The Defense Intelligence Agency will formally roll out its new Open Innovation Gateway, a major pillar in the agency’s push to move away from big, monolithic technology acquisitions and bring new innovations on board in small bites and in very short cycles.
Officials have not discussed many of the inner workings of the gateway prior to Wednesday’s official announcement, during which DIA will declare it has reached initial operating capability.
The agency has made clear for the past year that the intent is to give technology developers much more insight into the technical requirements that a new capability must meet before the agency will buy it.
That insight, DIA says, extends beyond publishing black and white technical standards. Via the gateway, the agency will give developers access to the actual computing environment DIA uses today — and eventually, the shared set of systems under the entire intelligence community technology infrastructure — so that they will know from the outset whether their technologies will integrate with DIA’s existing systems, and if not, what changes they will need to make.
Insight by Confluent: Learn about how agencies are benefitting from that concept of data-in-motion to improve mission outcomes in this exclusive e-book.
Dan Doney, the agency’s chief innovation officer, said the gateway is part and parcel of DIA’s broader innovation strategy, which aims to democratize the technology uptake process.
Speaking Tuesday at the agency’s second annual innovation symposium at its headquarters in Washington, he said DIA wants to buy whatever disruptive technology makes sense at any given time for whatever base of users needs it, in small, rapid cycles, whether the creator is a garage startup or a government contracting behemoth.
Proprietary solutions not welcome
While DIA’s strategy rests largely on the idea that the government must publicize its own technical standards, that decision should not be interpreted as a preference for open source solutions from vendors, Doney said.
“We will not kowtow to any proprietary solution. We cannot do that. However, when we’re driven by standards that allow us to replace what we have with something that’s better, if proprietary solutions meet our standards, then it’s game on,” he said. “I want the best thing I can get my hands on, and that requires us to clearly articulate our standards in a way that they’re not subjectively enforced. People have to know, ‘I’ve gotta get here if I want to do business with DIA.’ That levels the playing field, whether it’s proprietary or otherwise. We’ve also built in the ability for vendors to showcase their capabilities, while also protecting their intellectual property, so it’s an even playing field for folks who are building in an open model, as well as companies who have invested heavily in producing a capability and want a return on their investment.”
DIA has taken longer than it hoped to get the gateway off the ground. It was originally scheduled for a January launch. But a closely-related agile acquisition effort, Needipedia, has been up and running since late last year, and last Friday, the agency made its first contract award under that program.
Officials say they designed Needipedia to stay true to both the letter and spirit of acquisition law, but to move the process along much more quickly. To do that, DIA published a single broad area announcement that contains most of the boilerplate language the Federal Acquisition Regulation requires. But after checking those regulatory boxes, the BAA serves mostly as a pointer to a regularly-updated repository of capabilities DIA mission users would like to have, but do not as of now.
“Making one contract award may not seem like a game-changer, but it took tremendous effort from our contracting team, our legal team and our senior leadership to make it happen,” Doney said. “It’s broken the dam, and this model is different from the normal process in important ways. Now that we’ve learned the lessons, it’s going to let us act more efficiently on the ones that follow behind.”
Goals, promises unmet
The Needipedia process invites industry, inventors and academia to submit white papers if they think they can have a capability that can meet DIA’s wish list of new technologies or business processes.
During the program’s first eight months, it garnered 240 submissions, including 133 from offerors who don’t usually do business with the government. As of now, DIA is pushing 50 of those ideas into further review and potential acquisition.
The agency also created a classified version of Needipedia in May, so as not to disclose its objectives to actual or potential adversaries, and plans to launch a “sensitive but unclassified” version of the technology wish list this fall.
Still, the process is taking longer than Doney originally imagined when DIA first announced its agile acquisition ambitions a year ago. The intent at that point was to respond to those white papers within six weeks.
“We have not met all of our goals, we have not met all of our promises, and for that I’m directly responsible. We will not let this continue. We’re working to address this, we’re working to automate the process more than we have until now, and we’re delivering focused attention to make sure this doesn’t continue. We will not let any idea slip through the cracks. You will all receive good feedback in an appropriate amount of time,” he told the industry audience Tuesday.
Doney said there were several factors that stalled the agency’s plan to upset the traditional government acquisition process. Among them was the budgetary turmoil of the past year. Even though officials knew from the outset that finding a new way to do technology acquisition would be hard, it’s been more difficult than they expected.
Want to stay up to date with the latest federal news and information from all your devices? Download the revamped Federal News Network app
Balance speed, transparency, accessibility
DIA, Doney said, has struggled to craft its innovation program so that it wasn’t just fast, but also aligned with principles of competition, oversight and transparency that usually lead to very prolonged acquisition programs.
“People often tell me, ‘I’ve already got agile acquisition models, I can put things on contract very quickly.’ But they only deal with a small set of idea providers, and that’s unacceptable,” he said. “We have to reach out for the broadest set of ideas, and getting those from small businesses, nontraditional providers, coalition providers, we have to be able to bring them in. That has to be backed with efficiency, and many efforts to streamline acquisition have ended up in systems that become a lot like slush funds. We have to balance speed with transparency and efficiency and accessibility and auditability. We think we’ve made no sacrifices in this model, but there’s still a lot of room for improvement.”
Lt. Gen. Michael Flynn, DIA’s director, said rapid innovation is a priority, not just because DoD no longer has the financial resources it did a few years ago, but because he worries the department has lost its ability to innovate as quickly as it must in order to meet its responsibilities.
“Our defense programs and our intelligence community programs operate at such a turtle’s pace that we can’t stay ahead of the adjustments we need to make,” he said. “As I look at what we’re potentially about to do in Iraq, the kinds of tools and technologies and people we’re going to apply there aren’t going to be what we have on the shelf, because it’s not very agile. It does not provide the flexibility we need.”