Are Failed Mars Probes Price of Cost-Cutting?

Pieces of the future Mars exploration program--circuit boards, solar arrays, graphite frames and propulsion nozzles--were laid out in the high bay of Lockheed Martin Astronautics like an incomplete puzzle, open to revision and recrimination.

Here in this cavernous spacecraft assembly room, five Lockheed technicians are building a Mars lander virtually identical to the $165-million probe that vanished on its approach to the Red Planet earlier this month--all but certain the spacecraft on which they labor so diligently now will never be launched.

A few feet away, eight other engineers work on another Mars spacecraft, an orbiter similar to the one that disappeared near Mars in September.

Sobered by its back-to-back failures, NASA is determined to rethink its Mars exploration program and, if necessary, revamp these spacecraft designs. The agency plans to scrutinize project management at the Jet Propulsion Laboratory in Pasadena, which oversees the $356.8-million Mars exploration program, and its major contractor, Lockheed Martin.

It is the fifth time in barely a year that spaceflight failures have put Lockheed--the world’s largest aerospace contractor--at the center of a formal investigation. Each time, outside experts have sharply criticized the aerospace giant for mismanagement and an overemphasis on cost-cutting.

Lockheed has already acknowledged responsibility for an arithmetic error--failing to convert crucial navigation measurements from English to metric units--that sent the Mars Climate Orbiter plunging off course to its doom.

Now, in a frank assessment of their part in the Mars mishaps, Lockheed executives say that in their efforts to meet cost and schedule constraints, they did not test the Mars Polar Lander as thoroughly as they might have liked--especially the complex cluster of retrorockets used for its powered descent to the Martian surface.

They also say the need to economize drove major design decisions, including the choice of a novel “pulsed” landing system and the decision to forgo a telemetry transmitter that could have allowed flight operations engineers to maintain contact with the lander as it rocketed down to the surface. Still, Lockheed engineers are confident that the Mars lander did not fall victim to a mechanical or electronic malfunction. While proud of the probe they built, some Lockheed managers worry now that perhaps they “skinnied down” too much during its design and construction.

“We did what we thought was good enough,” said Edward A. Euler, director of Lockheed Martin’s Mars program.

Under less strict financial and schedule constraints, “we would have put it to more testing,” Euler said. “We did this job with a hell of a lot less people 1/8than in previous projects 3/8. I am worried that we may have gotten fooled.”

Earlier problems on the Mars project cost the company much of its $12.5-million award fee. The subsequent loss of the Mars Polar Lander and its sister spacecraft, the Mars Climate Orbiter, almost certainly will cost Lockheed any remaining profit to be had from its contract, company executives said.

“If we were responsible, there is no award fee,” said Noel Hinners, Lockheed’s vice president for flight systems.

Staffing, Testing to Be Beefed Up

In hindsight, several Lockheed managers said, there may have been too few engineers assigned to the Mars project and not enough oversight of those who were working on it.

Hinners said Lockheed may have pushed its people too hard. Eighty-hour workweeks were common, with a few people logging 100-hour workweeks. “There are hints that . . . we are working people too hard, stressing them with too much overtime,” he said. Under-staffing on the Mars exploration programs was a “very difficult” problem, he said.

Lockheed engineers said their working relationship with JPL was agreeably “collaborative, not confrontational.” But there certainly was not enough regular communication between spacecraft teams in Denver and Pasadena, they said.

C.W. McNally III, vice president and general manager of Lockheed Martin Astronautics, said overconfidence on the company’s part might have contributed to the Mars problems.

“This is a high-risk business. We may have deluded ourselves,” he said.

Lockheed managers are moving to beef up staffing, testing and quality assurance efforts, company officials said.

In the meantime, flight operations engineers are still trying to contact the Polar Lander. The orbiting Mars Global Surveyor has been photographing the suspected landing zone with its high-resolution camera, seeking vainly for some trace of the craft’s parachute.

But with no sign of the spacecraft or its two microprobes after weeks of searching, no one may ever know what happened.

NASA is assembling a panel for an inquest into the most recent Mars mishaps. JPL last week appointed its own 12-person review board, which is expected to report its findings in March.

“In catastrophe there is also opportunity, and that is the case here,” said Caltech planetary scientist Bruce Murray, a former JPL director who is serving as a consultant to the NASA panel.

“It is really the opportunity to not only deal with what shortcomings may have developed but to structure a stronger program for the future after the severe disappointments,” Murray said.

In the last decade, NASA has mounted only one wholly successful Mars mission in five attempts--the 1997 landing of the Pathfinder probe and its robot rover Sojourner. That craft was designed and built at JPL, at twice the cost of the more recent Mars probes.

Three other Mars missions failed completely, including the Polar Lander and its auxiliary Deep Space 2 probes. Engineers had to nurse a fourth--the Mars Global Surveyor--through a major malfunction to reach its proper orbit around Mars in 1997. A jammed antenna still affects its ability to transmit high-speed data to Earth.

All four spacecraft were built by Lockheed under JPL’s supervision.

The Mars failures followed five major launch mishaps in the past 18 months caused by faulty rocket boosters built by Lockheed or by its major U.S. competitor, Boeing Corp. More than $3.5-billion worth of critical military and commercial satellites were destroyed.

Surprisingly, the most serious Lockheed mishaps involved what had been the company’s most reliable heavy rocket booster--the Titan IV, which until August 1998 had suffered only one loss in 24 launches.

Then it hit three in a row.

First, a $1.3-billion spy satellite was destroyed when its Titan rocket booster short-circuited and exploded shortly after launch. Air Force investigators subsequently discovered damaged wiring in several other Titan IVB and Titan II rockets being readied for launch. Then a $250-million satellite was destroyed in April when someone left sticky tape on a wiring connector in the Titan’s upper stage. A few weeks later, an $800-million military satellite was wrecked when someone loaded a corrupted computer code into its Titan booster.

That same month, a Lockheed-built Athena rocket carrying an advanced Earth-imaging satellite crashed into the Pacific Ocean when a protective shroud did not separate properly and dragged the rocket back down through the atmosphere, where it burned up.

In the aftermath of those mishaps, two independent panels of experts concluded that poor quality control, an overemphasis on cost-cutting and a loss of experienced personnel were largely at fault. Those conclusions then were echoed by the NASA panel investigating the Climate Orbiter loss. The NASA inspector general also faulted management and staffing at Lockheed and JPL.

To head the team responsible for assessing NASA’s robotic exploration of Mars, NASA Administrator Daniel Goldin appointed A. Thomas Young. A retired Lockheed Martin Corp. executive, Young earlier this year was in charge of an independent panel investigating the company’s rocket booster problems.

Even so, independent space analysts view the most recent Mars failures as more than the problem of a single aerospace contractor. Rather, experts believe they are symptoms of broader ills that affect the industry and the government agencies that manage spaceflight.

“The military expendable launch vehicle 1/8programs 3/8 were understaffed, and the latest round of Mars missions were understaffed,” said John Pike, a space policy analyst at the Federation of American Scientists. “The common thread is that they all were trying to do it on the cheap.

“This is starting to look like a pervasive, industrywide problem,” Pike said.

Each Mission Is an Experiment

The concerns are not limited to the unmanned space program. NASA’s Aerospace Safety Advisory Panel, which annually reviews the manned spaceflight effort, is poised to issue a warning next month about the potential dangers of cost-cutting and personnel reductions in the Space Shuttle program.

By an independent accounting, NASA has designed and launched 16 robotic exploration missions since 1992 under its management philosophy of “faster, better, cheaper” space exploration. Seven of those missions either failed completely or suffered serious technical problems after launch.

The combined cost of the 16 missions is still less than that of the $3.3-billion Cassini probe, the last of NASA’s previous generation of large robotic space probes. The Cassini probe, managed by JPL, is now entering the asteroid belt after traveling 1.25 billion miles on a Lockheed-built propulsion system toward a rendezvous with Saturn in 2004.

Nonetheless, “that is a failure rate of 37%,” said Duke University technology historian Alex Roland. “That suggests something is systemically wrong.”

In the drive to reduce the price of routine spaceflight, congressional and agency budgeteers have been too aggressive, creating a series of low-cost failures, experts said.

“The basic philosophy is fine--to move to smaller, cheaper, faster missions,” Roland said. “ 1/8But 3/8 they have gone too far. They are trying to run the programs too cheaply. And every time you scrub a test--make one of those engineering judgments to cut something--you are asking for trouble.”

No matter how hard NASA engineers and their contractors try to reduce costs by using off-the-shelf components or recycling previously tested designs, their planetary explorers are still engineering experiments. Each craft embodies a unique mission that must unfold perfectly in the unforgiving extremes of space.

Lockheed Mars Surveyor operations manager James Neuman said that as a contractor working to a customer’s specifications, he faces a fundamental engineering dilemma when building an interplanetary spacecraft on a tight budget. He must leave some things out, and every choice entails some risk.

“What do you want us not to do, with the resources we have available?” Neuman said. “Make the money fit the project.”

Space policy analyst John Logsdon at George Washington University said NASA is already worried it may have ceded too much control to its contractors in its efforts to streamline mission management.

“Among other things,” Logsdon said, “these failures are wake-up calls to say the balance is not quite right yet.”

But so fast is the pace of planetary exploration today that NASA must speedily re-engineer those craft being readied for launch to the Red Planet in 2001, even before it can be certain what happened to the missing Mars Polar Lander.

Already, NASA planners are studying whether to incorporate the very features that earlier budget constraints eliminated from the missing lander, such as more robust radio capabilities during descent and a radar system that could detect landing hazards.

A better communications system might add $2 million to the new lander’s cost, said Euler at Lockheed. The hazard avoidance system might cost as much as $10 million.

But it may be that time now is the problem, not money.

‘The amount of technical changes we can do now is limited,” Euler said, “if we are going to launch in 2001.”