space.com's Andrew Chaikin. Executive Editor, Space & Science spoke recently with Donna Shirley, former manager for NASA's Mars Exploration Program at the Jet Propulsion Laboratory (JPL) in Pasadena, California. She offered her perspective on the failed mission. Shirley, now assistant dean of engineering at the University of Oklahoma in Norman, is also a member of space.com's board of advisors.

These successful missions stand in stark contrast to the dual failures of the Polar Lander and the Mars Climate Orbiter. Last fall, the orbiter failed to enter martian orbit because of a navigation error.

To understand NASA's most recent Mars debacle, Shirley said, you have to go back to another failure: the 1993 loss of the $1 billion Mars Observer probe, which fell silent just before it was due to go into martian orbit.

"The problem with Mars Observer," Shirley said, "was that NASA put all its eggs in one basket, and it failed."

Mars Observer's loss prompted NASA chief Dan Goldin to inaugurate his "faster, better, cheaper" approach to planetary exploration. The agency planned a series of smaller, less sophisticated orbiters to recover lost scientific data originally charged to the Observer.

NASA's budget for the new Mars Surveyor program: $150 million per year -- a bargain-basement price tag compared to Mars Observer.

To meet these goals, Shirley said, she and her colleagues at JPL worked with scientists across the U.S. and Europe to plan a decade's worth of Mars exploration. Each mission would take advantage of the biennial launch widow by launching two spacecraft, a strategy that also offered redundancy -- one probe could serve as a backup in case the other was lost.

The 1998 missions were to be a pair of identical orbiters, followed by two landers in 2001. And like the Mariner probes of the 1960s and early '70s, all the Mars Surveyor spacecraft would be built in an almost assembly-line fashion by a long-term "industrial partner," thus reducing their cost. But almost immediately, the plan changed, as Shirley explained to space.com:

Q: What happened to change your original plan for the Mars Surveyor program?

A: Originally the whole Mars Surveyor program talked about flying orbiters to recover Mars Observer science. The plan changed because NASA Headquarters decided that politically it would sell better if we had a lander as well as an orbiter on each opportunity. If we had flown two orbiters, they could've been carbon copies of one another, which would've saved a lot of money and reduced risk.

After Pathfinder and Mars Global Surveyor, which were relatively modest programs, there was rapid inflation in the number of experiments on a mission vs. budget dollars, to a point of total unreality. I thought the total unreality was going to happen in 2001. I was wrong. It happened in '98.

Also, a decision was made, completely separate from the Mars program, to have a smaller launch vehicle, and for the Mars program to be the first customer.

Q: How did that affect the Polar Lander?

A: When we went into this, we hadn't counted on a half-size launch vehicle. So you say, "I'm going to do [a Global-Surveyor-type mission] for half the cost. And you say, "Okay, I can figure out how to do that if I have less instruments, and I use new technology, and I have a kind of a production line going -- and so on." But then you add, "Well, it also has to be half the mass [because of the smaller launcher]." And, by the way, it's also going to have to carry more instruments. And you have to do an orbiter and a lander, which means you can't use complete commonality." If they'd been able to use two orbiters, it would've been a lot easier for Mars '98.

Q: In late 1996, the Mars program changed even more drastically, following the announcement by NASA scientists of possible evidence for fossil bacteria inside a martian meteorite.

A: That’s right. It was after that announcement that Dan Goldin urged a switch from a gradual understanding of Mars to a rush to look for life. And the key to that was the sample-return mission. Dan has always wanted a sample return because he believes it would attract public interest. And after the Mars meteorite in '96, sample return was the goal.

(Editor's Note: A sample-return had long been near the top of planetary scientists' "wish list," but it was seen as unaffordable within the strict Mars Surveyor program budget. And NASA planners looking ahead to human missions to Mars had wanted samples of martian soil to determine potential hazards and resources for astronauts. In 1996, at Goldin's urging, a sample return became the goal of NASA's robotic Mars exploration program, as a way to discover if martian life had ever existed.)

Q: What effect did the addition of the sample return mission have?

A: The whole science strategy that we had worked out very laboriously to fit within the $150 million per year, which included launch vehicles and everything, was out the window. We had this nice gradual gathering of knowledge about Mars, over 10 years. That all was thrown out and the program [was] focused on sample return.

Q: Some people have said that Mars Polar Lander (MPL) should have used airbags, like Pathfinder.

A: The MPL design was finalized before Pathfinder flew. The Lockheed Martin bid was for [a landing system using] rockets and legs. At the time [1995], everyone was going, "Ugh, airbags, what a stupid idea." It hadn't been proven. And in fact, the Pathfinder team didn't even know how to make it work yet.

Also, the airbags turned out to be much heavier than people originally thought. That's why they had to add a retrorocket for Pathfinder's descent, to keep it from hitting the surface too fast.

The reason Lockheed Martin chose legs and rockets for [the Polar Lander] was because [that system] had worked on Viking. [Lockheed Martin’s parent company, Martin Marietta] had two successes [i.e., Viking 1 and 2] with that technology, and none with airbags. What would you choose?

Q: So another factor is the fast pace of the Mars program?

A: Because you're launching every 26 months -- the 'faster' part of "faster, better, cheaper" -- you can't learn from one mission before you have to design the next one. You can only learn for one mission after that -- two opportunities later -- and even that’s tight, because it’s only four years after the mission you're trying to learn from [when it] arrives at Mars. So the 2001 lander, for instance, is now less than 16 months from launch. If you wanted to go in and make a major change in the 2001 lander, you couldn’t do it. I mean, you couldn’t go to airbags with 2001.

Q: As you’ve pointed out, we don't know why the Polar Lander failed?

A: That's right. The failure of the lander could very well have been that we didn't know enough about the environment at the [martian] south pole.

Q: What about the loss of the Mars Climate Orbiter?

A: One would never have expected a failure because of a metric-to-English units problem, but that was really a failure of not having enough staff.

Q: Is JPL spread too thin?

A: JPL has been spread too thin. NASA is spread too thin. The aerospace industry is spread too thin in general.

Q: Why is that?

A: The big reason is the end of the Cold War, which resulted in budget cutbacks and things like that -- the 'reinventing government' push to downsize government. One of the things that [resulted] was there were fewer and fewer people at NASA headquarters, so they couldn't really understand the issues [involved] in implementing the missions, and they were driven to make decisions for political reasons.

Q: How much of the problem [was created when] JPL underwent a reduction in force and lost a lot of good people?

A: I think that was a big problem. And JPL’s downsizing was totally political. It didn't save any money. It was because all the other centers were being downsized, so even though we're a contractor, we had to downsize. [In addition,] people are leaving because it's not a fun environment anymore… there’s so much pressure.

Q: Have you lost the institutional memory of people who really know what's involved in doing a mission?

A: I think we've lost a lot of it, yes. Bill Layman, for instance, retired. Bill is the best mechanical engineer in the world, I think, and he certainly is the best mechanical engineer at JPL. He was the chief engineer on the [Sojourner] rover and then we gave him to the [Pathfinder] lander, and he's absolutely brilliant and irreplaceable and he's gone. He retired. And Bill is less than 60 years old.

Q: So part of the problem is not just that the Mars program needs to slow down, but people need to be under less pressure and stay longer at JPL.

A: Or anywhere. It's not just JPL, it's everywhere. This is just a symptom of the entire business-driven, profit-driven, rush, rush, rush kind of thing that's stressing out the entire developed world.

Q: Are you encouraged at all by what you've heard since the failure, that things will change for the better?

A: I have heard that it's going to be studied. Whether it will change for the better, I reserve judgment.