The Space Reviewin association with SpaceNews

SBIRS satellite illustraton
Problems with military space programs like SBIRS (above) are the result of changes the Air Force made years ago about how it procured and managed space projects. (credit: Lockheed Martin)

The third shoe

The FY 2006 Department of Defense budget slashed funding for a number of military space programs. According to Aviation Week, the Transformational Satellite (TSAT) communications system was cut by $400 million, the Space Based Infrared System (SBIRS) High was cut by $50 million, and the Space Radar Program lost $125.8 million. All together, that is a cut of over $575 million; as the late Senator Everett F. Dirksen might have said, at this rate pretty soon we are going to be talking about some real money.

The cuts were hardly a surprise. The fact that a number of Air Force space acquisition programs were not doing too well has been obvious for some time. On top of the most recent concerns, the past few years have seen the Air Force pick up the tab for a far greater share of the Atlas 5 and Delta 4 program costs than was envisioned. Add to that the virtual abandonment of once-key aspects of the largest launch range upgrade program ever conceived, Range Standardization and Automation, and you have a less than sterling record.

How did this come about? Well, like most things in the military it was carefully planned and took a long time to accomplish.

So daunting were the challenges presented by the military utilization of space that it required a new way of doing business.

The USAF’s space efforts developed in an unusual manner. Satellites transitioned from scientific toys into military necessities with incredible speed. During the late 1950s and throughout the 1960s public attention was focused on the highly publicized and dramatic manned missions, but the most important progress was being made with unmanned satellite programs, especially those with military applications. Within a decade of the launch of Sputnik, military satellites were providing unequaled capabilities in such areas as weather forecasting, reconnaissance, navigation, and missile early warning. Other promising applications such as communications were coming up fast. So rapid was the pace of development that military space systems never really got out of the research and development phase; the success of one new system merely encouraged the development of an even more capable one. Rarely were any two sequentially launched satellites completely identical. “Tests” and actual “operations” became interchangeable in both terminology and approach. The normally torrid pace of aircraft development looked glacial in comparison to spacecraft.

So daunting were the challenges presented by the military utilization of space that it required a new way of doing business. One reason for this was the rapid pace of development, but an even more important reason was the unique manner in which the required tasks had to be accomplished. Specifying, designing, developing, assembling, processing, launching, and even operating a satellite system became one continuous process. All this meant that the skills and techniques required were so specialized and unique that only the spacecraft and booster builders themselves could handle the tasks required. But the most important driver of all was the fact that every launch was a test flight, and there were no second chances.

Spacecraft simply could not be delivered to a loading dock; ultimately they had to be delivered to the correct orbit, and in proper operating condition. First, though, the spacecraft had to leave the builder’s production facility and be delivered to what was, essentially, another factory. The same situation existed for the booster and the rest of the components required for the entire flight vehicle. All of the hardware that had to be assembled for a launch first encountered the rest of the equipment required at the launch base, and nowhere else. Spacecraft were never mounted on a booster until they got to the launch pad. All of the booster components were not assembled until they got to the launch base as well; in some cases this even included engines and flight control systems. Then there were the ground systems: boosters and spacecraft were checked out individually at the factory but usually were not interfaced with the actual ground control and support systems until they arrived at the launch base. Frequently there was no testing of such vital interfaces until they got to the launch facility itself.

And the worst problem of all was that there was no way to simulate the actual flight conditions—and there was no turning back after liftoff. Every launch was a test flight, because the hardware had never flown before. In many cases not even that particular combination of hardware types had been flown before.

The organization that was created to handle all of this management effort was perhaps not unique in structure but was unprecedented in its scope.

All this meant that the pre-launch processing at the launch base could only be viewed as an extension of the factory process, and processing and launch operations as part of the acquisition task. Contractors would provide the highly specialized skills required to accomplish all of the hands-on work, just as they did in their own factories. The System Program Offices (SPOs) would manage the contractors, just as SPOs did for production of aircraft and other more traditional military equipment. The SPOs would assure that contract requirements and the applicable laws and regulations were followed and also ensure that the Air Force conducted a thorough technical evaluation of the entire process.

The organization that was created to handle all of this management effort was perhaps not unique in structure but was unprecedented in its scope. The SPOs wrote, issued, and managed the contracts in all of their aspects, aided by government representatives at the contractor’s factories. The SPO “arms in the field”, the Aerospace Test Groups at Cape Canaveral Air Force Station and Vandenberg Air Force Base, controlled the launch process, including pre-launch processing activities, but did so under the authority granted them by the SPOs.

Meanwhile, there were SPOs at the launch ranges concerned with procuring range equipment: radars, computer systems, optical tracking gear, telemetry processing hardware, and communications electronics. Such an effort required an engineering organization at the ranges in order to understand, specify, and provide technical oversight of these procurements. Since handling all of the requirements of operating a test range and meeting a wide variety of users’ needs demanded considerable engineering expertise, this was neither out of character for the ranges nor beyond their intrinsic capabilities.

The Air Force Systems Command organization formed to accomplish all of this was known at various times as the Space and Missile Systems Organization, the Space and Missile Systems Center, or simply as Space Division. It was analogous to the other Air Force Systems Command product divisions that produced aircraft and electronics, but had much broader responsibilities. The three-star general that commanded Space Division was unique in that he had the whole ball of wax: specification, acquisition, production, test, launch operations, range support, and at least the initial phases of orbital operations. Unlike any other product division, he commanded not only the traditionally organized SPOs but also the ranges and launch organizations at Cape Canaveral and Vandenberg and the satellite control center at Sunnyvale, just south of San Francisco. The same general officer had responsibility for ballistic missile development and testing as well; there also were SPOs and related test organizations at the test ranges to accomplish that task.

In the early 1980’s this organization was expanded additionally to include the Air Force’s space-related laboratories, such as the “Rocket Lab” at Edwards Air Force Base. The concept here was that the labs ultimately were producing technologies that would be employed in the acquisition of space-related hardware and so should be driven by those requirements.

Thus, there was a “space force” with unique skills and techniques within the much larger US Air Force. While never recognized as a separate career field by the Air Force, the Space Division SPOs, ranges, test groups, and labs relied heavily on their ability to retain experienced people with the overall array of specialties they required. People were transferred between the larger Air Force and the space business quite regularly, but nonetheless Space Division was able to create and maintain a highly-experienced cadre of professional space acquisition and engineering personnel.

The space-related SPOs and other organizations eagerly sought out experienced officers and enlisted personnel. People that already had served in various space-related assignments not only were experienced in acquisition, but also had been indoctrinated in the unique attitudes and techniques demanded by space activities. They not only knew the organizational structure and the procedures required for procurement; they appreciated the consequences of failure - and failure was all too common.

In 1991 Space Command acquired the space launch ranges and launch organizations there. Overnight, many of the Air Force personnel at the launch ranges were no longer qualified to hold their jobs: they were now in “operational” rather than “acquisition” career fields.

This approach to manpower was essential to getting the job done but had its problems in terms of traditional military personnel management. Viewed from the perspective of the overall Air Force, someone who had rotated between assignments at the various space SPOs—Cape Canaveral, Vandenberg, Sunnyvale, and perhaps the associated labs or the ballistic missile SPOs—appeared to have acquired little broad experience. In fact, from a space acquisition perspective he had covered all of the bases, including launch operations. It was very difficult for the Air Force as a whole to recognize the kind of trained personnel required by the space acquisition business, and thus it proved to be well-nigh impossible to manage the people in a suitable fashion. This problem was never solved, and suddenly it got much worse.

Breaking the system

On September 1, 1982 Air Force Space Command was created, the intention being that it would become one of the operational Air Force commands similar to Strategic Air Command (SAC) and Tactical Air Command (TAC). Most space tracking and control assets were transferred to Space Command right away. These included the warning radars and analysis systems that SAC had gained with the shutdown of Air Defense Command a few years earlier, as well as the remote satellite tracking stations operated by Air Force Systems Command. The space acquisition workforce got a little smaller as a result.

Next, in 1991 Space Command acquired the space launch ranges and launch organizations there. The consequences of this transfer were enormous. Overnight, many of the Air Force personnel at the launch ranges were no longer qualified to hold their jobs: they were now in “operational” rather than “acquisition” career fields. Also, essentially overnight the “acquirers” of Space Division and the “operators” of Space Command had to figure out how to break up the factory-like launch process into its “acquisition” and “operational” elements. This was not easy; in fact, it represented a real unknown. Launch processing and operations had never been done that way and nothing had changed in terms of the technical requirements of the tasks and the considerable challenges that accompanied them. We were still launching exactly the same vehicles in exactly the same manner, and the results were often distressing.

By the early 1990s launch failures were far less common than in the early days of space flight, but failure was still far more likely than in any other endeavor. About 6% of space launches fail in some way, and about half of those represent failures to attain any kind of an orbit. Now, aviation related-analogies are favored by those who argue for an “operational approach” to space launch operations. To use such an analogy, if a similar failure probability applied to airplanes, about 150 airliners each day would arrive at an airport different than their intended destination and another 150 a day would arrive at a place on the ground where there was no airport of any kind (some might call this a crash).

The space acquisition career field—never truly a recognized specialty—had lost a large percentage of its jobs and a huge portion of its on-the-job training opportunities. Because of this, it was no longer a self-sustaining career field.

The operators of Space Command and the acquirers of Space Division gave the problem of dividing up a continuous process a lot of thought but failed to ever solve it satisfactorily. In one case, the two organizations agreed to give Space Command responsibility for launch vehicle integration “except for the engineering”, which would be done by Space Division. Since such integration is an engineering task, this was like giving someone the responsibility for all of the mathematics except for the actual calculations. A few years later, Space Command conducted a study that concluded that Space Division’s SPO’s had too much control over launch operations—but that this problem would be solved by a pending new agreement between the organizations that would give the SPOs more authority.

But aside from the difficulty inherent in such a task, another less-well recognized aspect was that the space acquisition career field—never truly a recognized specialty—had lost a large percentage of its jobs and a huge portion of its on-the-job training opportunities. Because of this, it was no longer a self-sustaining career field. To impact the Air Force’s aircraft procurement efforts in a similar manner you would have to close the test ranges at Eglin and Edwards Air Force Bases, where test programs are handled, and forbid personnel transfers to and from the Air Logistics Centers, where most aircraft major repair and modification work takes place.

The splitting up of space launch tasks was the biggest problem faced by the space SPOs, but it was not the only one. The SPOs relied on a specialized advisory contractor, Aerospace Corp., for much of their detailed technical expertise. In the early 1990s Congress mandated a reduction in funding for the company, which was forced to conduct its first reduction in force. Then Vice President Gore’s “Reinventing Government” initiative forced a 30% reduction in Air Force civilian personnel, and more experience went out the door. If that wasn’t enough, later in the ’90s the Air Force assistant secretary for acquisition directed that no SPO would have more than 50 people, regardless of the complexity of the task.

The impact of this draconian and largely arbitrary pruning activity was telling, but was not felt by just the SPOs. The National Reconnaissance Office (NRO) had depended on transfers of the most experienced personnel from the SPOs and launch bases to staff its own engineering, acquisition, and even operations efforts. By the late ’90s everyone was puzzling over the mysterious disappearance of the NRO’s once-legendary systems engineering capabilities. Then there was industry, which had always eagerly recruited experienced Air Force personnel, both the younger people and highly experienced retirees. One senior engineer with a major firm recently expressed his disappointment with the resumes he had been reviewing. One applicant might boast a series of impressive-sounding Air Force space launch assignments—all leading back to a liberal arts degree and no program office experience. Another resume might show an engineering degree combined with a hopscotch of assignments at various Air Force SPOs, but with no special experience in space programs. “I used to think that the space SPOs were made up of special people, that they got the cream of the crop. Not any more,” was how he summed it up.

The first shoe dropped in 1998 and 1999. Three out of four Air Force Titan 4 boosters launched from Cape Canaveral failed to deliver their payloads properly—and that wasn’t all. The first two Delta 3 boosters, launched in that same time frame, failed to achieve orbit, while one of the new Lockheed Martin Athena 2 boosters also suffered a failure, the second for the series.

The word is that the Air Force now is considering creating a cadre of space acquisition specialists. What a novel idea!

Such a series of failures simply was unprecedented. Not even the failures that shook the industry in 1985 and 1986 had been as numerous or as costly. While the Delta 3 and Athena failures had not occurred under Air Force control, they indicated that no element of the industry was problem-free—and even more importantly, that the Air Force could not rely on private company professionalism to make up for its own technical and managerial shortcomings. The Air Force convened an independent Broad Area Review Board that eventually concluded “You broke it; now go fix it.” To some this equated to “put it back the other way,” and, in fact, one member of the board, a former Air Force Chief of Staff, put it in just those terms. Problem was, it was not that easy.

The second shoe dropped soon after. A high-level study concluded that the space element of the Air Force should be managed under one organization, and that included the SPOs as well as the operational aspects. This was a radical suggestion, but the Air Force was in no position to ignore it: the leader of the study group was a remarkable gentleman named Donald Rumsfeld, and soon thereafter he was selected for a position that gave him considerable influence.

These latest funding cuts should be regarded as a “third shoe” dropping, or rather, perhaps, of it being banged on a podium, Khrushchev-style. The word is that the Air Force now is considering creating a cadre of space acquisition specialists. What a novel idea! Viewed from a longer perspective, this equates once again to the observation of six years ago: “You broke it; go fix it.” But perhaps first we really should understand how it was broken.