Current and former NASA officials gave much of the credit for the successful landing of the Spirit Mars rover Jan. 5 to the lessons drawn from a recent pair of failed attempts to explore the red planet up close.

Among those lessons is that the budgets for such complex undertakings must be flexible enough to accommodate the inevitable problems that crop up during development. But while setting arbitrary cost caps on missions can be a recipe for failure, bigger budgets do not guarantee success, they said.

“We are always on the razor’s edge of success or failure,” NASA Administrator Sean O’Keefe told reporters Jan. 6. “We never know which way it’s going to go and the best we can do is the best we can do.”

When it comes to Mars exploration, failure has been more the rule than the exception since the dawn of the space age. Spirit, whose twin Mars Exploration Rover, Opportunity, is slated to touch down on the martian surface Jan. 24, is just the fourth probe to successfully land on the red planet in 15 tries.

For NASA, the most recent Mars lander mission prior to Spirit was the Mars Polar Lander, which was lost in September 1999 as it began its descent to the distant planet’s surface. Just two months before that, NASA’s Mars Climate Orbiter disappeared without a trace as it was set to enter Mars orbit.

The failures of the Mars ’98 missions, so-called because they were launched in 1998, triggered a comprehensive independent review led by former Martin Marietta chief A. Thomas Young. The resulting “Mars Program Independent Assessment Team Report,” released in 2000, blamed the failures on an inexperienced contractor and civil-service workforce, inadequate testing and oversight, and poor communications.

The Mars ’98 missions, developed according to the so-called faster, better, cheaper philosophy championed by then-NASA Administrator Daniel Goldin, had a combined budget of $328 million. Donna Shirley, who managed the Mars ’98 program but left NASA prior to the loss of the two spacecraft, said the effort was undermined by inadequate funding. The Mars ’98 spaceraft development budget was about half that of the successful Mars Pathfinder lander and Mars Global Surveyor orbiter missions, she said. “We ended up with about twice as many requirements, she said. “I tried to tell people over and over again that that wasn’t going to work.”

After the Mars ’98 failures, NASA canceled plans to launch a lander in 2001, but continued work on an orbiter, dubbed Mars Odyssey, which launched in 2001 and currently is orbiting the red planet.

“After the failures, they decided to be a lot more sensible and reasonable. So they canceled the 2001 lander and put all the money into the orbiter, and guess what? It worked,” Shirley said. “It’s very straightforward. If you either control your appetite for requirements, things will work. If you don’t, they won’t.”

The term faster, better, cheaper reflects three key variables that space program managers typically have to work with: schedule, performance and cost. On Mars missions, for which there are narrow launch windows that arise once every two years, schedule flexibility often is limited, leaving performance and cost as the variables that best lend themselves to tinkering.

“We had three variables to work with: schedule, performance and cost,” said David Lavery, Mars Exploration Rovers program director at NASA headquarters. “We made a conscious decision up front that because schedule and performance were unmovable and couldn’t change, if we had to relieve pressure during development, the only variable we could do anything about was cost.”

Having that cost flexibility was key, because NASA did in fact run into developmental surprises on the Mars Exploration Rovers program. For example, NASA originally planned to use an parachute-and-airbag landing system identical to the one used on the $250 million Mars Pathfinder mission, only to find out that that substantial changes would be required to accommodate the much-heavier Mars Exploration Rovers, Lavery said. In addition, and in accordance with a recommendation from the Young report, NASA modified the rovers’ planned approach and landing trajectories several times to ensure that it could maintain contact with the craft during critical events, Lavery said.

Because NASA had no communication with Polar Lander during its descent, the team had very little telemetry to work with to recreate the failure.

As a result of these and other changes, the price tag for the Mars Exploration Rovers rose by about 15 percent, or $100 million, to about $820 million.

But O’Keefe said it would “oversimplify matters” to attribute the success of the Spirit landing to the fact that the Mars Exploration Rovers program budget was higher than that of the Mars ’98 missions. “It was 820 million bucks for two rovers,” O’Keefe said. “Whether or not that’s too much or too little is something we’ve got to continually look at.”

Lavery was similarly cautious about comparing budgets, saying the differences between the two programs are much broader. The science payloads for the Mars Exploration Rovers, for example, were built and waiting on a shelf when development of the vehicles got under way, he said.

Louis Friedman, a former NASA engineer and now executive director of the Planetary Society, a Pasadena, Calif., space exploration advocacy group, said it is far from certain that money in and of itself makes the critical difference between success or failure in Mars missions.

“That’s a very hard question to answer,” Friedman said. “Would [the Mars Exploration Rovers] have failed at $700 million? At $500 million? I don’t know. What this mission has shown, in comparison to Mars ’98 … is that we must never underestimate the risk.”

“They put extraordinary effort into it and it paid off,” Friedman said. “That’s more important than striking the right dollar balance. “Extraordinary efforts are needed to land on Mars. That’s the lesson I would draw from it.”

Lavery said a formal exercise to compile lessons learned from the Mars Exploration Rover program will get under way later this year. For now, he said, the team is focused on ensuring that Opportunity’s upcoming landing goes as well as Spirit’s.

“Nobody on the project is being lulled into a false sense of security about what we’ve got going on,” Lavery said Jan. 7. “We’ve got another one of these [landing] in another two-and-a-half weeks and it’s far from being a slam dunk that that’s going to go as well as this one.”