It is now 11 years since the Air Force announced that its Cold War missile warning satellites would be replaced by an ambitious ground and orbital architecture designed to accomplish a range of missions for diverse users. By the time that announcement was made in 1995, policymakers had been debating how to replace aging missile warning satellites for over a decade, and planners were growing concerned that the anonymously designated Defense Support Program satellites might cease functioning before a successor was in hand.

Loss of missile warning capability would be a true crisis. Unless U.S. leaders are alerted to the launch of hostile missiles in a timely manner, they cannot authorize the retaliatory strikes that lie at the heart of nuclear deterrence. Moreover, lack of early warning about approaching ballistic warheads would leave military and civilian personnel alike unable to protect themselves by taking cover or launching interceptors.

So missile warning really is a big deal, arguably the most important mission that space-based sensors accomplish. If the policymakers who conceived the “Space Based Infrared System-High” back in 1995 had been content to address that one overriding mission, a new generation of warning satellites might well be operational today.

But decision makers were not satisfied to replicate Cold War capabilities. The Soviet Union was gone, and missile defense looked more feasible against the rogue states that now occupied center stage in U.S. threat assessments. Also, there was pressure to justify the high cost of new reconnaissance satellites by making their output more relevant to the needs of tactical commanders.

So policymakers embraced an architecture that could not only provide missile warning, but one that could also support missile defense; provide battlefield characterization to local commanders; and collect technical intelligence for spy agencies.

The resulting complexity, compounded by mistakes in awarding and overseeing the contract to develop the satellites, has put SBIRS-High (or just SBIRS, “sih-bers,” as it is known today) billions of dollars over budget and years behind its original acquisition schedule. Not surprisingly, Congress and the military began to wonder whether the program was irretrievably flawed.

Late last year, amid new reports of cost growth in the program, Kenneth Krieg, Under Secretary of Defense for Acquisition, Technology and Logistics, disclosed a new approach to SBIRS that would reduce the scope of the program, focus on meeting missile warning needs fast, and provide a backup plan for satisfying that need even if the baseline program failed.

It was a plan that seemed to make sense, not only because of the high priority that strategists assign to reliable early warning, but also because SBIRS had become emblematic of all the missteps made in acquiring space systems during the Clinton years. Under-funded, overly complex and loosely managed, SBIRS was, in the words of a Defense Science Board task force, “a case study of how not to execute a space program.” With other space initiatives from the same era being restructured, it was only fitting that the faltering SBIRS be reorganized too.

Unfortunately, the backup plan won’t work, and is a clinical example of why the Pentagon keeps making the same mistakes over and over again in its space programs. It is a reflection of what happens when technically shallow people are put in senior positions, and then serve short terms covering only a portion of the time needed to bring complex systems to fruition.

Among the problems that SBIRS has encountered, two technical hurdles stand out as the most important drivers of delay. The first was a decision to equip geosynchronous satellites in the constellation with both scanning and staring infrared sensors. Cold War satellites carried only scanning sensors, and staring sensors are much more complicated. The second hurdle was the need to develop millions of lines of software code that could optimize the value of these sensors and their processed output for a variety of users.

As reporter Jeremy Singer pointed out in Space News on Dec. 19 [“Pentagon Scales Back SBIRS Program,” page 1], the latter hurdle was raised even higher by using an outmoded software language called Ada. The processor incorporated into the baseline spacecraft also looks antiquated, given technical advances in the years since it was selected. But the basic problem is that SBIRS is trying to do diverse things no one has ever done before, and the required technology takes a long time to perfect.

If you think you’re running out of time, and there’s a danger that basic missile warning requirements might not be met as legacy satellites go silent, the logical thing to do is cobble together a backup plan from mature technologies that can be deployed quickly.

That is not what the Pentagon has done. Instead, it has selected an even more complicated technology that currently exists only in laboratories, and has decided to scale up this so-called “wide-field-of-view” capability for use in the missile warning mission. The good news is that the proposed technology really could offer better performance than SBIRS. The bad news is that there’s no way the alternative sensor and software will be ready in time if SBIRS continues to falter.

Even a plain-vanilla sensor and software suite based on well-understood technologies would take the better part of a decade to design, develop, integrate and test, meaning it would not be completed until the eve of the 2015 launch threshold that policymakers have said must be met to avoid losing missile warning capabilities. So embracing the much more challenging wide-field-of-view system as an alternative to SBIRS increases program risk rather than reducing it.

Defenders of SBIRS say the alternative sensor is technology overkill, because it offers a level of performance far in excess of the requirements specified for SBIRS. But that’s beside the point — the intelligence community and other users would undoubtedly benefit from better performance. The real issue is that the alternative approach simply won’t be available within an acceptable timeframe.

How is it possible that policymakers could make such a gross mistake against the backdrop of so many similar missteps in space in recent years? Most of those previous “errors” were directly or indirectly traceable to naive optimism about the performance and pacing of new technology, and yet decision makers seem eager to repeat the mistake again.

It is understandable that policymakers don’t want to keep investing money in a system that seems outmoded. But the challenge they face on SBIRS at this point in the program’s history isn’t how to maximize future capabilities, it’s how to assure that essential mission needs can be met as legacy systems run down. The backup plan the Pentagon is pushing doesn’t do that, and thus raises doubts about whether policymakers have learned anything from their past mistakes.

Loren Thompson is chief operating officer of the Lexington Institute and a defense consultant. His clients include SBIRS prime contractor Lockheed Martin. He taught military technology courses at Georgetown University for 20 years.