F-35 test aircraft are seen at Naval Air Station Patuxent Aircraft, Md. (Lockheed Martin)
Two months ago, top officers and senior Defense Department officials huddled to discuss a serious matter of national security.
But the meeting was not about Afghanistan, North Korea or even Iran. It was about the requirements for the F-35 Joint Strike Fighter, the centerpiece of DoD's future combat aviation arsenal. The program, launched in the 1990s, has suffered years of delays and billions of dollars in overruns.
The meeting's participants sought to decide if the stealth fighter's requirements were realistic and whether slight modifications could improve cost, schedule and performance. In the end, some changes were made. Whether those alterations solve or mitigate the program's problems remains to be seen.
The F-35 program is one of a dozen that have undergone such a requirements review. The aim is to vastly streamline the onerous process of vetting weapons programs so they get to the battlefield more quickly and smoothly — and, theoretically, at less cost.
DoD officials spent nearly a year overhauling how they decide to benchmark capabilities for new weapons, and how they monitor requirements of existing systems. This process of determining requirements is called the Joint Capabilities Integration and Development System.
"We're trying to make it more of a meaningful and useful process," Adm. James Winnefeld, vice chairman of the Joint Chiefs of Staff, said at an April 11 Center for Strategic and International Studies conference in Washington. "We have a ways to go."
One goal is to give program managers "requirements relief," Air Force Brig. Gen. Richard Stapp, deputy director for requirements on the Joint Staff, said in an interview.
Under the new approach, requirements can be traded for cost, schedule and performance, giving the process more of a "consumer approach," Stapp said.
"You don't want to spend the last 15 percent of your program budget on the last 1 percent of your requirement," he said, noting there could be exceptions should the system not be able to complete a mission.
"What I don't want to see happen is to have us spend millions and millions and millions of dollars to get that much closer to a key performance parameter," Winnefeld said. "We're trying to check some reality into that without hurting the war fighter."
Scaling down the process
In the past it could take anywhere from 18 to 24 months for the Pentagon's top requirements-setters — known as the Joint Requirements Oversight Council (JROC) — to approve the program parameters and goals of a new weapon system. The council is made up of the vice chairmen of the Joint Chiefs of Staff and the military services.
"In this kind of information age, that was not conducive to good military acquisition," Stapp said.
One contributing factor to the lengthy time involved were the voluminous documents spelling out those requirements — in some cases, more than 400 pages.
Now, they are as few as 10 pages. This shrinks review times exponentially and allows senior Defense officials to debate top-level requirements and costs associated with them, Stapp said.
"You found out in those 300-, 400-page documents, almost nobody read them," he said. "So what you were really debating in a lot of those [JROC meetings] were really those top issues anyway."
For example, top-level requirements for a plane would include how far it can fly, the types of missions it can perform, when it will be battle ready and how much it will cost.
In another change, program managers now have control over small requirements, which will not be locked "into concrete" like they were in the lengthy documents, Stapp said.
"The program manager should be able to debate some of the lower-level stuff so those high-level things are actually delivered on cost and on schedule," Stapp said.
Only major stalemates during the document review process will be elevated to the JROC level. Yet while more of the requirement activity will take place outside the JROC, tripwires have been put in place to flag trouble spots.
If a program's cost jumps 10 percent, its schedule slips six months, or if there is a 10 percent change in quantity, it will be reviewed by the JROC.
"We are also inserting ourselves further on into the acquisition process to make sure that the requirements ... have not crept beyond what they should be, thereby increasing costs and adding bells and whistles, but also looking for ways where we can fine-tune and adjust the requirements to help with efficiency and just bring reality back into play," Winnefeld said.
While the number of documents is shrinking, so is the number of people attending JROC meetings.
Instead of hundreds of people attending — including lower-level action officers and even industry representatives — now only a select group of senior officials attend. They include the heads of the Pentagon's policy, acquisition, comptroller, cost-estimating and operation-training divisions.
"It really just was not conducive to good discussions and good debate, especially if there were controversial topics, because you just couldn't do it with a bunch of people because it would leak like a sieve throughout the entire building," Stapp said.
The new process also looks at a portfolio of systems versus a particular program. For example, the JROC will examine how an intelligence system fits within the larger context of this type of equipment.
"Very often in the past, an individual issue would come to the JROC and every program looks really good on its own merits, with a beautiful PowerPoint brief," Winnefeld said. "When you start to open up and throw that particular system or idea or requirement in a portfolio, then it really allows you to make more informed decisions that may actually make us more efficient."
Approving requirements faster
The new, streamlined requirements process is already showing signs of success. The JROC recently approved the documentation for a large Air Force program in 21 days.
"We can't do that with everything," said Stapp, who declined to identify the program. "This one was very important and we wanted to get it through."
Pentagon officials expect a normal program to go through the JROC process in 30 to 60 days, significantly shorter than the two years it has taken historically.
"What we wanted to be able to do is provide enough flexibility in the system that if something was critically important, we could expedite it, but those that were just normal processes, we also wanted them on a much faster time line," Stapp said.
The system could also accommodate an "urgent need" request from a battlefield commander in two to three weeks.
Stapp himself could approve these requirements. Should he find an issue, he could elevate it to Winnefeld, who could then discuss it with the relevant four-star combatant commander.
The revisions "make permanent several important initiatives that enable more rapidly delivered and affordable capabilities" to troops on the battlefield, acting Pentagon acquisition executive Frank Kendall wrote in his answers to policy questions from the Senate Armed Services Committee, which is reviewing his nomination to be undersecretary of Defense for acquisition, technology and logistics.
The changes also improve the "agility and efficiency in meeting the most urgent war-fighter needs in current and future contingency operations," he wrote.
In the case of cybersecurity and information technology requirements — where systems could be needed in days or hours — decisions could be made below the JROC level.
"You want to delegate those down so that decisions can be made in a much faster time line," Stapp said. "In a lot of those cyber requirements, you're going to see those delegated down to the" one-star general level.
More efficient acquisition
The changes to the requirements process should make the actual purchasing of systems more efficient.
"We kicked sometimes impossible problem sets over to the acquisition community, and they basically couldn't deliver because the cost and schedule were not appropriately debated and looked at," Stapp said.
Kendall and Christine Fox, director of DoD's Cost Assessment and Program Evaluation (CAPE) office, have been attending JROC meetings to provide input.
A Pentagon study has found about 90 percent of Pentagon programs were over cost and delayed. Often, the services submitted overly optimistic schedule and cost estimates, while the CAPE estimate ended up being more realistic.
"If you really looked at what CAPE had estimated over the same time frames, CAPE was right about 90 percent of the time," Stapp said.
Now, the JROC will review both estimates, but the CAPE estimate will be used, "especially if there are large deviations in cost and schedule," he said.
If realistic cost and schedule estimates are married to the requirements at the onset of a program, acquisition officials should be able to execute programs more effectively, Stapp said.
"If you deliver over more realistic schedules with realistic costs, what you find out is the acquisition community ideally should be able to deliver … more of a 90 percent on cost and schedule performance," he said.