NASA orders sweeping changes after Mars failures
BY WILLIAM HARWOOD
Posted: March 29, 2000
In the wake of two costly - and embarrassing - failures last year, NASA is restructuring its Mars exploration program, indefinitely delaying a planned 2001 landing mission amid sweeping management changes to improve communications, oversight and engineering expertise.
NASA science chief Ed Weiler also said the agency will re-examine its science goals on Mars and the technological approaches needed to achieve them. While he did not directly address whether plans to launch a Mars sample retrieval mission in 2005 will remain on track, the long-awaited flight faces major hurdles that would appear, at this point anyway, to make a delay inevitable.
In the near term, Weiler said NASA will focus on improving communications between NASA headquarters, the Jet Propulsion Laboratory and various contractors; implementing new programs to improve training; and setting up a Mars program director at NASA headquarters to beef up project oversight.
In addition, money will be held in reserve to handle unexpected problems during project development, a luxury that was not available to the cost-capped Mars '98 team responsible for the ill-fated Mars Climate Orbiter and the Mars Polar Lander.
The Climate Orbiter was lost last September when it plunged too steeply into the martian atmosphere while attempting to brake into orbit. The mishap later was blamed on the cumulative effects of small rocket firings that were based on English thrust units, provided by prime contractor Lockheed Martin, instead of metric units as required by NASA's Jet Propulsion Laboratory.
The Polar Lander was lost in December. Engineers may never know the exact cause because the spacecraft terminated communications with Earth - as planned - just before atmospheric entry. With no data, engineers can only make educated guesses about the most likely failure scenarios.
But an independent review team, in a report released today, said the Mars Polar Lander most likely crashed because of a false signal in the landing gear system that fooled the spacecraft's poorly programmed flight computer into thinking it had touched down on the red planet's surface when, in fact, it was still high off the ground.
Under this scenario, the shock of the probe's three landing legs snapping open and locking in place caused at least one of three landing sensors to generate a spurious signal. The flight computer, which was not programmed to differentiate between spurious and real landing sensor data, then shut down the Polar Lander's descent engines while the spacecraft was still some 130 feet up.
As a result, the Polar Lander crashed to the surface at a catastrophic 50 mph or so, according to the Mars Independent Assessment Team report released today.
He said the engines would have actually shut down "when the lander was about 40 meters, or 130 fee,t above the Mars surface and from that condition of falling without any way to decelerate, the lander would have hit the Mars surface at 22 meters per second or about 50 mph, therefore it would have been destroyed."
While the most probable cause of the failure was premature engine shutdown, "the underlying cause is inadequate software design and systems tests," Young said. "The software should have been designed to prevent spurious signals from causing premature engine shutdown. The systems test should have identified the presence of spurious signals."
He said Lockheed Martin carried out a full-scale test of the landing gear system before launch. But the touchdown sensors were improperly wired because of a design problem.
"As a result," Young said, "when the deployment actually took place during the test, no signals were generated by the sensors. After the wiring was corrected, the (test) was not repeated and therefore the systems test did not demonstrate this problem."
The landing sensor issue came to light in January, during tests of NASA's 2001 lander, which is similar to the Mars Polar Lander. In four of five tests, the landing sensors tripped when the legs deployed.
"In all four the lander would have failed had that been the circumstance that existed at Mars," Young said. "So I think it's almost certain that had the lander gotten to this particular part of its mission successfully then this would be the cause of the lander failure."
Young's panel concluded the Mars '98 program was under funded, its engineers ill trained and its otherwise competent managers inexperienced when it came to operating such complex programs. Prime contractor Lockheed Martin was blamed for promising more than it could deliver, failing to properly check out and test the flight hardware and failing to properly communicate with counterparts at NASA.
The company, in effect, promised NASA more than it could deliver in what Young described as an "aggressive" contract proposal. As for communications, "Lockheed Martin, in our view, had failed to provide a formal identification to the customer, JPL, that indeed there were deviations from standard practice, there were risks that were becoming larger than they should on a program like this," Young said.
While Young's report provided an embarrassing look at a high-profile program gone awry, Weiler said the public should keep NASA's successes and failures in perspective.
"If you look at the total amount of launches we've had in terms of dollars, that is, what's the total value, dollar wise, that we've launched since, say, 1992, it's about $18 billion worth of space hardware," he said. "Out of that $18 billion - counting these two Mars failures - we have lost $500 million. That's 97 percent (success)."
In the case of the Mars '98 failures, he said, 20-20 hindsight shows NASA asked the Jet Propulsion Laboratory to "do the impossible."
"We pushed them too far and I will not condemn them because they failed," he said. "In fact, I salute them for trying to succeed against these substantial odds. ... These two failures are not a damnation of 'faster, better, cheaper.' Remember: 97 percent."
About the author
William Harwood has covered the U.S. space program for more than a decade. He is a consultant for CBS News and writes for The Washington Post and Space News. He maintains a space website for CBS News.
MPL potential failure modes - possible causes of the Mars Polar Lander failure from the official report.
The planned touchdown of the Mars Polar Lander probe is shown in this computer simulation.
PLAY (407k QuickTime file)
Post-touchdown activities planned for MPL are depicted in this computer simulation.
PLAY (166k QuickTime file)
Download QuickTime software to view these files.
Explore the net
Mars Polar Lander - official website at NASA's Jet Propulsion Laboratory.
Deep Space 2 - official home page for the twin Deep Space 2 probes.
Mars Microphone - the Planetary Society's project to listen to the sounds of Mars.
Mars Volatiles and Climate Surveyor - a suite of instruments aboard the Mars Polar Lander.
Mars Descent Imager - details from Malin Space Science Systems.
Search Mars Polar Lander websites. (Provided by SpaceRef)
Sign up for Astronomy Now's NewsAlert service and have the latest news in astronomy and space e-mailed directly to your desktop (free of charge).