Embracing Risk in Defense Tech with Lt. Gen. (ret.) Michael Nagata

Bonnie Evangelista [00:00:18]:
All right. I'm Bonnie, with the chief, Digital and AI. Office, joined by Captain Sean Cooper with the 18th everyone corps, right?

Shawn Cooper [00:00:26]:
Yeah. How's it going?

Bonnie Evangelista [00:00:27]:
Good. Not too bad. Thanks for joining me. What group in the 18th everyone Corps are you with?

Shawn Cooper [00:00:32]:
Yeah. So we have a new cell on our core staff called the data activities, and there's a wide portfolio that we do everything from soldier software development, have a couple of applications that we're working on, and then we have an innovation portfolio as well, which I'm in charge of the ground up innovation portfolio. We have a cloud and data portfolio as well.

Bonnie Evangelista [00:00:53]:
Is this within the data warfare company.

Shawn Cooper [00:00:56]:
Or separate we don't call it the data warfare company anymore.

Bonnie Evangelista [00:00:59]:
I'm already OBE on the terms. Okay.

Shawn Cooper [00:01:03]:
So initially, it was a separate company. What we're doing is the entire intent behind the company, and the reason it was stood up was to upskill the core staff so that we can operate in a data AI. Enabled all those buzzwords environment. And what we've found is that having a separate organization doesn't necessarily accomplish that. So we're figuring out how to take those skill sets and integrate them into the core staff to really meet that end state.

Bonnie Evangelista [00:01:31]:
So somebody described you in a very unique and particular way, and I kind of want to get your reaction to it. They said, Sean is the apex nerd.

Shawn Cooper [00:01:41]:
The apex nerd? Oh, man. No, that could be a good way of doing it. It depends on what you want to qualify Apex as. Like, the one who herds them all or the one who well, we were.

Bonnie Evangelista [00:01:56]:
Just talking about neural networks and magic the gathering.

Shawn Cooper [00:02:01]:
Yeah, no, that definitely qualifies. I'll own that one.

Bonnie Evangelista [00:02:05]:
Is that a real thing?

Shawn Cooper [00:02:06]:
Yeah. So I was kind of bored, so I did a master's degree and applied in computational mathematics. And one of my projects in that, I wrote a neural network from scratch to pick cards in the game magic the gathering and mimic human behavior.

Bonnie Evangelista [00:02:23]:
So do you get to play? Could you play with it? Is that what you're saying?

Shawn Cooper [00:02:27]:
Yeah. I didn't work too much on the user interface or anything, but essentially you could show it the cards that you had, and it would tell you which one was the best one to pick for a certain situation.

Bonnie Evangelista [00:02:36]:
Did you actually win using tool?

Shawn Cooper [00:02:40]:
I haven't played, like, against other people with it, but at the time that I did it, this was probably in late 21, early 22, the best machine learning algorithms out on the net were getting about a 40% to 50% accuracy rate compared to human picks. And then mine with a somewhat limited data set, was able to get up to 60% accuracy. Okay, still a little ways to go, but it was cool to see with just a little bit of data, what you could do.

Bonnie Evangelista [00:03:12]:
So how are you translating some I would call that a little innovation on your part. How are you translating some of that to the 18th Airbran Corps?

Shawn Cooper [00:03:19]:
Yeah. So the reason I got picked up for the whole innovation cell in the first place was, I think, largely due to my background in that and what I found. And what has been really exciting about this position has been like, I might be called the apex nerd, but there's a lot of really good nerds out there in the operational force. And a lot of times we don't give them the tools or the publicity or the resources to succeed. So having done a little bit of myself, a little bit of it myself, I think that I'm well placed to identify those people and empower them. Just some quick examples. So, as part of the innovation portfolio, we run the Dragons layer competition, which is kind of a Shark Tank style innovation competition that we do once or twice a year. We had a soldier from the 25th Infantry Division.

Shawn Cooper [00:04:12]:
He was a pilot, and he wanted to get flight feedback from his training flights, go back and look at it in how he did. So he asked the army, hey, can I plug in and get the data off of the flight computers? They're like, absolutely not. No way. So he's like, okay. And he went and programmed a computer vision algorithm that they could put in the aircraft, and it would look at the flight instruments and just take the data from the flight instruments, and he could take that and go back and recreate its flight himself. So all on his own to make.

Bonnie Evangelista [00:04:46]:
Him a better pilot.

Shawn Cooper [00:04:47]:
Right, to make him a better pilot. And he was one of the winners for our Dragonslayer competition. And that's just one of dozens of examples from artificial intelligence work up through materials and policy recommendations that we get from across the core.

Bonnie Evangelista [00:05:02]:
So I talked to Bill Wilder last week about dragons layer and understanding the mechanics of how it works, because bottoms up. Innovation is hard, I think, in the Department of Defense in particular, because there's so many layers and having that connective tissue that if you have the pilot, like you're describing in your scenario, the pilot on the ground, he has an idea. How do you get it up to the other divisions of labor across the department that can maybe give him the resources and tools that you're talking about? So how either in that scenario or how is your group working to build that connective tissue?

Shawn Cooper [00:05:40]:
So when we first started off the program, it really was looking at how can soldiers solve problems at their level, whether that's through additive manufacturing, software programming, even things like tailoring to fix their bags and their backpacks and so on and so forth. And we quickly realized that there weren't that many mechanisms out there to scale that. Luckily, there's a lot of good examples across the DoD, like Afworks has their spark cells as well. And we did a lot of learning from the Air Force very early on on how to pair ground up innovation with top down modernization is how we'll kind of make that distinction. So over the last few years, we've really matured that process. And what we do is initially the competitions where you would have an idea, you'd go pitch your idea, and then we'd try to figure out how to scale it or how to get it integrated with our development partners across DoD or academia. And now we've sort of flipped that on its head, and we identify semi finalists months in advance, and then we start reaching out to our partners at the program executive offices, at the capability development centers, so on and so forth. We're like, hey, is this in your portfolio? Is this something you can provide feedback on? And especially for the development centers, is this something you'd be willing to work on? Prototyping and Maturing, the prototype that this soldier has, a lot of that's been done through a very robust agreement with Army Research Labs, which is called the Pathfinder Program, and they provide almost all the prototyping materials for the soldiers at the maker spaces that we run.

Shawn Cooper [00:07:16]:
Once they've gone through that three month long process of development and working with their partner, when they come into pitch, they should be able to look at the three stars and four stars that we bring to the competition and say, this is the problem we try to solve. Here's what we've done, and these are the next steps. I'm already working with the Armament Center, or I'm already working with this program executive office. All we need is a letter of support from you and $10,000 in three months.

Bonnie Evangelista [00:07:47]:
What's your favorite success story? Maybe doing what you just described.

Shawn Cooper [00:07:52]:
So very recently, for Dragonslayer six or Seven, we had an Airman actually submit an idea. It was for a cooling vest that would go under your plate carrier, but it would also integrate with your water system as well, so it would keep your body cool and your water source. Because he won Dragonslayer, the Air Force helped him with an entrepreneurship off ramp program for the idea. And the company that he was able to tech transfer his IP to was awarded a Siber direct to phase two siber that was funded in part by the Air Force and in part by Army Research Labs. And then they additionally got matching funding from their state that they're in, which is the state of Kentucky, for a grand total of $300,000 to continue to prototype the idea. So that was one of the biggest monetary wins that we've seen as well.

Bonnie Evangelista [00:08:45]:
That's a great transition story, right, with.

Shawn Cooper [00:08:50]:
The tech transfer, and they're actually out there doing testing in CENTCOM right now.

Bonnie Evangelista [00:08:56]:
Wow. What would you like to see from people like me? So I'm in the acquisition community. I'm also at the OSD level. What do you want to see from all these other groups or divisions of labor across the department to better support what you're doing?

Shawn Cooper [00:09:14]:
So the biggest thing that we need from the army side is those more robust pipelines for taking ground up feedback. If there's a company out there that's selling something, there's a million different ways that you could contract with them today to get their piece of equipment, whether it's in a prototype fashion or anything like that. And a lot of times that soldiers come up with something, it gets looked at as just tinkering. But what it actually is, it's a very real demand signal of what the end user actually needs, because most of these soldiers are very passionate about their training, and the people who are out there innovating are the ones who really want to be the best at their job. So when they're bringing a problem up to a higher Echelon to say, hey, this is really something I'm trying to solve, that's usually a very good indicator that that's an unaddressed gap or need by that unit.

Bonnie Evangelista [00:10:12]:
That makes a lot of sense.

Shawn Cooper [00:10:13]:
There's a million different ways to support, whether it's just through training, whether it's through actually adopting something, depending on where you're at at OSD or in acquisitions or anything like that. But the most important thing is just recognizing that that ground up feedback represents a very real gap and pain point.

Bonnie Evangelista [00:10:33]:
Yeah. So from your perspective, where my mind goes is where is that pain point? How is that being embedded or incorporated into the requirement? So our traditional acquisition system is requirements driven because of what you're saying here. I'm wondering if there's a disconnect. So from your perspective, what are you seeing? What does that process or conduit look like to get those pain points to an actual requirement or we lack one.

Shawn Cooper [00:11:08]:
So in the army, we have the capabilities Development Integration directorates ceded, and they're all based off of the various functional areas that we have engineering, maneuver, maneuver support. And that's where our requirements in the operational force really get solidified. And we've really started to make great inroads with relationships with those seededs. In fact, we had a briefing to all of the seated directors in July. So earlier this month or last month a great success story there is between demonstrations that soldiers from 18th airborne corps were able to give with first person view drones as well as soldiers in special operations. The maneuver seated was able to get an appendix to a joint capabilities development document to incorporate FPV drones which had previously not been in there. And that whole process from when those soldiers got to the requirements writer to when that got put in was like two or three months, which is pretty light speed for that ground up requirement getting integrated. One of the big things that we want our program to do though, is also support continual development while that requirement is in either a draft phase or getting reviewed or anything like that.

Shawn Cooper [00:12:26]:
So we've got that first ever valley of death from when a requirement is handed over to a writer to when it actually gets approved or anything like that. And a lot can change, especially with the barrier to innovation being as low as it is today. They're out there in Ukraine printing thousands of drones out of a single maker space and a lot of times we're know how can we incorporate one or two into our and that change has know over the past few months.

Bonnie Evangelista [00:12:59]:
What I would like to see a little bit more of. I can totally understand appreciate what you just laid out in terms of how we traditionally define, identify, and then start to implement or act on our requirements. But I think there's a lot of work we can do before we even get to the requirement. Because sometimes something I talk about a lot is we don't really know what we don't know and sometimes we just need to experiment or play. So why not let the people doing their job to your point, people who want to do it the best they can be, let them play, and that's going to inform requirements later. So do you see a lot of activity or outside of dragon layer but that support from either the PMS or the Peos giving you resources to experiment and play.

Shawn Cooper [00:13:56]:
So from the PMS and the Peos, a lot of times they're focused on the delivery of the capabilities that they've been directed to deliver. But what we have found is through our integrations with the seeds back in that early requirements phase, we've been doing what essentially amounts to continuous experimentation to develop these requirements. And okay, I'll stick to the FPV example or just the small UAS robotics example. In general, in fiscal year 21, I think the Army Futures Command had like 113 total soldier touch points for all of their programs. Our small UAS developers and experimenters in the core have probably had close to that many just in the last six to eight months with the maneuver seated itself. So there's constantly TDYs, whether it's the developers coming out. We had personnel that are working on the new regulations for small UAS fly out and sit down and talk to all of these end users and soldiers who've been experimenting in our maker spaces and out on our ranges, take feedback, write it down and actually incorporate that into new regulation. And simultaneously we've been flying soldiers out to what's called the maneuver Battle lab.

Shawn Cooper [00:15:22]:
That the maneuver seated. And out there they're able to take these robotics out break them without having to worry about being charged as a soldier for breaking something like that. See how they work in a tactical setting and the frequency of that, the increased frequency of those interactions is what's really enabled by these tactical innovation cells out there is they're tied into the unit's training plan so they can be like, is this going to affect the unit? They're tied into what the CD is doing and they understand or the Peos or whoever what their schedule is. They can very easily make these small trips successful.

Bonnie Evangelista [00:16:02]:
I'm curious because that sounds awesome, but also sounds like a lot of work on the soldier's part, right? So from your seat, how are people balancing? Soldier has to do a job, but also we need soldiers to give us that feedback that you're talking about and to create this space for more goodness to come out of it, like in this experimentation environment.

Shawn Cooper [00:16:23]:
So I would say my job was made a little bit easier by an Inspector General report from Army Forces Command, which found out that at the platoon level and higher, everyone's really busy. Their time is soaked up for the entire day filling out spreadsheets or whatever we do in the army. And then they looked at the junior enlist and they found that they spend about 4 hours a day just waiting to be told what to do. And that energy. Some of them want to go work out, some of them want to spend time with their family, some of them want to go invent something, make their lives easier and be a nerd sign up to be the next apex nerd. And that's the energy that we really want to harness there. That's current, as we see, currently untapped. So by having innovation managers at these tactical units, they can identify that.

Shawn Cooper [00:17:23]:
They can make sure that it does not have a negative effect on the unit and their training plans while getting the most out of the energy that those soldiers have.

Bonnie Evangelista [00:17:32]:
What are you most excited about right now about your role and what you're doing?

Shawn Cooper [00:17:39]:
So one of the biggest things that's coming up is we're working on creating an innovation officer skill Identifier, or for NCOs, an additional skill Identifier. What that's going to allow us to do is most of these very bright and motivated innovation officers, whether they're at our maker spaces or they're at the division level or the brigade level, everything that they've learned has been very self taught. And to see a lieutenant stand up and explain what PEO is working on, what and what PM is there, and be able to immediately tie a soldier problem statement to that correct portfolio and also walk them through the whole process not only of navigating the army's or the DoD's innovation or modernization ecosystem, but also understand the theoretical basis of innovation. Whether that's design, thinking, methodology or value canvas or any of these things that they'll teach in MBA programs is amazing. To see them walk through this whole thing. But again, that's all been self taught. So we're trying to work with 75th Innovation Command, defense Innovation Unit, defense Acquisition University, as well as other organizations across the army to create a formal program to capture all that knowledge that they've been self teaching.

Bonnie Evangelista [00:19:01]:
Do you have any airing of grievances? I'm not looking to knock anybody, but what could we do better from your vantage point?

Shawn Cooper [00:19:12]:
So the army and I guess the DoD as a whole is very concerned about Manning for very real reasons right now. And because of that, it's really easy to squash innovation efforts and say, we don't need you to be tinkering, we need you to be training, we need you to be in a ready or anything else like that. But what we'll find, and I'm sure is about 80% of your operational effectiveness comes from probably 20% of your people out there. And that's the case for almost anything, any organization. Yeah, if you allow that 20% or that 10% or that 1% to really reach their full capability, you're going to get outsized returns. And that's the basis of any investment is that maybe I don't get you to be on gate guard this weekend, but maybe I get quadrupeds with electronic warfare sensors on them that can automatically communicate to a small UAS, which talks to your end user TAC device because a soldier took the time to make those pieces of equipment communicate. And then that saves 510 20 lives down the line. So innovation is an investment, and it does require leaders to say, I am willing to give you a little bit of time and a little bit of attention and maybe sometimes a little bit of money.

Shawn Cooper [00:20:49]:
Because I think that you can have a bigger impact with your ideas and your inventions than you can with just your pure manual labor or time.

Bonnie Evangelista [00:21:00]:
I love that there's a whole system that supports that individual. I love that. Thank you so much.

Shawn Cooper [00:21:09]:
Absolutely awesome.

Bonnie Evangelista [00:21:10]:
Thanks for taking some time.

Shawn Cooper [00:21:11]:
For sure. Great.