. 24 but has not been heard from since. He said they would try to use the camera to locate it.

In another piece of good news, European scientists said Friday that Beagle 2’s mother ship, Mars Express, had used its infrared imager to directly detect small amounts of water in the Martian atmosphere. The finding confirms observations by other orbiters inferring the presence of frozen water under the surface.