On October 29, 2009 the Government Accountability Office (GAO) issued a report entitled, “DEFENSE ACQUISITIONS: Challenges in Aligning Space System Components” (GAO-10-55).  The report surveyed 8 DOD satellite programs with aggregate total mission costs currently expected to exceed $50 billion. The DOD satellite programs in the GAO study were:  Advanced Extremely High Frequency (AEHF) Satellite, Mobile User Objective System (MUOS), Wideband Global SATCOM (WGS), Space-Based Infrared System (SBIRS), Space Tracking and Surveillance System (STSS), NAVSTAR Global Positioning System (GPS), Space-based Space Surveillance (SBSS), and National Polar-orbiting Operational Environmental Satellite System (NPOESS).

 

The table below summarizes the programs and the dollars associated with each.

Summary of DOD Satellite Systems Reviewed by GAO
		FY 2009 Dollars in Millions
DOD Mission	Program Name	Total RDT&E Cost	Total Procurement Cost	Total Mission Cost
Communications	AEHF	$7,267.3	$3,150.1	$10,417.4
Communications	MUOS	3,807.7	2,671.4	6,479.1
Communications	WGS	377.9	1,738.0	2,115.9
Missile Warning/ Tracking	SBIRS	7,724.7	2,655.7	10,380.4
Missile Warning/ Tracking	STSS	2,174.0	N/A	2,174.0
Navigation/Timing	GPS	4,485.9	4,937.6	9.423.5
Space Object Tracking	SBSS	514.1	N/A	514.1
Weather	NPOESS	$8,125.5	$2,943.4	$11,068.9
Totals		$34,477.1	$18,096.2	$52,573.3

DOD satellite systems typically have up to three main components:  (1) satellites, (2) ground control systems, and (3) user terminals. According to GAO, ground control systems are generally used to download and process data from satellite sensors (and disseminate this information to warfighters and other users), and to maintain the health and status of the satellites (including steering the satellites and ensuring that they stay in assigned orbits).  User terminals are typically procured by the military services and managed separately from associated satellites and ground control systems; they can range from equipment hosted on backpacks to terminals mounted on Humvees, airborne assets, or ships.  Terminals can be used to help the warfighter determine longitude, latitude, and altitude via GPS satellites, or securely communicate with others via AEHF satellites. The GAO study focused on how DOD aligns (or fails to align) development and operational capability of each of those components, in order to maximize satellite utilization and deliver maximum capability to the warfighters.

 

GAO found that “for six of DOD’s eight major space system acquisitions, DOD has not been able to align delivery of space assets with ground assets, user assets, or both.… For the five space systems requiring user terminals, none were aligned. In some cases, capability gaps resulting from delays in the fielding of ground control systems or user terminals are 4 or more years.”

 

Unlike many GAO reports we have reviewed, in this one GAO took pains to provide context for its findings, stating—

 

Notwithstanding the fact that alignment gaps are undesirable, several factors provide insight into the inherent challenges associated with managing alignment. First, alignment may be relatively easier to achieve in some programs than in others. For example, some space systems may require only a ground system or few user terminals and may even manage these acquisitions within one organization. By contrast, other programs may require literally tens of thousands of terminals that must be installed on a wide span of weapon systems, including ships, planes, vehicles, and even other space systems—which are owned and controlled by various military services. Second, an inherent difficulty in aligning satellite launches with ground and user terminal programs is the lead time needed to schedule satellite launches—about 2 years—which makes it difficult to hold back satellite deployment if a ground or user terminal is experiencing a considerable delay…. Third, it is difficult to measure the extent to which warfighters and other users are being affected by delayed capability or even the extent to which capability is delayed. … satellites themselves only offer initial capabilities until enough satellites have been launched to provide the coverage needed to achieve full capability.  This process alone can take years and will vary system to system as the number of satellites required to achieve full operational capability depends on mission requirements and coverage offered by satellites, among other factors.  At the same time, ground control systems can be delivered in phases, the first of which may focus solely on controlling and maintaining the health of the satellite, with subsequent phases delivering software that can collect and process sensor data. User terminals can take years to install as they can span a broad spectrum of weapon systems and their installation is usually done alongside other upgrades.

 

GAO commented on individual satellite programs—

 

 

• GPS achieved full operational capability in 1995…. However, the plan and the capabilities it is being designed to provide have been delayed and are significantly over budget. As a result, some new capabilities are not now available to the warfighter because the ground control system features needed to command and operate the capabilities have not been completely delivered. For example, updated user equipment possessing a capability to prevent spoofing of navigation information started being delivered to the warfighter in 2004. However, the Architecture Evolution Plan, representing the current ground control system, is not capable of providing two important aspects of this capability and is not expected to do so until early fiscal year 2010.

 

• The first SBIRS satellite5 will carry scanning and staring sensors designed to provide early missile warning capabilities. However, DOD will not be able to fully utilize the data collected from the staring sensor when this first satellite launches, currently planned for September 2010, because the ground control software that is to process the sensor’s data is not planned to be fully functional until at least 2014. This means that complete, usable data from the staring sensor will not be available until about 4 years after the satellite is on orbit.

 

• The first WGS satellite launched in October 2007, but its associated ground mission planning software—the Consolidated Network Planning Software—does not work properly. This planning software was designed to compute required bandwidth for all users simultaneously accessing WGS satellites. It would then disseminate that information to various satellite operation and support stations located globally so that all stations had a real-time view of the availability of WGS satellite capabilities. However, because the development of the mission planning software has had problems and is not well coordinated with WGS satellite capability, the dissemination of information does not occur as designed, and the information has to go through a time-consuming and labor-intensive work-around through a single ground station before it reaches the warfighter.

 

GAO commented on alignment of user terminals as follows—

 

 

• The Air Force’s FAB-T program is designed to provide antijam and protected communications for nuclear and conventional forces as well as many airborne assets and ground command posts. As one of the primary user terminal programs associated with AEHF, FAB-T has recently experienced numerous problems and is not currently aligned with the AEHF satellite program. Specifically, contractor performance problems, which caused design teams to be restructured to improve performance and efficiency, caused a delay in the start of initial production from fiscal year 2007 to fiscal year 2010. In addition, design changes and contract cost growth have more than tripled development costs since the contract was first awarded. While AEHF will be able to provide capability through other user terminals, current estimates show that FAB-T will only have 2 percent of its terminals fielded when AEHF is scheduled to reach its initial operating capability in 2011. Further, estimates are that FAB-T will not have all of its terminals fielded until fiscal year 2019.

 

• JTRS is a family of interoperable, digital, modular, and software-defined radios that is planned to provide the capability to receive, transmit, and relay voice, data, and video. In the past, tactical military radios could not work well with each other. The JTRS radio is also being designed as the primary user terminal for the new MUOS satellite capability to help the warfighter achieve information superiority. Although MUOS will be able to provide capability through other, legacy user terminals, DOD estimates that less than 20 percent of JTRS terminals will be available to access the MUOS satellite when it achieves operational on-orbit capability in December 2011. In 2014, when MUOS is expected to reach full operational capability, 32 percent of JTRS terminals are expected to be available to the warfighter. DOD expects to field all the needed JTRS terminals by 2021—about 7 years after MUOS is expected to be fully operational. … Officials from one warfighting command (users of the capability) told us that because of the 2-year gap between when all MUOS satellites reach on-orbit capability and when the MUOS-capable user terminals (JTRS) first become available, the MUOS satellites will have spent a portion of their expected lifespan less than fully utilized. This issue concerns the combatant command because MUOS is replacing the aging Ultra High Frequency Follow-On space system, which currently serves more military customers than it was originally designed to handle. While waiting for the JTRS capability, the command will likely have to lease commercial satellite capability and user terminals to increase bandwidth capacity and improve the speed and effectiveness of information and communication transfers.

 

GAO summed up other problems.  We report them here verbatim because of the potential to learn lessons from them.

 

The satellite, ground system, and user terminal programs we studied have had execution problems that have caused substantial delays in schedule that in turn have made it more difficult to align delivery of all three space system components. Most prominent are requirements changes, technical problems resulting from underestimation of complexity, and poor contractor oversight. The first satellite delivery of SBIRS, for instance, has been delayed at least 7 years in part because of poor oversight, technical complexities, and rework. The first satellite delivery for NPOESS is over 4 years late. AEHF has experienced delays of about 3 years for these reasons along with requirements changes that occurred earlier in the program and difficulties meeting information assurance requirements for its satellite. The GPS IIF system has also had about a 3-year delay because of technical and workmanship problems and requirements changes. Ground systems and user terminals have experienced similar problems. JTRS, for example, has experienced significant delays because of problems in maturing critical technologies, and as noted earlier, FAB-T delays have occurred because of contractor performance problems. Also, as noted earlier, the WGS ground system has experienced technical problems that have prevented it from working properly with WGS satellites now in orbit.

 

*****

 

Underestimating software complexity has also been a problem. The complexity of software on any system, including space systems, is often denoted by the amount of software, or number of lines of software code. Generally, the greater the number of lines of code, the more complicated the software system development, and ground control systems typically require significantly more software than the satellites. This means that software development for ground control systems is oftentimes the higher risk. In some cases, unanticipated software complexity can lead to lack of synchronization between the satellite and ground systems of space system acquisitions. For example, on the AEHF space system, the prime contractor has experienced quality control problems with the software for the mission planning element of the ground control system. In testing so far, the government has identified numerous significant software deficiencies and continues to find deficiencies as testing continues. Ground control system fielding will be delayed until the deficiencies are corrected. Also, our past work has shown that the MUOS ground control software represented one of the greatest risks to the program because of the size and complexity of the design. On SBIRS, the total estimated lines of code on the ground control system software grew from approximately 1.55 million in August 2004 to approximately 1.88 million in December 2008.

 

*****

 

Space system acquisition programs sometimes shift funds from the development of ground control systems to their associated satellite development efforts to meet unexpected obstacles—an action that can create new problems. For example, when the GPS IIF satellite program encountered development problems, the program shifted funds set aside for the GPS ground control system to address the satellite problems, causing a delay in the delivery of some ground control capabilities. Similarly, SBIRS officials reallocated funding from the ground control system to address satellite software issues, which may have contributed to the system’s initial inability to utilize the staring sensor data from the first geosynchronous earth orbit satellite. Program officials told us that they like the flexibility of being able to move funds from ground control systems to the satellites if priorities warrant. However, as we indicated above, this can put the development of ground control systems at a disadvantage compared to development of the satellites for space systems, for example, GPS and SBIRS.

 

We don’t have much to add to GAO’s insights.  We continue to be proud of America’s technological prowess and worried about the apparent lack of management oversight given to DOD’s satellite programs.  These are important—critical—space systems, and it is important that they be given appropriate attention by the various stakeholders.  See the entire 37 page GAO report here.