The students at Stanford University were tackling one of the Pentagon's most perplexing problems: How do you constantly monitor important regions of the world for emerging developments, such as preparations for a ballistic missile launch in North Korea?
The students thought a relatively new technology, synthetic aperture radar, could address the gaps in coverage left by large government spy satellites. Last year, they started talking to defense agencies about their idea to put hundreds of such radars in vehicles the size of breadboxes and launch them as a satellite constellation.
The idea was so promising it was picked up by the Defense Innovation Unit Experimental (DIUx), the Pentagon's Silicon Valley outpost, as one of its first projects. With DIUx as a customer and start-up funding provided by venture capital firms, the idea morphed into a full-fledged company called Capella Space. The Pentagon is now counting on Capella Space to launch the first constellation of synthetic aperture radar satellites in early 2018.
The students who started Capella Space didn't connect with Defense Department agencies by chance. They were participating in a four-credit university course called "Hacking for Defense," which connects graduate students with government agencies and their real-world national security problems. DIUx, meanwhile, provides a pathway into the DOD bureaucracy for small, start-up companies.
Hacking for Defense and DIUx are among several initiatives the Pentagon has launched in recent years to circumvent the military's famously slow bureaucracy and inject new technology, ideas and talent into DOD. While they started out small, the initiatives have begun to grow in both size and influence.
Hacking for Defense is now offered at seven universities in addition to Stanford. The House's fiscal year 2018 defense policy bill includes a provision allowing the defense secretary to spend up to $15 million per year on the program, potentially paving the way for more expansion.
DIUx was established in 2015 by then-Defense Secretary Ash Carter as a way to bring more Silicon Valley tech companies into DOD's industrial base. After a slow start, the office was re-organized, granted acquisition authority and expanded to locations in Austin, TX and Boston, MA. In a little over a year, DIUx has awarded $184 million for 59 pilot projects and recently transitioned its first project to a production contract worth a potential $750 million.
As initiatives like Hacking for Defense and DIUx grow, however, some wonder about the role of these increasingly institutionalized innovation insurgents.
"As an office grows in size and scale, there's a risk they become a part of the bureaucracy," said Paul Scharre, a former Pentagon official and senior fellow at the Center for a New American Security. "What happens to these things as they go forward?"
A new administration takes over
When President Donald Trump's transition team showed up at the Pentagon late last year, defense analysts questioned what would happen to Carter's innovation projects, in particular DIUx. The Silicon Valley office still faced skepticism from lawmakers and traditional defense contractors. Would the new administration see the program as superfluous and wasteful?
But Defense Secretary Jim Mattis, who spent his brief retirement from government service as a visiting fellow at Stanford, has embraced DIUx and Carter's broader innovation agenda.
"There is no doubt in my mind that DIUx will not only continue to exist, it will actually grow in its influence and its impact on the Department of Defense," Mattis told reporters during an August trip to DIUx headquarters.
The Strategic Capabilities Office, another Carter project started in 2012, has also been welcomed by Mattis' Pentagon. SCO is tasked with working with DOD services and agencies to find new, innovative ways of using existing technologies and weapon systems. For instance, SCO recently led a project to give a Navy anti-aircraft weapon the ability to target ships at long ranges.
In November 2016, two months before leaving his post as defense secretary, Carter created a permanent place for SCO within the Pentagon bureaucracy. The directive he signed calls for the U.S. military to "establish a new era of power projection."
The new administration has continued to back the secretive office, which has seen its budget grow dramatically since being established five years ago. In FY-14, SCO's budget was $125 million. Now, the Trump administration is seeking $1.1 billion for the office in FY-18, $200 million more than the Obama administration allocated for SCO last year.
Elsewhere in DOD, Carter's innovation projects continue to grow. The Defense Digital Service, a team of self-described tech "nerds" from the private sector who sign up for short-term "tours of duty," was established by Carter in 2015. Initially, DDS began a campaign called "Hack the Pentagon," which uses citizen hackers to find vulnerabilities in the military's public websites.
DDS' white-hat hacking program continues, but the organization has taken on larger, enterprise projects as well. In September, Deputy Defense Secretary Patrick Shanahan tasked DDS with developing a strategy to accelerate DOD's adoption of cloud computing technology, marking a potential overhaul of the department's sprawling IT infrastructure.
DIUx, SCO and the Defense Innovation Board are also involved in the cloud project. The latter was set up by Carter in 2016 as a new DOD advisory panel and has enlisted tech luminaries like Alphabet Executive Chairman Eric Schmidt and astrophysicist Neil DeGrasse Tyson. The board has continued advising the Pentagon under the Trump administration, and its recommendations on issues from software procurement to artificial intelligence have made their way into new DOD strategies and defense legislation.
Scharre said Carter deserves credit for creating various organizations and "shaping the bureaucracy in a way that I think may take a long time to be appreciated."
But the experimental organizations still face questions about their long-term place at the Pentagon. In an August interview with the Associated Press, House Armed Services Committee Chairman Mac Thornberry (R-TX) suggested DIUx might not be necessary.
"This question is: What is this office doing that’s different from what others are doing?" he said.
Other policymakers wonder how the new organizations fit in with more established DOD entities. In questions posed to the Pentagon's new acquisition chief, Ellen Lord, before her July confirmation hearing, lawmakers on the Senate Armed Services Committee asked whether there has been "appropriate interaction" between the Defense Advanced Research Projects Agency and SCO.
"Both DARPA and SCO are key elements of our DOD research ecosystem, and I believe their interaction is essential," Lord responded. "If confirmed, I will study the interaction between the two and ensure that it continues as appropriate and productive."
Acquisition restructure and lingering questions
DIUx and SCO were established outside the traditional DOD hierarchy -- both organizations have a direct line to the defense secretary. But a plan to restructure the Pentagon's acquisition structure may change that model.
Under the initial proposal put forward by the Pentagon in August, both DIUx and SCO would shift under the new under secretary of defense for research and engineering, reporting to the assistant secretary of defense for advanced capabilities.
The middle management of DOD is sometimes described as the "frozen middle," as it's often where turf battles and bureaucratic processes strangle innovative ideas, Scharre said. Both DIUx and SCO were designed to short-circuit the frozen middle, he said.
"As you start to see more layers inserted between them and senior leaders, how do you make sure the people in between are still incentivizing innovation and free thinking and trying new things, and they don't become risk averse?" Scharre said.
Mattis sought to quell concerns that the restructure would minimize DIUx during his trip to its headquarters in August, saying the organization will still "have direct access to me."
Despite its uncertain future within the bureaucracy, DIUx "is starting to identify the networks they need to use and the people that are best suited to work with them," according to Pete Newell, one of the co-founders of Hacking for Defense and a retired Army colonel who ran the service's Rapid Equipping Force.
DIUx partners with defense agencies and commands on every pilot project it initiates. Newell said he learned the partnership approach is key for organizations like DIUx and REF as they seek to develop and transition ideas to warfighters quickly.
"By the time I left [REF], I wasn't touching things that I couldn't build a consortium of folks around," he said.
Establishing those "organic" connections among different DOD organizations as well as private industry, academia and elsewhere is key to fostering a long-term "innovation ecosystem" at the department, according to Adam Jay Harrison, director of the MD5 National Security Technology Accelerator.
Established in 2015 within the Pentagon's manufacturing and industrial base policy shop, MD5 now runs the Hacking for Defense program for DOD and has started several more initiatives aimed at building a "bottom-up problem solving competence," Harrison said. The relatively unknown organization also has the backing of Congress, which appropriated $25 million for MD5 in FY-17.
Harrison points to Capella Space, and its hand-off from the Hacking for Defense class to DIUx, as an example of the partnerships MD5 wants to foster.
"We need more of that organic interaction," he said. "I would say the network building, the outreach, the opportunities to convene, the knowledge sharing, more of the tacit types of interactions, that's something we need to make a very explicit decision to invest in in the department."
However, what drives and connects DIUx, SCO, MD5 and other offices beyond a broad innovation banner is not clear.
The civilians who take leadership roles within the new under secretary for research and engineering office will play a key role in managing the organizations and establishing a culture for them to succeed, according to Scharre.
"When you create a culture within an organization that says, 'This is the direction we're going,' you also create a lot of opportunity for bottom-up growth," he said.
Harrison said he believes Pentagon leadership needs to more clearly define DOD's innovation goals.
"I think if you asked 10 people in DOD what innovation is, the answer you get nine times out of 10 is, technology -- that's just flat-ass wrong," Harrison said. "It's not technology. Technology is an outcome of the innovation process."
Stephen Rodriguez, a senior fellow with New America's International Security Program and founder of an independent defense tech accelerator called One Defense, agreed that "too often we get focused on the technology aspects of these organizations." Instead, DIUx, SCO and others should be organized around the common mission of countering Russia's and China's emerging asymmetric advantages over U.S. forces, according to Rodriguez.
"We're not just doing all this because we want to be cool or cute or fancy or we all want to hang out at Andreessen Horowitz conferences together," Rodriguez said, referencing the Silicon Valley venture capital firm. "We're doing it because we have a critical mission set that we need to orient these organizations around."