In looking into reasons why the engine quit, they found faulty software that mistakenly told the lander it was on the ground when in fact it was still above the surface.
In a 154-page report out Tuesday, the Mars Polar Lander Failure Review Board pointed to poor training, inadequate testing, minimal oversight and a lack of people and money as primary reasons behind the failure of the $165 million spacecraft on December 3.
Had the potential engine glitch been detected in time, controllers might have been able to save the spacecraft with a single line of computer instructions beamed from Earth. That command would have made sure the engine didn't quit early.
"One line of code would have done it," said John Casani, former head of flight programs at the Jet Propulsion Laboratory (JPL), who led the failure review board.
Instead, the lander took a suicide plunge to the surface. The beginning of the end came when its legs were deployed in preparation for landing and sensors on the legs gave erroneous signals that the spacecraft was already on the surface.
In fact the Polar Lander was 132 feet (40 meters) above the ground when the engine cut off and it crashed into the surface at 50 miles per hour (80 kilometers per hour) -- five times faster than intended.
"It probably would have pancaked," Casani said. "The legs would have broken off, it might have cart-wheeled or started rolling. You had two propellant tanks that were half full and those would have broken up, though I don't think there would have been a fire or an explosion.
"In the end, you'd have a lot of pieces to pick up," he said.
NASA Associate Administrator for Space Science Ed Weiler said the agency is working to improve communications, training and oversight of its programs.
Complicated missions like the Polar Lander, he said, were under-funded by about a third and money will now be put into reserve at headquarters to make sure managers have enough cash to do the job.
"In hindsight all mistakes are stupid," Weiler said. "We pushed [JPL] too far and I will not condemn them for it We will continue to do missions faster and cheaper, but we're going to focus on doing them better."
Casani's team couldn't verify the engine shutdown scenario because the spacecraft wasn't carrying the necessary communications gear to tell controllers what was happening on its way down to the surface. Cash-strapped managers had failed to install the equipment, which would have cost a few extra million dollars.
"Given the total absence of telemetry data and no response [by the lander] to any of the attempted recovery actions, it was not expected that a probable cause, or causes, of failure could be determined," the report said.
But the board put engine shutdown at the top of seven possible failures. It pointed to "compelling evidence" from the spurious signals in the lander legs.
The signals issue first came to light in early February during tests on the 2001 lander at spacecraft-maker Lockheed Martin's facilities in Denver. Investigators learned of those tests around February 8.
In the case of the twin Deep Space 2 microprobes that piggybacked aboard the Polar Lander, the board found the mission went ahead despite being neither adequately tested nor ready for launch.
For those microprobes, "there was no one failure mode identified as most probable," the report said. The microprobes could have bounced on the surface, suffered an electronic or battery failure or landed on their sides, interfering with the operation of their antennas.
Casani's board listed 22 separate recommendations resulting from the Polar Lander investigation. Among them: more oversight of contractors, more systems engineering, more testing of flight components before launch and less overtime of project workers.
The report found the 10-member Polar Lander team at JPL was overworked and under extreme pressure to meet cost and schedule restraints. Workers routinely put in 60-hour weeks; a few worked 80-hour weeks.
"You had young people and we put them in a box there was no way to get out of," Casani said. "If anything, they were going too fast. The staffing was so thin that many functions went only one person deep. That's too light."
The JPL team should have had at least another 20 people. The Lockheed team, which numbered about 100 people, could have used 30 more, Casani said.
One section of the board's report listed 23 separate recommendations for future landers at Mars. Those included better communications gear to avoid the information-blackout that controllers had with the Polar Lander. The board also advised improvements in the propulsion and thermal systems, computer software and the structural frame of future landers.
Polar Lander was the second Mars probe to fail last year. The companion spacecraft, the $125 million Mars Climate Orbiter, either broke up or burned up in the martian atmosphere due to controllers' confusion over metric and English measurements in plotting its trajectory.
SPACE.com's Pasadena Bureau Chief Andrew Bridges contributed to this report
advertisement
