DIU Increasing Innovation Impact
Rapid technology development is rapidly becoming the norm.
Officials leading the Defense Innovation Unit, the Defense Department’s one-of-a-kind rapid prototyping organization, intend to increase the unit’s influence, largely by focusing on technologies with the broadest applicability.
The organization, now known as DIU, was founded in 2015 to help the department better keep up with emerging commercial technologies. Focus areas include artificial intelligence (AI), autonomy, human systems, information technology and space. The unit’s small staff of about 75 includes active duty and civilian personnel at its offices in Mountain View, California; Austin, Texas; Boston and the Pentagon. Its modus operandi is to work with both traditional contractors and smaller businesses and entrepreneurs without a lot of experience working in the national defense arena.
The DIU team has largely normalized the use of a procurement tool known as other transaction authority, which was once a niche process used only by NASA and the Defense Advanced Research Projects Agency. The process essentially eliminates much of the red tape and development time associated with more traditional procurement processes. The team also extensively uses agile software development operations, commonly referred to as DevOps, which allows greater innovation and rapid development.
“The benefit we bring is that we’re able to get commercial companies on contract sooner than a traditional acquisition process. Our goal is within 60 days to get a prototype contract and then to get that prototype done successfully so that we can rapidly field new capabilities,” explains Michael Brown, DIU managing director. “We’re getting access to leading-edge capability, getting it there sooner and saving the department money in the process.”
Arguably, the DIU already has significantly influenced the Defense Department. It has awarded roughly 100 contracts in the few years since its founding and introduced many new suppliers in the process. Furthermore, the use of other transaction processes and DevOps are becoming more widespread. The Air Force, for example, has developed its own DevOps lab and intends to use lessons learned from the DIU for developing and deploying a wide range of technologies across the service.
That doesn’t mean, however, that the DIU leadership will be resting on their laurels. “As far as I know, we are the only [organization] going across all of the department that is focused exclusively on commercial technology. We want to have more impact on that mission,” Brown says.
To expand DIU’s influence, Brown and his team are looking for technologies with broad applicability across the department. “It’s not just about the quantity of what we’re doing. It’s about the quality. One project that has the potential to be across all the services and save $10 billion or more can be as important as expanding from 100 contracts to 150,” Brown says. “We’re really focused on the impact we can have as opposed to the quantity of contracts or the amount of money awarded.”
He cites the Predictive Maintenance project as one example. The DIU is working on the project with C3, a Redwood, California, company specializing in AI. The company says on its website that its platform enables the department to aggregate and keep current enormous volumes of disparate data, including sensor reports and maintenance logs, in a unified cloud-based data image running on the GovCloud solution from Amazon Web Services.
The project initially focused on the F-16 fighter jet and the E-3 Sentry Airborne Warning and Control System. It uses AI to reduce maintenance costs and can be expanded for use on vehicles of all kinds, or “anything that moves,” Brown says. “That’s a huge opportunity, saving potentially billions of dollars as you apply what’s already in use in the commercial airline space to military aircraft,” he adds. “We’re now expanding it to other types of aircraft and the Bradley Fighting Vehicle. We’re working with all three branches—Army, Air Force and Navy.”
Brown also points to the Kessel Run project with the Air Force. Air Force personnel once organized aerial refueling missions by entering mission data on an Excel spreadsheet and visualizing that data by drawing lines with dry-erase markers and arranging magnets to organize each day’s mission, explains Airman Magazine, an official Air Force publication. Coders from DIU—which was then referred to as the Defense Innovation Unit Experimental —rapidly developed a program that allows one person to accomplish in three hours a task that once took a six-person team eight hours.
Project Kessel Run now has its own DevOps lab—the Kessel Run Experimental Laboratory. Under the direction of the Program Executive Officer Digital, it is expanding the use of DevOps across the Air Force. “We not only piloted a solution but an entire methodology, agile software development,” Brown asserts, pointing out that the accomplishment came before his tenure at DIU began last September.
One of the benefits of DevOps is that it is broadly applicable and transformative. “If you think about the way the commercial world works today in software, it’s really all about how to have more continuous capability that gets fielded. That’s everything from your Apple product to your Tesla car. The military really needs to be on a development system or methodology that mirrors that so that we are not trying to design something, specifying it to death, and then waiting so long to get what the military needs,” Brown elaborates.
The Drone/Counter-Drone Strategies project is another that the DIU team considers transformational. The DIU has partnered with Dedrone, a San Francisco company that combines sensor hardware and machine learning software to provide early warning for, classification of and mitigation against drone threats.
“The low-cost drone market, those that might cost $1,000 or less, is pretty much owned by China, and in particular one company, DJI Industrial Solutions. They have 80-90 percent of the world’s market for low-cost consumer-oriented drones,” Brown explains. “That puts us in a dilemma as the Defense Department.”
He adds that the Defense Department also wants to field low-cost drones but needs technology it can trust. “We’d like to have the capability to have our soldiers and airmen and sailors to be able to use drones. We don’t want to be buying those from China.”
Furthermore, those low-cost drones pose a threat to U.S. forces. “Because our adversaries, whether it’s ISIS, al-Qaida, whoever, have access to those low-cost drones they can buy over the Internet, we need to have a set of technologies that will help us to counter those,” he adds.
The partnership with Dedrone was announced in August. “We already have counter-drone strategies being deployed today to protect our troops in the field,” he adds.
In addition to focusing on technologies that can be deployed across the department, DIU officials say they can continue to expand and improve the acquisition tools already in use, including the commercial solutions opening (CSO), a rapid acquisition source selection approach. “We do believe the department is set up with the correct authorities in place in order to engage the commercial market, but what we’ve been focusing on at DIU is developing repeatable business models that can be emulated and scaled across the department,” explains Maj. David Rothzeid, USAF, DIU director of acquisition pathways. “The first instantiation of that was the commercial solutions opening, which is the primary mechanism by which DIU engages in solicitations that then result in prototype contracts.”
DIU officials have confronted a degree of resistance to change since the organization’s inception. “There’s still quite the cultural and policy barrier that’s inhibiting other organizations to be able to leverage the authorities to the maximum extent practicable. There’s definitely a bit of inertia, but it’s starting to crumble around the edges,” Rothzeid asserts. “We’ve been participating at the highest levels with those who would be writing the policies to help provide the top cover and the mechanism and the rules of the road by which other Defense Department organizations could leverage these authorities.”
Even with the use of other transactions, CSOs and DevOps, the Defense Department’s acquisition processes probably will never please everyone, Rothzeid indicates. “There are always going to be detractors stating that the acquisition system could be better. I don’t think they’re wrong, but we just have to be flexible in the way that we then approach and accomplish acquisition,” he states.
Still, Brown indicates that officials across the department and services accept that change is necessary. “Everybody in the Defense Department recognizes we’re in a technology race with China, and we need to move fast. Everybody’s eager to take advantage of whatever is available.”