Midcourse Correction

ith the flight of the first female space shuttle commander and the promise of captivating findings about life on Mars from two robot probes, 1999 should have been a banner year for the National Aeronautics and Space Administration. Instead, it will be remembered for its disappointments.

Air Force Col. Eileen Collins and her crew successfully delivered to orbit a $1.5 billion X-ray telescope in July 1999, despite a short circuit that shut down backup computers on two of shuttle Columbia's three main engines during launch. If redundant computers had been needed but failed on just one engine, Collins would have been forced to attempt an unprecedented and questionably survivable emergency maneuver-turning the $2 billion shuttle around in midair and flying it back to Florida's Kennedy Space Center. Only after Collins landed Columbia safely at the end of the five-day mission did inspectors discover the source of an unrelated low-fuel reading seen during liftoff: three dime-size holes in an engine nozzle. The punctures were made by a metal pin, not properly tested after it was used to fix a fuel injector plate, that jarred loose when the engine ignited. With Columbia back in its hangar, electricians hunting for the cause of the short circuit found a more serious problem. On all four shuttles, miles of wiring had been nicked and exposed for years by technicians careless about where they stepped or set up scaffolding during maintenance.

Two months later, a mix-up of English and metric navigation units by Lockheed Martin engineers sent the first Martian weather satellite, the Mars Climate Orbiter, on a suicide dive into the red planet's atmosphere. In December 1999, a machine was supposed to land on Mars, dig up soil, bake out any water that might be trapped in the dirt, and record natural sounds with a tiny microphone. Instead, the Mars Polar Lander crashed because a sensor not tested properly before launch signaled an early touchdown, causing the descent engines to stop prematurely. The combined price tag for the two failed Mars missions: $365 million.

The close calls with the shuttle and the back-to-back losses of the Mars probes weren't the only setbacks in a run of hard luck besetting the U.S. space industry in recent years. Suddenly, NASA seems to have lost its road map. Today, less than 36 months before the launch window for a new Mars mission, and in the midst of the most ambitious human mission ever attempted-building an international space station-the space agency is at a crossroads, wondering which way to go.

Faster, Cheaper, But Not Better

A string of investigations and a meteorlike shower of reports in March all bombarded NASA with the same conclusion: Administrator Daniel Goldin's eight years of efforts to explore space "faster, better and cheaper" haven't paid off in the way he expected. NASA is doing more things in less time and with less money, but at the expense of quality. "It is important to recognize that space missions are a 'one strike and you are out' activity," said the report of an 18-member independent team chaired by retired NASA and industry executive Thomas Young that assessed both Mars failures. "Thousands of functions can be correctly performed and one mistake can be mission-catastrophic."

Key projects are underfunded and employees are overworked, the Young team and other blue-ribbon panels concluded. Among the concerns cited: NASA headquarters administrators imposed arbitrary deadlines and budgets and demoralized employees by placing equal emphasis on safety and job cuts. "NASA went too far-from oversight to what I'll call 'out of sight,'" says Art Stephenson, director of NASA's Marshall Space Flight Center in Huntsville, Ala., and chief Mars Climate Orbiter investigator. Stephenson's group compared the results of eight different failure investigations and identified four recurring problems: inadequate reviews, poor risk management, insufficient testing and verification, and deficient communication. "The success of 'faster, better, cheaper' is tempered by the fact that some projects and programs have put too much emphasis on cost and schedule reduction," the 10-member panel found.

Additional evidence that NASA continues to operate on the edge of unacceptable risk came from the agency's inspector general, who audited three faster, better, cheaper experimental vehicle projects and found management similarly lacking. A September 1999 IG audit questioned the "cost reasonableness" of the X-33, the $1.2 billion NASA-subsidized prototype of Lockheed Martin's proposed VentureStar reusable launch vehicle. VentureStar has been touted as a space shuttle replacement. The X-33's advance composite fuel tank flunked a key qualification test in late 1999. Uncertainties about flight test requirements have grounded the $186 million X-34 reusable launch vehicle technology demonstrator. In light of all the recent mishaps, X-34 builder Orbital Sciences Corp. has gone back to the drawing board for an avionics redesign. Rapid prototyping of a $950 million, seven-passenger space station lifeboat, the X-38, is on target, but the $125 million craft, being developed and tested by a small band of NASA engineers, "warrants more risk management than current NASA policy requires," the NASA IG found in February.

Despite recent failures and the doubts they raise about the faster, better, cheaper philosophy, some still counsel NASA to stay its current course-but not at full speed. "We need to slow down some, not rush too quickly into important programs and projects, plan and implement them more carefully, and move away from fixations on costs and near-term gain," said Anthony Spear, the retired project manager for the successful 1997 Mars Pathfinder mission, in a March 1999 review of the faster, better, cheaper planetary missions.

"It needs some small midcourse corrections, but the basic direction Goldin has been trying to go is absolutely right," says Harvard government professor Steven Kelman, a federal procurement policy expert. "The alternative to faster, better, cheaper may end up being slower, more expensive and no better, and that would be a terrible shame."

Stepping on the Brakes

Since 1992, Goldin has guided NASA away from traditional missions-the kind that cost as much as $1 billion and take a decade or more to complete-toward smaller missions, typically costing from $150 million to $350 million and taking three or four years to complete. With the savings in time and cost, NASA can launch 10 or more robot explorers each year. Where astronauts are concerned, Goldin has directed the hand-over of routine space shuttle operations to a private contractor. His intent is to free thousands of civil servants to research and develop more affordable technology for space access.

In its early years, NASA was a lean machine with a mean mission: to beat the Soviet Union to the political high ground in space. NASA pushed the technology envelope in the 1960s and 1970s, suffering its share of failures but achieving monumental successes. In the 1980s, the last decade of the Cold War, space technology budgets grew. The cash infusion produced revolutionary advances in spacecraft and subsystem design. The Pentagon continued to test the boundaries, but NASA-preoccupied with the space shuttle-resisted advancing technology, streamlining management and adopting commercial-style business practices. With government workers overseeing contractor employees' every move, NASA developed an aversion to risk.

Then came three humiliating, costly failures. The shuttle Challenger exploded, killing seven astronauts little more than a minute after liftoff on an icy-cold morning in January 1986. A flawed rubber rocket seal got the primary blame, but the presidential commission that looked into the catastrophe identified schedule pressure and carelessness as underlying causes. In 1990, days after the $3 billion Hubble Space Telescope was placed in Earth orbit, astronomers discovered a misshapen mirror. It was a manufacturing defect that should have been found and corrected before the instrument left the ground. Then, in 1993, the $800 million Mars Observer was lost to a propellant tank rupture en route to the red planet-NASA's first planetary mission failure in 27 years. It became obvious that the fattened bureaucracy needed to diet.

"When I arrived at NASA money was the magic ingredient. It was a momentum model. From '83 to '92 the budget doubled and I had to take some very drastic action. I stepped on the brakes," Goldin said, defending himself before a House Science Committee hearing in June. He declined to be interviewed for this story. "Had I not stepped on the brakes, the NASA budget wouldn't be $14 billion a year, it would be $28 billion a year today," Goldin told the House panel. Taking a machete to budgets and thousands of jobs all at once made Goldin an enemy of the workforce and an ally of Presidents Bush and Clinton. Rep. Ralph M. Hall, D-Texas, calls Goldin "the best politician on the Hill," pointing out that he is a Bush appointee working in the Clinton administration.

Darling Clementine

History may hail Goldin as the creator of faster, better, cheaper, but the truth is that he did not originate it. Goldin took the idea and ran with it after the National Reconnaissance Organization's bargain-basement Clementine probe completed a successful lunar mission in 1994. Pedro Rustan, then director of small satellites for the NRO, used Clementine to demonstrate how to build small, efficient spacecraft by commercial methods still foreign to most government organizations. NASA participated in the project. "During this program, we experienced a great deal of frustration dealing with a NASA bureaucracy that, in our opinion, believed that at least five years and several hundred million dollars were required to build and fly a spacecraft that could achieve a credible scientific mission," Rustan told the House Science Committee in June.

Despite the impediments, Clementine was built in 22 months for $80 million. It mapped the surface of the moon and suggested the presence of ice in a large crater at the lunar South Pole. "Clementine's dramatic success encouraged NASA to reappraise large and expensive missions and return to the streamlined management practices of the 1960s and 1970s," Rustan told legislators.

Vowing to "turn the old way of doing business upside down," in 1994 NASA requested proposals for a series of focused science missions that could be developed and flown in three years for less than $150 million each, excluding the launch vehicle. Costs were capped at $250 million. Principal investigators were to lead the missions and be solely responsible to NASA for their success. Eight missions have been selected so far. Two-the Near Earth Asteroid Rendezvous, launched in February 1996 to orbit the asteroid Eros, and Mars Pathfinder, a small surface lander and rover that landed on Mars in July 1997-were in development by the time the Lunar Prospector was selected in 1995 as the third of the eight.

Alan Binder, who headed the Lunar Prospector mission, is director of the nonprofit Lunar Research Institute and CEO of the for-profit Lunar Exploration Inc., which seek to build a commercial moon base by 2012. The first peer-reviewed and competitively selected of the faster, better, cheaper Discovery missions, Lunar Prospector sought in part to verify Clementine's results. Launched in January 1998, it flew a trouble-free 19-month mapping mission and found no conclusive evidence of water.

A simple spacecraft with no redundant parts and no onboard computer, Lunar Prospector was developed privately, outside the NASA hierarchy. Binder and several colleagues conceived the idea several years before they proposed the mission to NASA, so the payload, spacecraft and mission already were well defined and validated. Binder and a 35-person Lockheed Martin engineering and support staff built the spacecraft and launched it in 22 months for a total of $65 million. "I decisively demonstrated with Lunar Prospector that inexpensive missions can be successfully defined, developed and flown with an absolute minimum of both management and NASA oversight," Binder said at the June House Science Committee hearing.

Like Lunar Prospector, some faster, better, cheaper missions have been wildly successful. Take the $185 million Mars Pathfinder with its anthropomorphic rover that scurried around the red planet looking at rocks named after cartoon characters. It was such a hit that Goldin challenged NASA to carry out two new Mars missions in 1998 for the price of that one. Pathfinder's orbiting companion, the $154 million Global Surveyor, recently beamed back the strongest evidence yet of the existence of water on Mars-in spite of a mechanical problem that delayed its mapping work for a year.

Losing Lewis and Clark

Other projects haven't had it quite so easy. Even Clementine had trouble. It was supposed to leave the moon and fly to an asteroid but couldn't because someone had inadvertently loaded unchecked software into the probe's computer while the exhausted ground control team was enjoying a weekend off. "During the slow time after the successful first part of the mission, I let people go out and celebrate," Rustan told the House Science Committee in June. "This shows that if you keep pushing and pushing the system, the system will fail."

The August 1997 failure of the $71 million Earth-observing spacecraft, Lewis, was another warning that faster, better, cheaper wasn't working. Developed under the fast-track Small Spacecraft Technology Initiative, a program that let satellite builders incorporate commercial design and qualification standards in a two-year development cycle, Lewis fell from orbit after launch. Investigators determined that TRW, the firm that designed and built Lewis, hadn't properly tested the attitude control system. Lewis had a companion, Clark, which was so far behind schedule and over budget that NASA terminated it in February 1998 losing $55 million.

Lewis investigators' assessment was prophetic. In a June 1998 news release upon publication of the February 1998 Lewis report, Christine Anderson, who chaired the failure board, raised the same concerns blue-ribbon panels expressed this spring. "I do not think that [faster, better, cheaper] is flawed," Anderson said. "What was flawed in the Lewis program, beyond some engineering assumptions, was the lack of clear understanding between NASA and TRW about how to apply this philosophy effectively. This includes developing an appropriate balance between the three elements of this philosophy; the need for well-defined, well-understood and consistent roles for government and industry partners; and regular communication between all parts of the team."

What's Better?

Faster. Better. Cheaper. Even the NASA Advisory Council, a group of outside experts who meet several times a year to help chart the agency's course, has argued with Goldin over the meaning of the words. The minutes of a 1997 session reveal that Goldin "asked the group to look at the approach retrospectively to understand the issues and prospectively to make recommendations.

"In response to a question about what he meant by faster, better, cheaper, he said he meant 'faster, better and cheaper,' " according to the council minutes. "It is not theoretical . . . he said his words have been manipulated to sell programs. He wanted it on the record that he was furious."

Of the three words, "better" is most elusive. "We've never been able to define what better is in any meaningful way," says Donna Shirley, manager of the Mars exploration program at the Jet Propulsion Laboratory from 1994 until she retired in 1998. "What is better? More science with simultaneous observations? Incredible resolution with no coverage? You need both. As the joke goes, you can't have faster, better and cheaper. Pick two."

When is a mission too fast and too cheap? The Aerospace Corp., a private, nonprofit California firm that operates a federally funded research and development center for the Defense Department, analyzed 48 NASA and Defense Department low-cost planetary and Earth-orbiting satellites. The corporation used a "complexity index" devised from 21 identifiable cost, schedule and technical parameters common to spacecraft projects. NASA's impaired or failed missions are significantly more complex than the average mission, according to the statistical analysis. In particular, the small planetary missions cost more, are developed faster, and fail more often than do the Earth-orbiting missions. Whether more money or extra time would have improved the odds for any given mission wasn't known, the May study said, "but this much is clear: When NASA fails, it appears that they are beyond a discernable threshold where resources are insufficient."

In a separate comparison in 1999 of low-cost and traditional robotic space projects, The Aerospace Corp. tried to answer the question "Is faster, cheaper really better?" It found that 44 percent of faster, better, cheaper missions suffer impairments or catastrophic failures, whereas only 30 percent of traditional ones encounter similar fates. It quantified science return, or return on investment, from faster, better, cheaper missions at 79 instrument-months, on average-significantly less than the 305 instrument-months returned from traditional missions. "Given the reduced science return and the high failure rate," the study found, "faster, better, cheaper missions cannot be considered 'better.'" But the study also found that faster, better, cheaper missions are 57 percent more cost-effective because of their economy in time-they are twice as fast-and price-they cost 85 percent less.

The failed Mars missions "died a death of a thousand cuts," Shirley says, because the cost targets set by NASA headquarters held project managers to fixed budgets despite the fact that the requirements set by headquarters officials grew steadily. Shirley, now assistant dean of engineering at the University of Oklahoma, was among those who thought the Mars Climate Orbiter and Polar Lander missions would succeed. She worried, however, about the series of missions planned to take place every 26 months in the next 10 years. In written testimony to Congress for the June House science hearing, Shirley said she and other Jet Propulsion Laboratory managers complained to Goldin more than once that NASA officials were imposing unrealistic requirements on the Mars program without providing additional resources to mitigate the risk. Goldin testified that he didn't remember a conversation with Shirley but acknowledged that "on many occasions people came into my office and I threw 'em out on their heads when they wanted more money."

Tight-Fisted Turnaround

Despite the Young team's finding that the failed Mars missions were underfunded by 30 percent, it doesn't appear that NASA will be given significantly more money than Goldin already has requested. His $14 billion budget request for fiscal 2001 represents a 3.2 percent increase over this year's appropriation of $13.6 billion. It reflects future year increases above the rate of inflation-to $15.6 billion in 2005. Science and technology expenditures account for 42 percent of the budget this year-up from 31 percent in 1991 on the way to an expected 51 percent in 2005. "I'm steadfast in my position that money is not the magic ingredient," Goldin said in his testimony at the June House science hearing.

If money isn't the magic needed to get NASA back on course, it's legitimate to ask what is. The answer, at least for now, appears to be a mix of increased inspection, improved technology, restructured missions and beefed-up training.

At the behest of a 13-member assessment team headed by NASA Ames Research Center Director Henry McDonald, the shuttle program has undertaken a rigorous review of 76 pieces of equipment designated "Criticality 1." The failure of any one of these items would lead to a catastrophic loss of the orbiter and its crew. The technology to make them fail-safe did not exist in 1972 when the shuttle's design was frozen. The intent of the review is to determine whether the technology exists today to improve the equipment. The shuttle program also devised new access methods to prevent damage by orbiter repair maintenance technicians and added about 150 people to its Kennedy Space Center launch site staff. NASA space flight chief Joseph Rothenburg acknowledges that the shuttle program needed a shakeup. "After 14 years of successful missions, you tend to get a little complacent. Having an independent team come in and ask you all these questions raises your sensitivity."

In response to the Young team's assessment, NASA canceled plans to send another lander to Mars in 2001-though plans to send an orbiter still stand-and moved Scott Hubbard, a top Jet Propulsion Lab manager, to Washington to oversee Mars projects from cradle to grave. The agency vowed to conduct a Mars mission at every opportunity but retreated from the idea of requiring two probes each time while it reevaluated its exploration goals. Rather than arbitrarily dictating what should fly and when it should fly, Hubbard says, top-level management has adopted a more sensible strategy: Fundamental science-with the goal of finding ancient or modern water on Mars-will determine the mix, timing and scope of NASA's future missions to the planet. "This new approach gives the taxpayers a much better return on their investment," says Hubbard. Where there is water there should be life. So uncovering evidence of living organisms, if they ever existed, will be a coincidental but intended bonus.

But inspecting and restructuring are cosmetic adjustments. Ed Hoffman, head of NASA's Academy of Program and Project Leadership, says correcting the agency's course should include properly equipping employees to do their jobs.

In the past, trainers planted the seeds of basic skills in a classroom setting and left the nurturing of new hires to the more experienced employees in the workplace. "Over the last few years, that paradigm has changed dramatically, and it has caught us by surprise," Hoffman says. "I saw it happening, but I thought it would be a little bit more evolutionary." As NASA drastically reduced its workforce through buyouts, retirements and reductions in force, it dramatically increased its num- ber of projects. "People are mo-ving up faster, taking over greater responsibilities than ever before. All of a sudden, the senior folks aren't there," adds Hoffman. The new paradigm requires a significant change in the agency's employee development methods. There will be less emphasis on the classroom, increased use of the Internet to deliver just-in-time training, and more emphasis on challenging simulations. "A lot of the technology is going to have to veer off of training just for the sake of the lesson," Hoffman says. "We're getting more requests for tools and techniques that can really be used."

No one is recommending that NASA forget faster, better, cheaper and find a newer way of doing business. "It's the key to the agency's future," says McDonald. "It's a matter of fine-tuning the system more than anything else." There is no easy solution for Hubbard, whose Mars projects are the most out of tune. "It is something you live with every day," he says. "When you get up and look in the mirror and brush your teeth in the morning, you ask yourself-at least, I do-'Are we hitting the right balance?' It's not something you can write a prescription for, go away for three years, come back and find that everything is perfect. Ongoing assessment is part of the process."

Beth Dickey is a freelance writer who has covered the U.S. space shuttle program since 1981. She reports from Kennedy Space Center.