30 juillet 2018 |
By: Mike Gruss
The long-held thinking in Washington is that if the Department of Defense wants to stay ahead of its adversaries, it will need improved capabilities, many of which are being developed outside of the Beltway.
To that end, Pentagon leaders launched the Defense Innovation Unit Experimental, commonly referred to in defense circles as DIUx, as a way to attract new companies into the fold and accelerate the pace of acquisition. In fall 2017, Secretary of Defense Jim Mattis visited the organization's headquarters and said its influence would grow. A few months later, DIUx's director, Raj Shah, left the organization for the private sector. Capt. Sean Heritage has served as the acting managing partner since then.
Heritage spoke recently with C4ISRNET's Mike Gruss.
C4ISRNET: What's the status of the DIUx reboot?
CAPT. SEAN HERITAGE: The last two years have been largely focused on fielding capability across the department. We like to say we can solve problems faster, better, cheaper using commercial technologies with nontraditional companies that are out there.
We've learned that we have some amazing talents on our team and rather than just defaulting to fielding capability, we are creating tailored solutions using organic expertise. We've been able to do this primarily in a suborganization we refer to as the Rogue Squadron — it works in the UAF, counter-unmanned aerial system world.
We're sharing our lessons learned across the department. We're coaching people on how we do things so that they can emulate us. So, we'd like to think that if we are the only ones doing CSOs [commercial solutions openings] and OTAs [other transaction authorities], in our image a year or two from now, then we failed.
C4ISRNET: How do you know this is working?
HERITAGE: We know all our prototypes aren't going to transfer to production. Our target goal is between 40 to 60 percent success rate. The feeling is that if we are more successful than that, then we probably aren't being innovative or creative enough and assuming enough risk on behalf of the department. With regard to our creating line of effort, we're very proud of the [Air Operations Center] Pathfinder project. Kessel Run was a great example of fielding capability and then leveraging organic talent to create tailored solutions and transition to a program of record.
Then the coaching effort, we're defining success by creating somebody in other organizations in our image every six months.
C4ISRNET: On talent development, is part of that asking leaders to rethink how they're using their people or is it something else?
HERITAGE: The best example would go back to Kessel Run. That really started with the Defense Innovation Board. We were able to help the Air Force explore the art of the possible, which made a compelling case for the DIB's original recommendation: you guys need software developers as a core skill at the unit level to solve these problems in real time.
They like to use the example that Home Depot has 6,000 software developers to sell you hammers. You can say the same for any service. How many software developers do you have? Well, the answer is we don't know, because we don't have a specialty for software developers, but no doubt once you know people across your team, there are many very capable people who do software development on the side. How do you leverage that? How do you more deliberately recruit to that? And how do you develop them?
C4ISRNET: What's one area where you guys have been able to say, “Without us, we're not sure this would've happened as fast?”
HERITAGE: We are very proud of the work we are doing in the artificial intelligence world. So, we are doing some projects on predictive maintenance. Again, for the Air Force, focusing on the [E-3 Sentry airborne warning and control system] out of the gate. But it's scaling across other platforms. We're already able to deliver value by solving that problem, reducing unplanned maintenance.
Another contract was let out that's scaling to the Army with their Bradley Fighting Vehicles. We are going to be doing predictive maintenance with them. Something that may be lost on many people is our involvement in the formulation of the JAIC, the Joint AI Center. That would not have happened without us, but, more specifically, that would not have happened if we weren't here in the Valley and had access to create relationships with some significant talent in the civilian world.
C4ISRNET: I wanted to reframe what you were saying. If DIUx wasn't in existence, do you think that there wouldn't be the same level of talent?
HERITAGE: The talent we are able to attract to our team, to work through any problems and inform strategic thinking, wouldn't have happened. What we have is very open-minded senior leaders who are yearning for diverse thinking. And, as a career military guy, it's not all that diverse. What we're able to do is reach into a talent pool that has lots of credibility and expertise that isn't resident in the department.
C4ISRNET: What's something DIUx is doing now, that it wouldn't have done three years ago?
HERITAGE: The biggest difference is the deliberate embracing of our role in creating solutions and scaling ... coaching. Those were things that DIUx didn't do in the past. We have much more support from across the department. We no longer need that direct cover of the Dep. Sec. Def as a direct report.
The percentage of our team that is military is higher than at the beginning. That's not necessarily by design. We are attempting to recruit more individuals, commercial executives, to help coach us. We don't want to be a military organization that happens to be in Silicon Valley, Boston and Austin. We want to truly be representatives of the culture.
C4ISRNET: You talked about that 40 to 60 percent goal. What are the other ways you measure success?
HERITAGE: The first one is the number of customers within the department that are coming our way to ask us to help them solve problems using our relationships, authorities and methodologies. That's one metric and that number continues to go up.
The other is the number of nontraditional companies who want to help solve problems. A commercial solutions opening is a honey-pot for all these creative companies to say, ‘Can I contribute?' The number of companies continues to go up. It was 40-plus on the last CSO. The record to date is over 70. Now we're in a position where we have to say no to projects. We're a little more selective.
C4ISRNET: What is DIUx's role with AI?
HERITAGE: A lot of folks claim to be able to contribute to the AI mission area. There are 593 initiatives across the department that claim to be AI. It's our ability and our connections with Project Maven that have afforded us the opportunity to influence the way ahead for the department and facilitate some conversions. Some of the work that we're able to do with these nontraditional companies is not only shaping understanding of the power of AI across the department, but it's focusing everybody as well. We're able to help make them a bigger part of the conversation or maybe inspire them to a bigger part of the conversation.
C4ISRNET: How do you convince a company that working with the Pentagon is the right path, and it's not a quagmire like it has been at Google?
HERITAGE: We don't spend time trying to convince companies to do or think anything. There are plenty of folks out here and elsewhere, who are passionate about contributing to the cause, and appreciate what they are learning just through having access to the data that we're able to provide them. But it's a challenging conversation to be a part of because we don't want to turn people off. We're here to provide people with opportunity.
C4ISRNET: Where can you help the most in IT?
HERITAGE: [Information technology] is an example of how a DoD organization can leverage technology and access to a network that is still secure, yet different than what the rest of the department is using. It's things that you probably take for granted. We have people come to our team for a short period of time, after a career within the department going, “Wow, look at how my productivity has changed since I've been here.” I have access to Slack and the G Suite and a whole host of other tools that are not allowed, and for good reason, on DoD networks.
C4ISRNET: One concern is whether new vendors can provide overmatch. Is that the case?
HERITAGE: We look at the companies to make sure they're going to be around long enough to provide this capability. When we talk about our responsibility to help navigate the “Valley of Death,” and make sure that these prototypes don't just die, these companies have to have legs underneath them and investors behind them to be viable down the road so that we don't spend our time developing a prototype that they won't be around to deliver.