The Space Reviewin association with SpaceNews

Obama at KSC
President Obama hopes to turn more responsibility over to the private sector for human spaceflight, but past experience in the defense field shows problems with that approach. (credit: B. Ingalls/NASA)

Looking for a silver bullet

Tempering optimism with a dose of history

Bookmark and Share

The new space plan is here.

Gone is Constellation, the US plan to return to the Moon by 2020. Gone is the Ares 1/5 rocket architecture, Orion spacecraft, and Altair lunar lander that were to be developed to support this goal (well, Orion is only sort of dead, but more on that later). So it’s out with the Moon and out with the Constellation program that had the blogosphere so riled up for years. And it’s in with…

That’s the fuzzy part.

When variants of this idea of circumventing the government acquisition system have been tried in the past, not only did they not work, but they turned out to be unmitigated disasters.

President Obama’s space speech on April 15—almost certainly the only space speech he will make during his administration—was full of inspiring rhetoric but quite devoid of detail. The key thrust of the plan, though, involves scrapping the Ares 1/Orion replacement for the Shuttle and substituting a commercially developed vehicle. The premise is that NASA will buy rides on commercially offered spacecraft, and move toward a paradigm closer to buying tickets on an airline to space, rather than acting as an entity that builds an airliner every time it wants to fly.

This sounds like a great idea. It seems logical that private industry, free of the bureaucracy and inefficiency of government, could develop spacecraft more cheaply, faster, and for less risk than an “in-house” NASA spacecraft and then offer use of these spacecraft to the government at lower cost than present human launch services. This would be phenomenal for the space industry.

But it’s just not that simple. This is not a new idea. The Obama Administration is not the first to suggest the idea of circumventing the government acquisition system by contracting directly with private industry. And the problem is, when variants of this idea have been tried in the past, not only did they not work, but they turned out to be unmitigated disasters.

Killing the $800 screwdriver: the silver bullet

Of all the functions of government, defense procurement arguably has one of the most unsavory and sullied reputations of all. Most Americans hold the view that the defense procurement process consistently produces systems that are over budget, behind schedule, and under performance targets. To the extent that this is true, it is a complex and multifaceted problem, rooted in a variety of structural issues inherent in defense procurement: congressional micromanagement, “requirements creep”, poor management, companies forced to underbid lest they lose their technical teams, disagreements over contract interpretations, overregulation, and a host of other problems. Horror stories of $400 hammers and so forth, while they are often nothing more than accounting quirks, contribute to the perception that the government simply cannot produce weapon systems on time and on budget.

As is the case in many industries, rather than taking carefully considered steps to address these long-term problems, defense procurement reform has instead traditionally taken a silver bullet approach: generate some magic new fix that will instantly eliminate all these problems, all at once. Needless to say, this approach has never worked.

The Obama administration has essentially pitched its proposal thus: NASA will say to private companies, “go and produce a working spacecraft, and then we’ll use it”, and then, in effect, sit back and watch. If the company delivers, great; if not, NASA hasn’t lost much. This arrangement sounds great if it could successfully produce reliable, quality products time and time again. But in fact, this approach is frighteningly reminiscent of a process that today causes many in today’s defense industry to shudder: Total System Performance Responsibility (TSPR).

TSPR did not work. In fact, not only did it not work, but as former Martin Marietta head A. Thomas Young recently stated, “Projects were a disaster and TSPR was judged by all to be a total failure.”

TSPR was an idea hatched in the days of the Clinton Administration as part of an attempt to drastically reduce the cost of defense acquisition in the aftermath of the Cold War. The underlying premise behind TSPR was that too much money was being spent on government employees who specialized in defense acquisition and who oversaw defense contracts let to private companies. Instead of relying on government to oversee the development of weapons systems, TSPR had the contractor itself totally responsible for the performance of the system it delivered—hence Total System Performance Responsibility. This is quite reminiscent of what the Obama Administration is proposing; rather than have NASA oversee spacecraft development at each contractor, NASA will simply specify its requirements and then let the private contractors take responsibility for delivering the spacecraft, supposedly precipitating huge economies in cost and time.

Except, TSPR did not work. In fact, not only did it not work, but as former Martin Marietta head A. Thomas Young recently stated before the House Science and Technology Committee, “Projects were a disaster and TSPR was judged by all to be a total failure.” In essence, what happened is that the USAF and NRO eliminated many of their key project managers and systems engineers, thereby losing decades of institutional knowledge and memory for systems engineering and management of large-scale space projects. Space projects that were “beneficiaries” of the TSPR approach include the Future Imagery Architecture (which the New York Times called “the most spectacular and expensive failure in the 50-year history of American spy satellite projects”), the NPOESS weather satellite, (whose cost had increased by about $7 billion—more than 100%—since its inception in 1994, and which was cancelled in February) and SBIRS High (which a 2003 DoD report said “could be considered a case study for how not to execute a space program”—and it was only about $1 billion and 33% over budget when that report was written; today it is more than $7.5 billion over budget and nearly a decade behind schedule).

NASA, of course, does not develop its spacecraft in house. As the president himself pointed out during his KSC speech, since the dawn of the Space Age NASA has always had private contractors develop its manned spacecraft, from McDonnell’s Mercury to North American’s Apollo to Lockheed Martin’s Orion. While NASA has had a key role in setting the design parameters for these spacecraft, at the end of the day it is the private companies that executed the detailed design and construction. It is thus unclear what advantage would accrue in letting private companies “develop spacecraft for LEO access” unless NASA were to withdraw the stifling bureaucratic and governmental overhead that causes government acquisition to be so expensive in the first place. In other words, the new Obama paradigm will have private companies, using government funding but without the overhead of government management, delivering technically complex space systems for a single-customer market.

Sound familiar?

Thus, unfortunately, the new Obama space plan smacks of TSPR. The approach that has been laid out by the President, while vague, seems to have the same premise: take government out of the equation and let the contractors oversee development and performance of the product, supposedly leading to cheaper systems that are delivered faster.

This is the primary difference between the Obama plan and the Constellation acquisition strategy for Orion, and its key flaw, as discussed by Young in his Congressional testimony and also retired USAF Lt. Col. Wayne Eleazer (see “All things old are new again”, The Space Review, October 20, 2008) is the loss of institutional engineering knowledge and corporate memory. “By being more innovative, we may be gaining advantages in cost and capability, but we are also reaching far back, past the current experience base, and recreating older failures,” Eleazer wrote. “Most importantly, what is being lost is not just specific knowledge… but an entire body of knowledge, a philosophical approach to engineering, a unique attitude.”

Unfortunately, the new Obama space plan smacks of TSPR.

In some sense, anyone can design a spacecraft on paper. But the decades of institutional memory, expert systems engineering experience, and management skills for large space projects that will be lost by dissolving NASA’s role in spacecraft development can never be recovered. And it was just this long-term institutional capability that allowed NASA to be so successful in the past. As Young explained in his testimony, “successful management occurs when the continuity of expertise of NASA or the Air Force or the NRO is combined with the implementation capability of industry.” Government-civilian partnership in areas like space and defense should be a synergy, not a buyer-dealer relationship.

The Galaxy and the requirements problem

But wait, you might say. Why doesn’t NASA then adopt a hybrid approach whereby they contract with industry in such a way that NASA is only responsible for a fixed, pre-agreed cost, but still retain project management oversight? This way, the contractors bear the responsibility for keeping costs on target, but NASA retains its management responsibilities.

That’s been tried too, and the result was another acronym that strikes fear into the heart of defense professionals: Total Package Procurement (TPP).

TPP came out of the McNamara DoD of the 1960s, when it was felt the government was paying too much to develop new aircraft, with too many cost overruns originating from the contractors. The new idea was applied to the procurement of the massive C-5 Galaxy cargo aircraft, built by Lockheed. The TPP concept was that the government would basically pay one negotiated, upfront price for the entire “cradle-to-grave” development of the whole system. In other words, the government would pay Lockheed its fixed, prenegotiated sum, and then Lockheed, by hook or crook, would deliver a working airplane. Lockheed would pay for any cost overruns. If Lockheed delivered the aircraft under the fixed cost, it would get to keep whatever extra money remained. Logically it would be in the interest of the contractor to deliver the aircraft for as low a price as possible. Sounds like a great way to avoid cost overruns, right?

It did—and then reality set in. When Lockheed constructed the first C-5 Galaxies, the Air Force ruled that Lockheed’s landing gear had to be modified to make them more rugged. After the landing gear redesign, however, a new problem cropped up: since the new, tougher landing gear was now heavier, the aircraft came in a few thousand pounds overweight (out of nearly 380,000 pounds (172,000 kilograms) of empty weight). The Air Force command that was procuring the aircraft was prepared to accept it, but upper-level Pentagon management refused on the grounds that the whole point of TPP was to not allow Lockheed leeway on the product it delivered. Lockheed could not change the fuselage dimensions because they too were mandated under the rigid TPP requirements. Lockheed was thus forced to remove weight from the wings, one of the only places on the aircraft where they were contractually allowed to do so. During this period, the repeated delays and redesigns caused the C-5 costs to skyrocket by more than 250%. Lockheed was forced to cover these costs and it nearly bankrupted the company, which would have had disastrous effects on the nation’s aerospace workforce. Years later the C-5 fleet had to have its weakened wings replaced, costing more billions of dollars. The lesson to be learned? By holding the contract requirements to be more important than the quality and nature of the final product delivered, the government nearly bankrupted Lockheed and received a substandard product years behind schedule and billions over budget.

While TPP is long dead, its lessons are not. One of the key factors that makes government and military products so much more expensive than civilian versions are the intense and stringent requirements levied on these products within the contract. This is what can lead to things like “$1,200 pliers”; it’s not that the pliers themselves cost that much, but the pliers must be modified so that they can withstand a nuclear blast or an airplane crash.

Companies like SpaceX are extraordinarily good at launching rockets on the cheap, but this is precisely because they have been free of NASA’s requirements.

Undoubtedly among the most stringent requirements that can be levied by any government agency are those necessary for human-rating a spacecraft. The standards required to certify that one can safely launch astronauts to the satisfaction of the government—or private industry—boggle the mind. Given NASA’s longstanding and justified concern for the primacy of safety, it is difficult to imagine that they will relax those standards in the slightest for these private spacecraft services. And since NASA will be the only customer for these spacecraft (at least for five to ten years), it will wield enormous leverage over the standards to which these spacecraft will need to be built. Again we have a contradiction: if NASA will require these companies to meet its human-rated safety standards, and given that NASA already buys its spacecraft from private contractors, how can anyone expect this acquisition to be cheaper than current spacecraft procurement?

Companies like SpaceX are extraordinarily good at launching rockets on the cheap; but this is precisely because they have been free of NASA’s requirements and the standing army of technicians, scientists, accountants, managers, auditors, and comptrollers that all ensure that these requirements are being met. Once such a company is producing NASA spacecraft built to NASA standards, well… you’re not really doing business in a new way, except that the company may have slightly more freedom in the design stage than it otherwise would.

It may be that NASA will relax its requirements and allow more freedom to its contractors. But this is not a realistic expectation. As mentioned, it contradicts NASA’s history to believe that it would suddenly devalue safety requirements, because a space accident that claims the lives of astronauts will always be far more troubling and public for NASA than a spacecraft contract that costs $6 billion extra and is four years behind schedule. This is as it should be, but the point is that NASA is not an organization that is structurally prepared to be open to low-cost, relatively risky manned spacecraft operation.

For this reason, it is difficult to understand how the Obama plan will change NASA’s spacecraft acquisition model in such a way that the resulting product will reduce costs. To express this another way, it is unclear how Obama’s policy, as outlined, will differ from the way in which NASA currently procures its spacecraft unless NASA is a) willing to compromise on safety and/or other requirements, which is unlikely for the reasons outlined above; or b) adopts a “hands-off”, TSPR-like approach to the development of the vehicle, saving itself dollars and responsibility but potentially putting at risk the schedule, budget, and performance of the final product.

page 2: a few little things >>