For the second year in a row, the chairman of the House Armed Services Committee released a sweeping set of proposals to reform the Defense Department’s acquisition system on Tuesday, saying he believes Congress should focus its attention toward making the existing system more agile rather than continuing to create special rapid acquisition lanes that ignore its underlying problems.
Rep. Mac Thornberry’s (R-Texas) discussion draft, released a month before the committee is set to begin serious work on the annual Defense authorization bill for 2017, would mandate the use of open architectures for military weapons systems and continues a trend begun last year to devolve more acquisition authority to the military services.
The proposal would require that all major Defense programs have a strategy to implement open architectures when they reach “Milestone A,” the early phase when DoD decides to move a program into technology maturation.
The intent, Thornberry said, is to break big platforms (say, a ship’s hull or a fighter jet’s airframe) apart from subsystems whose potential capabilities are evolving much too quickly for the current procurement process.
Insight by GitLab: During this webinar executives from the State Department, U.S. Securities and Exchange Commission, U.S. Patent and Trademark Office and GitLab will discuss how institutionalizing a DevSecOps approach to software development is a journey that must bring together the technology and business sides to change an organization’s culture.
“We were down at Quantico a few weeks ago looking at the F-35, and one of the pilots told us that this is basically a big, flying computer,” he told the Brookings Institution Tuesday. “What we’re trying to do is differentiate components from platforms and make sure our systems are designed in a way that components can be upgraded whenever the technology allows. That’s going to allow us to field new capabilities in a much faster way. On average, it takes 9 years for a major defense program to be fielded.”
The Defense Department frequently and publicly proclaims its support for open architectures, but congressional staff members who have been working on the latest reform proposal said DoD’s actual implementation of the practice has been far too sporadic, and that it’s time to legislate the open architecture as the normal course of business for big systems.
The bill would require DoD to establish and publicly release technical standards for interfaces that would let components plug into its major platforms, and demand that all the components it buys adhere to those standards.
That route, Thornberry said, will encourage new companies to enter a Defense market that’s been effectively closed to all but the largest traditional government vendors and create new competition as small and medium-sized companies vie to deliver an individual subsystem on an aircraft carrier rather than an entire aircraft carrier.
“You don’t have to be the big defense company, and I’ll even go so far as to say it will reduce bid protests,” he said. “When you’ve got a chance to work on incremental upgrades instead of entire platforms, you’re not missing out on the opportunity to be a part of that platform for the next 10 or 20 years.”
Thornberry’s proposal also theorizes that separating platforms from their components will shorten the overall acquisition process for a major system and cut its costs. The bill includes provisions intended to encourage DoD to only integrate technologies onto its major weapons systems once they’re proved and stable.
Meanwhile, the bill would segregate unproved technologies into a separate funding process in which the military services would be urged to test new ideas to get those components to the point at which they’re ready to be added to a major platform.
“The more stable the technology, the less need there is for things like cost-plus contracts,” he said. “On the other side, we need to create funding for experimentation and prototyping. That’s where you’re supposed to fail, and if you don’t, you’re not pushing hard enough. The problem we’ve had is that our big programs and our technology development have been locked together: You can’t get funding for prototyping or experimentation without being tied to a big program of record, so we haven’t done much of it. So when technology experiments failed, they failed spectacularly. We need to be willing to fail early, and pulling experimentation apart from our major programs has some major advantages.”
The bill also tackles the thorny issue of intellectual property. It aims to simplify the debate over who owns which data rights by breaking IP into three distinct categories: If the “black box” a company has built to plug into a weapons system’s interface was built with its own funds, that firm owns the intellectual property. If it was built entirely with government funds, the government would own it. If government and industry shared the costs, the parties would have to enter into a negotiating process.
“We’re not going to solve all of the intellectual property issues with this year’s bill, but I’m very concerned that companies are walking away from the Defense business. I know that because they’ve told me they are,” Thornberry said. “It’s especially the case with technology companies where a relatively small fraction of their business comes from government sales, and that’s where a lot of our upgrades are going to come from. We can’t afford that. Hopefully, a little clarity will let us upgrade faster, and the purpose off all of this is to get technology into the hands of warfighters faster.”
Thornberry said he believes the acquisition process can also be sped up by devolving the decision making process about any given weapons system lower in the food chain.
Last year’s Defense bill gave the military services their own authority to move programs to their next major milestone. Before that, for the largest weapons systems, the authority resided with the undersecretary of Defense for acquisition, technology and logistics.
The new proposal involves a similar devolution of authority for joint programs. Beginning in 2019, for multi-service systems like the F-35 or the Joint Light Tactical Vehicle, one service would be assigned as the executive agent to manage the program after the existing Joint Requirements Oversight Council has approved its requirements and the Office of the Secretary of Defense (OSD) has assigned firm cost and schedule targets.
Thornberry said the change would add accountability to an acquisition process whose responsibilities are diffused through too many levels of bureaucracy.
“Right now, when we in Congress ask who’s to blame for a program that’s gone overbudget or delayed, we get a lot of finger pointing. ‘It’s the services, it’s OSD, it’s the testers.’ No one’s accountable,” he said. “When we look to the outside world, you’ve got corporate oversight of budgets and policy, but the division managers actually build the stuff. If you have a division manager who’s not pulling his weight, he’s replaced. That’s what we have to get to, rather than layering on a whole additional layer of oversight to compensate for the inadequacies of the underlying organization.”
Want to stay up to date with the latest federal news and information from all your devices? Download the revamped Federal News Network app
House Armed Services staffers argued the draft bill is in line with the original intent of the 1986 Goldwater-Nichols Act: A strong Defense secretary who can broadly oversee the military services’ acquisitions and veto their decisions when necessary without an affirmative signoff and separate paperwork shuffle for every decision a military service makes.
“We’re going to expect the services to build things, and if they’re unable to manage their programs, the secretary of Defense must be much more decisive in replacing whoever he needs to replace,” Thornberry said. “But the secretary will have cost caps and schedule targets on every program , and if something starts to go off-track, lights start blinking so that he can ask what’s going on.”
Another aspect of the Thornberry proposal would add more blinking lights at every stage of a program’s development should it begin to exceed its cost or schedule projections.
At each milestone of a program, DoD would have to produce a one-page scorecard comparing program management officials’ estimates of how well their programs are doing against the independent estimates produced by existing organizations such as the office of Cost Assessment and Program Evaluation.
“It’s kind of a dashboard,” Thornberry said. “There’s a lot of data out there about our programs, but it’s not in as usable a form as it needs to be. We’re not asking them to invent something new, but it’s going to be an important management tool.”