They figured it out. Flash memory.
Mars Lander Spirit's Problem Traced
Engineers believe they can restore the craft to nearly full operation in a matter of weeks. Second craft to land tonight.
By Thomas H. Maugh II
Times Staff Writer
January 25, 2004
Engineers have traced the malfunction in the Spirit rover to a computer memory problem and believe they can restore the craft to nearly full operation in a matter of weeks, project manager Pete Theisinger of the Jet Propulsion Laboratory said Saturday.
"We're upgrading its condition from critical to serious," Theisinger said.
Controllers were able to reestablish contact with Spirit overnight Friday and received data indicating that the problem lies in the craft's flash memory, which functions much like the memory sticks in a digital camera.
Theisinger said he was confident that the problem could be either repaired or circumvented.
"It will take us a couple of weeks to get back close to routine operations," he said. "And we are probably three weeks from driving the rover."
Meanwhile, NASA's Opportunity rover was "in perfect order" Saturday as it headed for its 9:05 p.m. encounter with Meridiani Planum on Mars.
At noon, the spacecraft was cruising at about 6,700 mph, but the strong pull of the Red Planet's gravity accelerated it to more than 12,000 mph before it began the fiery "six minutes from hell" descent through the atmosphere.
Controllers hoped to hear from the craft before midnight Saturday, but cautioned that contact could come as late as Sunday evening if the rover came to rest in an awkward orientation with its communications antenna pointed away from Earth.
Mission controllers cautioned that the landing was unlikely to go as smoothly as Spirit's, which proceeded flawlessly onto Gusev Crater on Jan. 3.
The recent problems with Spirit began Wednesday morning, when the craft abruptly stopped transmitting data to Earth. The team subsequently found that Spirit's on-board computer was rebooting itself about once every hour and that the craft was not going to sleep during the long, cold Martian night to conserve its batteries.
The team had some very slow communications with the craft Thursday, then got an unexpected burst of data during the Martian night via the Mars Odyssey orbiter. That data gave them a hint that the flash memory was involved.
When controllers made contact with the craft again, they ordered it to go into a "cripple" mode, which bypassed the flash memory, and then to reboot. After it rebooted, they had a one-hour, low-speed communication session with the craft with no apparent problems. Significantly, reboots did not occur after that, Theisinger said.
The team then ordered the craft to go to sleep, so that its batteries could be recharged.
The rovers have three types of memory: RAM (random access memory), flash and EEPROM (electrically erasable programmable read-only memory).
RAM is similar to that in a home computer, and is used for temporary storage of instructions and data. Any information in the RAM is lost when the system is shut down.
Flash memory provides longer-term storage and is used for scheduling the rover's daily activities and storing scientific information until it is transmitted to Earth. Information stays in the flash memory until it is deliberately erased.
EEPROM is similar to flash memory, but its information is harder to erase and serves as a more permanent form of memory.
The rover's core programming is contained in EEPROM, and it does not seem to be affected by the current problem.
Once Spirit's batteries have been recharged, the team will wake the craft up and place it in cripple mode again — a task that must be carried out each day. They will then set up a high-speed communications link, probably through the Mars Odyssey orbiter, to download the contents of the flash memory so they can figure out what happened.
The team has not yet discounted the possibility that a hardware problem might be the root cause of Spirit's ailment, Theisinger said. The whole episode began while the craft was adjusting the mirror on its thermal emission spectrometer. "We still do not know why that happened," he said.
Spirit can function without flash memory if necessary, Theisinger said. The most important effect would be that scientific data collected during a Martian day would have to be beamed back to Earth before the craft entered sleep mode, or else the data would be lost. That might reduce the amount of data collected somewhat, but would not make the mission a failure.
Spirit is halfway around Mars from Opportunity's planned landing site in Meridiani Planum. Geologists targeted that site because it boasts a feature seen nowhere else on Mars, or on Earth — an Oklahoma-sized deposit of an iron-containing mineral called hematite.
A shiny crystalline ore, hematite is most commonly produced in water, and scientists think that the massive deposit in Meridiani was laid down by an ancient lake. That suggests there once were large quantities of water on the surface of Mars and, where there was water, there very well could have been life.
"The surface of Mars is desolate, cold, dry and barren — not an enticing environment for life," said geologist Steve Squyres of Cornell University, principal investigator on the rover missions. But places like Meridiani offer "hints that it could have been a very different sort of place in the past."
The mission of NASA's twin rovers, as well as of Britain's failed Beagle 2 lander, is to look for evidence of "whether there was water and whether it was favorable for life," said project scientist Joy Crisp.
Meridiani Planum is within a large region that has been known as Meridiani since the earliest days of telescopic study of Mars because it lies near the planet's arbitrarily designated prime meridian, or line of zero longitude.
"Planum" means plain, and the name fits. Meridiani Planum is one of the smoothest, flattest places on Mars. That makes it one of the safest potential landing sites. But it is the hematite that makes it one of the most interesting.
The gray hematite was discovered by an instrument called a thermal emission spectrometer on the Mars Global Surveyor orbiter. The mineral covers an estimated 15% to 20% of the surface in the vicinity of the proposed landing site. In photographs, it appears as a dark layer atop a brighter layer of Martian soil that is exposed in many areas of the region.
"The hematite outcrop is right on the equator ... and it just kind of called to us to go to it," said geologist Ray Arvidson of Washington University, the deputy principal investigator.
On Earth, hematite is most often formed in association with water, either in iron-rich lakes or in hot water percolating underground. If Opportunity finds sedimentary rocks of the type found in lake beds associated with the hematite, that would be strong evidence that a lake once existed on the site.
Alternatively, the presence of certain characteristic minerals in conjunction with the hematite would indicate that it had been formed by hot water circulating in the soil. The instruments on Opportunity's flexible arm will be looking for the presence of those minerals.
A third possibility — and the least desirable one, from the scientist's viewpoint — is that the hematite was produced by the direct oxidation of volcanic rock without the presence of water. If Opportunity finds only volcanic rocks at Meridiani, that would support this possibility.
"Is this a big pile of volcanic sediments where hematite might have been altered, or is it a thick stratum produced in an ancient ocean?" Arvidson said.
The goal of the mission, he said, is to get "the real Mars to stand up." |
