Editor’s Note: An exclusive interview with Gen. Hyten is available here with more details.
Details are emerging about another restructuring of the contract for the new GPS ground system, a deal that pushes completion of the project back another two years and recasts the remaining work to fit within the Air Force’s strained financial profile.
Editor’s Note: An exclusive interview with Gen. Hyten is available here with more details.
Details are emerging about another restructuring of the contract for the new GPS ground system, a deal that pushes completion of the project back another two years and recasts the remaining work to fit within the Air Force’s strained financial profile.
Called the Next Generation Operational Control System, or OCX, the new ground system involves a rebuilding nearly from scratch of the infrastructure that monitors and controls the GPS satellite constellation. Project terms call for prime contractor Raytheon to replace or refresh all the hardware and software at the Master Control Station at Schriever Air Force Base, Colorado, and the Alternate Master Control Station at Vandenberg Air Force Base as well as upgrading four legacy antennas and 17 monitoring stations scattered around the world. The company will also install all new hardware and software into an existing building at Schriever to create a transition software facility.
The new schedule delays the delivery of the full OCX system by another 23 months to 2018 — about two years after the first of the new GPS Block III satellites is supposed to be available for launch and three years after OCX was originally supposed to be completed.
The lag between the availability of OCX and the new generation of spacecraft is particularly important because the GPS III satellites cannot be brought online and made operational without OCX. They can be launched — in fact, Raytheon will soon begin delivery of the Launch and Checkout System (LCS), which can be used (as the name implies) to loft and assess the post-launch health of satellites including GPS III. But LCS is an interim step in building the system, and use of the new signals and other capabilities of the modernized satellites will have until OCX is completed.
The LCS also has been delayed — but only by about a year. Raytheon will start delivering hardware to Schriever Air Force Base, the operational home for the LCS, in December. The hardware will undergo testing there, and acceptance is now scheduled for November of 2015.
Because the GPS III satellites have also been delayed, with the first satellite expected to be available in 2016, that lessens somewhat the gap between the availability of the satellites and the completed OCX.
“The schedule that we’ve completed is approved by the Air Force,” said Matthew Gilligan, a vice-president at Raytheon and the OCX program manager, in an exclusive interview with Inside GNSS. “It meets the needs of the enterprise.”
New Chapter, Old Story
This is not the first time that the OCX program has been rejiggered and delayed. In fact, the LCS was created in 2012 as part of an effort to reduce the effect of delays triggered when the Air Force asked to Raytheon to find some $100 million in cost savings. The resulting reorganization included modifications to the technical baseline, formal recognition of shifts in the contract’s requirements, and reductions in the contract’s scope to “address affordability challenges,” according to then-Raytheon spokesman Jared Adams.
The initial OCX contract, awarded in February 2010, had a baseline duration of just over six years. Delivery of the ground system — which included the new hardware and software for the control stations plus the deployment of advanced monitor stations at remote sites — was expected in 2015. That first contract was worth $866 million and had options for five more years of sustainment work. Altogether the total potential value of the original deal was $1.535 billion.
Gilligan declined to reveal the value of the latest contract, referring the question to the Air Force. The Air Force did not respond to a query about that and several other questions related to OCX.
Without the numbers it is not possible to tell if the OCX program is nearing a level of cost overrun that would require a report to Congress. The Nunn-McCurdy provision in the 1982 National Defense Authorization Act requires Congress be notified if a program’s cost exceeds the original estimate by 15 percent. The program is to be terminated if the cost overrun tops 25 percent unless the secretary of defense certifies the criticality and unique need for the program.
That would seem like it should be a source of serious concern for OCX. Actually, not so much.
“Nunn-McCurdy was in response to what Congress and probably the rest of the known universe perceived as runaway inflation in major military programs — and that goes back years,” explained Frederick Downey, a consultant who served for 13 years as senior counselor and military and foreign policy legislative assistant for Senator Joe Lieberman before becoming vice president for national security for the Aerospace Industries Association. “Most major programs that I’ve been familiar with had substantial inflation over the life of the program.”
Few of those programs actually get cancelled, he said, and he thinks it unlikely the OCX program would be at risk, especially in light of the need for the GPS constellation.
In his view, he said the program “is so critical, both to joint war fighting and GPS to the nation at large, that there’s really is little risk that the program would be canceled. You can always be surprised in Washington but if you are going to be surprised in acquisitions — especially in major acquisition programs — you’re more likely to be surprised that they go on as long as they do, not that they get cut.”
Cybersecurity Concerns
The complexity of the OCX software is widely cited as the main reason for the delays. Raytheon is not reusing any of the existing ground station software. It is replacing it with entirely new code that incorporates sophisticated information assurance (IA) measures – that is, the software is being comprehensively structured to weather cybersecurity attacks, discussed in a November/December 2013 Washington View column.
The need for such security is clear — particularly given the dependence of several categories of critical infrastructure on the timing and positioning information obtained from the GPS system.
The number of attacks on computer systems, including state-sponsored cyber incursions, appears to be increasing. The Senate Armed Services Committee released a report September 17 saying hackers with the Chinese government successfully penetrated the computers of military contractors involved in the mobilization and deployment of U.S. military forces. The same day the Wall Street Journal reported that U.S. action against the militant group Islamic State could trigger cyberattacks.
Cybersecurity is ”the fastest-growing threat facing the United States of America,” Michael Morell, former deputy director of the CIA told Inside GNSS in November. “The threat is here today. The number of intrusions into government, military and commercial networks is massive.”
But staving off cyberattacks is a monumental challenge. In Raytheon’s case programmers are not only writing new OCX code, they are assessing and upgrading the commercial off-the-shelf (COTS) and free open source software (FOSS) code incorporated into the system.
The COTS and FOSS software, which includes well-established packages like Linux, is being integrated as a way to keep costs down. The COTS/FOSS code, however, is not being incorporated blindly into OCX. Each package is being evaluated and refactored — that is, edited and updated — to erase the imprint of old, less-than-secure coding habits and incorporate up-to-date countermeasures against new threats. As of a year ago some two million lines of code had been refactored and many millions more had been assessed.
Raytheon representatives proudly point out that the Air Force considers OCX to be a “pathfinder program” when it comes to information assurance. So, Raytheon’s work on OCX is setting the bar for IA programs that follow. They acknowledge, however, that implementing the IA provisions has been time-consuming.
“One of the drivers for the extended schedule, one of the primary ones,” said Gilligan, “has been the pathfinding in this level of IA compliance.” To go through the scanning and mitigation for each of the FOSS packages, he said, “was something that took a fair amount of time.”
But what does that really mean?
Gilligan told Inside GNSS in late August that 70 to 80 percent of the IA work was complete. To put things into perspective, he said in an interview at the end of 2013 that 70 percent of the IA work had been finished.
To move things along faster Raytheon had been doing segments of the work concurrently — now that is going to change.
“We started to get a stackup of some of the iterations [or work packages]. So, we had concurrent engineering of some software development and some of the systems engineering — some of the systems engineering that goes on in parallel,” Gilligan explained.
Raytheon will now “decompress” the overlapped tasking, he said. Starting at the beginning of this year the company made an effort to complete all of the segments and elements-level systems engineering for the balance of the program — something Gilligan expects will be completed this fall. This, he said, will be “a lower risk approach to completion.”
Money Worries
Hopefully, the changes will go a long way towards keeping the program on schedule. But the complexity of the coding and integration of the IA requirements is not the only factor in the OCX delays. The program has faced funding challenges as well — and those appear likely to continue.
Congress has cut funds for the project repeatedly, both directly through reductions in the OCX line item in the Department of Defense (DoD) budget and indirectly through cuts in the money requested by the Department of Transportation for civil GPS funding. Money from both sources goes to fund the program but the civil funding shortfall has been particularly nettlesome.
Over the last couple of years, experts familiar with the civil funding request have told Inside GNSS that the Federal Aviation Administration, which is tasked with requesting and forwarding the money for the civil contribution on behalf of the whole community, has failed to sufficiently support the request before Congress. A glitch also occurred in the way the forwarded money was handled for a time. The GPS program apparently delayed obligating the monies, giving Congress the impression the funds were surplus which prompted further cuts.
As a result of this spotty funding history the civil community was, as of last year, some $50 million to $100 million short on its original commitment to the GPS program. This year the House zeroed out the $27 million request and the Senate slashed it top $10 million. The prospects for restoring the full $27 million, much less getting any more, appear grim at best.
The DoD has stepped up in the past to make up the shortfall and keep the program on track. Gen. John E. Hyten, the new commander of Air Force Space Command (AFSPC) recently hinted he might support doing so again.
“Unfortunately, our civil partners have faced the same fiscal scrutiny that we in the military have had to contend with over the past several fiscal years,” Hyten said in a written answer to a question on the issue submitted by Inside GNSS. “The result has been a delay in the amount of money the civil community is paying for their share of modernized GPS civil services. The GPS Enterprise impacts result in delays to the Next Generation Operational Control System (OCX) program. This can be avoided if the DOD funds civil funding shortfalls in order to keep the program on its revised schedule.”
‘2016 Scares Me’
It is not entirely clear, however, that AFSPC will have the resources to follow through — in the near term or in the long run. Although the GPS program got nearly all of the money it requested for fiscal year 2015, program managers are weighing a new competition for a contractor to build the last tranche of GPS III satellites, and contests cost money. Sequestration is also kicking in again in 2016 and the pressure is on. Hyten warned the impact of sequestration would be dire.
“If we go to a sequestration budget it will break Space Command,” he told Space News in a recent interview. “We’ve already taken every fungible asset we can try to figure out. Now we have to go after real capability. What is it you want to stop doing? You want to stop doing GPS? You want to stop doing [missile warning]? You want to stop doing ground-based missile warning? You want to stop doing deep space surveillance? What mission do you want to stop doing?”
Hyten said that AFSPC officials had “put real options on the table” as a way to persuade Pentagon and congressional leaders of the difficulty of the budget situation. Although he offered no real details on what those choices might include, he specifically mentioned ground-based assets, including radars and telescopes, as things that might get cut.
“The options are close this, close this, close this,” he said. “The worst case would be a constellation, but I don’t think anybody believes that the United States will tell us to shut down a constellation. That means you have to go to the ground-based architecture and start closing down elements of the ground-based architecture.”
One possible option, and one the Air Force has used before, is to slow modernization and keep the GPS constellation as it is. The system is working, after all, and there is a ground system in place — albeit one without the IA elements that are being integrated into OCX. What that means for the OCX program and for all those who depend on GPS positioning and timing signals for communications, electrical power grids, and other critical services is potentially stark.
Hyten may have summarized it best: “2016 scares me.”
