Host Intro (Gene Kim)

[00:00:17] So in 2020, we had Adam Furtado, chief of platform at the groundbreaking U.S. Air Force Kessel Run program, along with Lauren Knausenberger, now CIO of an entire U.S. Air Force.

[00:00:28] So I was so delighted that last year another group in the U.S. Air Force presented: Derek and Mike Snyder talked about the work that they were doing at another one of these amazing software efforts called Ski Camp, home of the first supersonic Kubernetes deployment demonstration.

[00:00:43] I got to spend two days with them in April in Ogden, Utah, earlier this year, and I got to see and experience so many amazing things. One of them was meeting U.S. Navy Captain Andy Biehn, who is currently executive assistant to Assistant Secretary of the Navy for Research, Development, and Acquisition.

[00:00:59] Before that posting, he was in a leadership role for the Aegis integrated combat system program, which included its modernization.

[00:01:06] So as a kid in the 1980s, I remember reading about the Aegis missile defense system in Tom Clancy books like The Hunt for Red October and Red Storm Rising, and how groundbreaking and critical that system is to protect tens or maybe even hundreds of thousands of sailors.

[00:01:20] So in Ogden, Captain Biehn presented one of the most amazing engineering talks I've ever seen, describing the incredible challenges they faced modernizing systems that were built in the 1970s. It created so many engineering breakthroughs back then, and he and the team created their own breakthroughs, which I think those amazing engineers in the 1970s would be very much proud of as well.

[00:01:38] I love the talk for so many reasons, including that it may be the best example of: if they can do it, what is our excuse? Here's Andy.

CAPT Andrew Biehn

[00:02:01] Okay, how are we doing, everybody? Awesome.

[00:02:06] It's been really fantastic to be here. It's my first DevOps Enterprise Summit, and what's been really cool is meeting all these people who have been doing this for quite a while, who really pioneered some of these efforts. A lot of folks have said, "Oh, yeah, when I was at the 2014 DevOps Enterprise Summit..." I was like, wow, because that's pretty cool, because in 2014, that's where I was, and I didn't know anything about DevOps.

[00:02:27] So I'm not a computer guy. I'm a career Surface Warfare Officer. That means I drive ships. I've been doing it for 27 years: five deployments at sea, one paid vacation to Afghanistan. One out of five. Do not recommend.

[00:02:45] So, not an engineer or computer scientist. My 13-year-old is in algebra. My role in helping with algebra is pointing him to his physics-major mom, because when the alphabet shows up in math, Dad's out.

[00:02:57] But I was fortunate to command that ship right there, USS Truxtun, an Aegis destroyer, and 300 of America's finest men and women. Then after that, the Navy asked me to become an acquisition professional, and as Gene said, most recently I led the Aegis program as the program manager. So we're going to talk about some of the stuff I did there before I got sent to the Pentagon to do penance for a year.

[00:03:20] First off, I normally give these kinds of talks to a Navy audience that doesn't know a whole heck of a lot about DevOps. This is exactly the reverse: you guys all know more about DevOps than me, but you may not know that much about the Navy.

[00:03:32] So, your Navy: about 500,000 people total, 295 ships, 2,600 aircraft, and on watch, as of last week, October 13th was our 247th birthday: on watch 24/7 for 247 years. Why we're there: to support and defend the Constitution of the United States.

[00:03:55] Against all enemies, foreign and domestic. That's the oath that all 500,000 of those people, military and civilian, take. And how do we do it? We conduct prompt and sustained operations incident to combat at sea. That's actually our mission from Title 10. So being in the Navy is all about being combat credible, and everything we do stems from that.

[00:04:14] Some foundational principles, some constants for the Navy. First, warfare is all about attacking effectively first. There's nowhere to hide in the ocean. So attacking effectively first is about finding the enemy and then developing a fire control solution, putting ordnance on target before he can do the same to you.

[00:04:31] We have to operate inside the enemy's observe, orient, decide, and act loop. That means we need to think quicker than the enemy tactically and, as you'll see here, technologically as well.

[00:04:42] The biggest constant is we operate on, above, and below the sea, the most demanding environment on planet Earth. That shapes everything we do.

[00:04:54] In my business, the combat systems business, we're all about the fire control loop: bringing the sensor and the weapon together and making sure that the accumulated errors throughout that fire control loop are small enough that the missile we're shooting can hit the target we're shooting it at.

[00:05:11] So that dominates everything we do. Variables: technology. Technology has evolved throughout the history of naval warfare, and we are a technology-driven business, whether it was the ironclad in the American Civil War, or the rise of nuclear power, or the development of naval aviation. Technology has changed naval warfare applications multiple times.

[00:05:32] The targets have changed. We no longer shoot at just other ships that are close by. We're shooting, obviously, at land, and then at missiles, including ballistic missiles, and figuring out how to attack hypersonic, in excess of Mach 5, cruise missiles.

[00:05:46] The range and speed of the weapons we're using has changed and grown along that time, as have the number and variety of sensors that we're using to go into that fire control loop.

[00:05:55] So let's talk a little bit about the evolution of war at sea. We'll begin with naval gunfire. On the left-hand side of the slide you've got the USS Constitution, oldest commissioned warship afloat. That's in the War of 1812, shooting cannons against the British frigate.

[00:06:08] On the right-hand side, you've got the USS San Francisco steaming into San Francisco Bay at the end of the Second World War.

[00:06:18] Gunfire at sea has always been platform-based. It's been around the individual ship. The sensors have been on the ship. The fire control has come from the ship. It has conveniently been a really good example of an open architecture. As technology has changed, whether in the guns or the sensors or the concept of fire control, we've been able to include that pretty quickly.

[00:06:39] San Francisco is in there because she shows the state of technology at the end of World War II, with the U.S. advances in radar technology. It helped defeat the Japanese in the South Pacific, and that evolution throughout the Second World War is an instructive lesson. I recommend the book Learning War by Trent Hone, which really talks about the Navy as an adaptive learning organization throughout the Second World War.

[00:07:00] And so that naval gunfire could evolve as quickly as the technology did, whether it was the weapon or the sensor. It was really just dependent on the ability of the organization to learn, integrate, and adapt.

[00:07:14] All of that changed here. This is in the 1967 war between Israel and several Arab nations. Those are Egyptian patrol boats on the left-hand side. They're shooting Soviet-made Styx missiles at the ship on the right-hand side, the Israeli naval ship Eilat. That was a former British corvette that had been given or sold to the Israelis. It was a modern warship with a well-trained crew, and it sank that day.

[00:07:38] Because the weapons and capabilities it had could not stand up to an anti-ship cruise missile that had been pioneered and then deployed by the Soviet Union at the height of the Cold War. So that was a real game-changer in surface warfare and naval warfare, and it opened a lot of people's eyes to how we needed to adapt technologically as a Navy in order to compete.

[00:08:01] And so that led to Aegis, which is the program I had the privilege of leading. Aegis is a 50-year-old technology program. It began in the '70s, and it is also platform-based. At least at its inception, it was about connecting the sensor, in this case a phased array radar known as the SPY-1, to a missile, the Standard Missile-2, and bringing those together within that fire control loop to defeat anti-ship cruise missiles. This was really designed to escort convoys across the Atlantic, assuming that there would have been a World War III against the Soviet Union.

[00:08:40] It was a very tightly coupled architecture, and there were two things that drove that. The first was the size of the radar and the missile launchers, et cetera, really demanded that the ship be built around them. They had to be designed from the keel up. You couldn't just strap them on. So there's a tight coupling between the ship itself and the weapon system, which literally is throughout the ship.

[00:09:04] Secondly, between the software and the hardware. As you can imagine, there's a little bit of math involved in shooting an anti-ship cruise missile that can be going faster than the speed of sound. So the limits of computing at the time in the '70s and early '80s really demanded tight coupling between the software and the hardware architecture, because we were using every ounce of compute that was available to do those calculations and support the crew in conducting area air defense.

[00:09:28] It was superbly effective. Humble brag: that's my ship shooting a missile. We hit. It works really well, but it was difficult to rapidly evolve. So in 2014, I was doing that kind of fun stuff, and then a few years later I'm the Aegis program manager, and I've got to figure out: how do I evolve this to get capability out there? Because my fleet, the customer, was demanding increased capability rapidly.

[00:09:58] But there are some significant modernization challenges. On the left-hand side you see an Aegis destroyer in the dry dock, and she's going through Aegis modernization. What are they doing? They're ripping out those computers right there. That's a UYK-43 computer. They're ripping that out, replacing it with a modern computing environment, plus replacing all the cabling, et cetera.

[00:10:19] Lieutenant Commander Biehn, 20 years ago, was a combat systems officer and did his cruise with UYK-43 computers. They weren't new then. They're still out there on multiple ships now. That gives you an idea of how difficult that is.

[00:10:31] So: new hardware, new software to go with it, or new hardware because we're bringing new software; very intrusive ship changes. You have to cut a hole in the side of the ship, which means taking the ship out of the water, because holes in the side of ships are bad.

[00:10:44] In order to bring it out. So it's time-consuming. We take these ships offline for one to two years. That's a ship that's not deploying, doing the nation's business. That means somebody else is deploying more. Finally, it's pretty expensive.

[00:10:58] So this is the challenge I was presented with as a program manager: how do we deliver combat system capability at the speed of technological change?

[00:11:05] And I had no idea. Somebody said, "Hey, you should do some of that agile stuff." I was like, awesome, what's agile? I'll go figure that out. So I start reading on that, and then I found DevOps and then The Phoenix Project. I read The Phoenix Project, and my wife turned to me one night: "So what are you reading?" I'm like, "I'm reading a novel about modern software development." She's like, "You are such a loser."

[00:11:24] But I learned the Three Ways, and so I want to talk about how we've been applying the Three Ways to transform Navy combat system development.

[00:11:33] First: flow, systems thinking. The biggest challenge we had was that inextricable link between our hardware and our software. We had to break that, because I can't just push stuff to the cloud. I can put a cloud on every ship.

[00:11:45] And so that's what we did. We virtualized Aegis, and we began doing that to simplify our test and evaluation process. Then we realized, hey, we can take it to sea and actually use this as a foundation for the Aegis weapon system, and that gets us a lot of good benefits in being able to avoid obsolescence, keep compute updated, et cetera.

[00:12:06] The second thing we need to do is really change how we're going from requirements to delivery. There you'll see on the bottom right-hand corner of the slide, we used to have very fixed requirements. We would get a set of requirements that would take about two years to make its way through the Pentagon. Then we'd issue a contract. They'd take about three years for a contractor to develop it. Then they'd give it to me, then we test it for a year.

[00:12:29] At that point, I have a really bad decision as a program manager. I can either kick it back to the contractor and punish them by giving them more time and more money to fix the things we found, or I can push it to the fleet who needs that capability, but knowing that I'm pushing them technical debt as well.

[00:12:45] So what we did is actually really changed everything from our contracting process to our development process, and built our own single DevOps pipeline called The Forge, which is our first combat system software factory. We're actually in the middle of that transition right now through a contractual competition, but breaking down the barriers between requirements, dev, and ops to one single flow where we've got the developers, the testers, the cybersecurity folks, operators, and our users right there throughout that process. That has been a pretty long journey, but it's been pretty awesome, and I've been fortunate to get great support along the way.

[00:13:23] The second one is to amplify feedback loops. This is USS Monterey. She was actually just decommissioned earlier this year. We did a pretty awesome experiment on her on her last deployment.

[00:13:32] As we talked about getting senior-level support, I was fortunate to have senior leader support from the Director of Surface Warfare in the Pentagon. It was a two-star admiral. I was at a different meeting with two four-star admirals briefing them, and we started talking about virtualization. He said, "What we really need to do is go do an experiment on Monterey because she's got the oldest combat system in the fleet. If you can do it there, you can do it anywhere." And the admiral said, "Yeah, it sounds like a great idea." I said, "Awesome." And then the team did.

[00:14:00] So what you see there is a modernized common display architecture that we delivered that provided up-to-date displays. What was really cool is, throughout the height of the pandemic, the team was able to deliver monthly updates to the ship, taking feedback from sailors, fixing things, giving them new features.

[00:14:20] When the ship was on quarantine and we couldn't actually take hard drives on board, which is how we did deliveries, we did our first deliveries. We had T1 lines pierced, so that was exciting. Then we even did two deliveries while they were at sea on deployment in the Persian Gulf, one delivering a cybersecurity fix, the other delivering new capability. Pretty awesome that the team was able to figure that out.

[00:14:39] Continual experimentation. If you want to convince a bureaucrat of something, bring more PowerPoint. If you want to convince an engineer, make a prototype. If you want to convince a senior naval officer, blow something up.

[00:14:51] On the left-hand side is the first successful missile shot from a virtualized Aegis weapon system. We did that, so people began to see that virtualization made sense, that it would work for us, that we could execute the fire control loop in a virtualized environment.

[00:15:06] Along the bottom you'll see the unmanned surface vessel Ranger. That's an autonomous ship with some containers on the back. That picture on the bottom right-hand side is one of those containers opening up and shooting a missile out of it that's controlled by a virtualized Aegis weapon system. So we were able to take a missile launcher, put it in basically a standardized shipping container, and then shoot it at sea off an unmanned platform.

[00:15:27] If you look at the top, that's a press release from Naval Forces Europe. That is somewhere in Europe. That is a containerized SM-6 launcher being deployed somewhere in Europe that didn't exist a couple years ago. It's in Europe for obvious reasons now, and pretty awesome that we were able to deliver that capability. I did not plan on making an Aegis truck when I became the program manager. Trucks normally aren't what I do, but that was pretty cool.

[00:15:55] Continual experimentation is a great way to get that senior-level buy-in, and I was fortunate to have enough resources to be able to do that without asking for a whole lot more money. Then when we demonstrated success, we started to get real reinforcements in senior-level buy-in.

[00:16:14] So what does that mean for us? DevOps is very much influencing the Navy's vision now. The Navy's vision is laid out here in the Chief of Naval Operations' Navigation Plan. I'll ask you to take a look at the text there, but I'll highlight a couple things: where we talk about a more agile approach to experimentation and force modernization, more partnerships, a culture that embraces new concepts, talking about rapidly adapting and fast-following in others, and then including sailors early in the development process, listening to their ideas, and driving healthy feedback loops with the industrial and innovation base. That really sounds a lot like DevOps.

[00:16:51] Now, Admiral Gilday, my boss, wrote that. He's not a big DevOps practitioner, but the concepts of DevOps are percolating throughout DOD because we're being driven to understand what we need to do technologically to compete with peer competitors in a really challenging world. I am one of several hundred folks who've played a small role in this, but it very much is influencing the Navy's vision and how we're going to go forward.

[00:17:21] Why does that matter? Because in the end, we talked a lot about platform-based weapon systems. Well, networks will defeat platforms every time.

[00:17:31] This is the Russian cruiser Moskva, sunk in the Black Sea earlier this year. When I had command of Truxtun and I took Truxtun up into the Black Sea, and spent a lot of sleepless nights — this was the first time Russia invaded Ukraine in 2014 — I spent a lot of sleepless nights worrying about this cruiser and all of her friends, because there were a whole bunch of Russian ships up there, and me and my friends on the ship.

[00:17:57] It's fascinating to me to see how quickly the game changed when faced with an adaptive, well-trained, well-supported enemy that was able to rapidly integrate new sensors, sources of intelligence, and weapons, and execute them effectively in a distributed manner to really change the nature of the war in that area.

[00:18:22] Now, I'm not stupid enough to have any vision of how Russia and Ukraine is going to work out. Predictions are hard, especially about the future. But certainly, for those of us in this business, there are a lot of professional lessons learned there about how distributed networks and adaptive people and an open architecture, rapidly learning environment can tip the balance of warfare in the modern world.

[00:18:57] So what kind of help am I looking for? Three things that I would ask from you. First, I'd love to talk to anybody who's transforming legacy real-time software into a modern architecture to support CI/CD. Real-time software: we're operating real systems operating in the physical environment, and I've got a big monolithic code base that we're trying to break down to support continuous integration, continuous delivery. If you've got experience with that or ideas about that, I would love to talk to you and hook you up with the people that are doing this now.

[00:19:25] Secondly, normally when I go to an event similar to this, I talk to pretty much the same defense industrial base, and we love those folks. There's nobody better in the world at what they do. But I really encourage innovative people and companies to explore how they can contribute their skills to support our sailors and Marines, because a lot of the work that's being done by companies and people here is very different but has immediate applicability. So if you've got some ideas on that or your company does, I would really love to talk to you about that and help you get connected with some of those opportunities.

[00:19:59] Thirdly, I'd ask you: if you've got any talented people in your lives that you mentor, advise, et cetera, who are considering what they're going to do, ask them to consider a career in public service. We frankly have a greater separation between the civilian world and the military world than we've had in a long time in this country. One of the positive benefits of the all-volunteer force is we get great young men and women to come in. One of the negative sides is it's in some ways becoming a family business, and then a lot of people who don't know anything about the military don't have any opportunity to know people who served in the military or have those experiences.

[00:20:41] So I'd ask, if you've got people like that in your lives, encourage them to consider an opportunity for a career in public service, whether in uniform or out. First off, we're looking for really talented people. We do really cool work. It literally is rocket science, and it's a remarkably rewarding life, making a real difference for our country. So thanks very much for your time.