 Hey all, welcome back to theCUBE's coverage of KubeCon North America 22, CloudNativeCon. We're in Detroit, we've been here all day, covering day one of the event from our perspective. Three days of coverage coming at you. Lisa Martin here with John Furrier. John, a lot of buzz today. A lot of talk about the maturation of Kubernetes with different services that vendors are offering. We talked a little bit about security earlier today. One of the things that is a hot topic is national security. This is a huge segment we got coming up. It really takes all that nerd talk about Kubernetes and puts it into action, where you actually see demonstrable results. This is about advanced artificial intelligence for tactical decision making at the edge to support our military operations because a lot of the deaths are because of bad technology. And this has been talked about. We've been covering at Silicon Angle. We wrote a story there now on this topic. This should be a really exciting segment. So I'm really looking forward to it. So am I. Please welcome back one of our alumni, Nick Barsai, Senior Director, Customer-Led Open Innovation at Red Hat. Great to have you back. Greg Forrest joins us as well from Lockheed Martin, Director of AI Foundations. Guys, great to have you on the program. What's been your, Nick, what's been your kind of perception before we dig into the news and break that open of KubeCon 2022? So KubeCon is always a wonderful event because we can see people working with us in the community, developing new stuff, people that we see virtually all year, but it's the time at which we can really establish human contact, and that's wonderful. And it's also the moment where we can make big topic, move forward, and the topics have been planted at this KubeCon, from Microsoft to KCP to AI, all domains have been covered. Greg, you're the Director of AI Foundations at Lockheed Martin, obviously well-known contractor to the military, a lot of intellectual property, story to history. Sure. Talk about this announcement with Red Hat, because I think this is really indicative of what's happening at the edge. Data, compute, industrial equipment, and people. In this case, lives are at danger or to prevent, to preserve peace. This is a killer story in terms of understanding what this all means. What's your take on this relationship with Red Hat? What's the secret sauce? Yeah, it's really important for us. So part of our 21st Century Security Strategy as a company is to partner with companies like Red Hat and Big Tech, and bring the best of the commercial world into Department of Defense for our soldiers on the ground. And that's exactly what we announced today, or Tuesday in our partnership. And so the ability to take commercial products and utilize them in theater is really important for saving lives on the ground. And so we can go through exactly what we did as part of this demonstration, but we took MicroShift at the edge and we're able to run our AI payloads on that. That provided us with the ability to do things like AI-based RF sensing, so radio frequency sensing. And we were also able to do computer vision-based technologies at the edge. So we went out, we had a small UAV that went out and searched for a target on the ground. It found a target using its radio frequency capabilities, the RF capabilities. And then once we're able to hone in on that target, what Red Hat device edge in MicroShift enables us to do is actually then switch sensing modalities. And then we're able to look at this target via the camera and use computer vision-based technologies to actually more accurately locate the target and then track that target in real time. So that's one of the keys to be able to actually switch modalities in real time on one platform is really important for our joint all-domain operations construct, right? The idea of how do you actually connect all of these assets in the environment in the battle space? Talk about the challenge and how hard it is to do this. The backhaul, you'll go back to the central server, bring data back, connecting things. What if there's insecurity around connectivity? I mean, there's a lot of things going. Can you just scope the magnitude of like, how hard it is to actually deploy something at like a tactical edge? It is, there's a lot of data that comes from all of these sensors, whether they're RF sensors or EO or IR. We're working across multiple domains, right? And so we want to take that data back and train on that and then redeploy to the edge. And so with MicroShift, we're able to do that in a way that's robust, that's repeatable, and that's automated. And that really instills trust in us and our customers that when we deploy new software capabilities to the edge over the air like we did in this demonstration, that they're going to run right on the target hardware, right? And so that's a huge advantage to kind of what we're doing here that when we push software to the edge in real time, we know it's going to run, right? And the real time is absolutely critical. We talk about it in so many different industries. Oh, it's customers expect real time access, whether it's your banking app or whatnot. But here we're talking about literally life and death situations on the battlefield. So that real time data access is literally life and death. It's paramount to what we're doing. And this, in this case, the aircraft started with one role, which was to go find a radio frequency admitter and then switch roles to then go get cameras and eyes on them. So where is that coming from? Are there people on the ground? Are there people, dangerous people on the ground? And it gives the end user on the ground complete situational awareness of what is actually happening. And that is key for enhanced decision-making. Enhanced decision-making is critical to what we're doing. And so that's really where we're advancing this technology and where we can save lives. I read a report from General Mattis when he was in service that a lot of the deaths are due to not having enough information really at the edge. Friendly fire. Friendly fire, a lot of stuff that goes on there. So this is really, really important. Nick, this is like, you're sitting there saying, this is great. My customer's talking about the product. This is your innovation, Red Hat Device Edge in action. This is real, this is industrial. So it's more than real. It's actually, this type of use case is what convinced us to transform a technology we had been working on, which is a small form factor of Kubernetes to transform it into a product. Because sometimes, you know, us engineers have a tendency to invent stuff that are great on paper, but it's a solution trying to find a problem. And we need customers to work with us to make sure that the solution do solve a real problem. And Lockheed was great, worked with us upstream on that project. And helped us prove out that the concept was actually worth it. And we waited until Lockheed had tested the concept in the air. Okay, so Red Hat Device Edge and MicroShift, explain that, how that works real quick for folks that don't know. So one of the things we learned is that Kubernetes is great, but it's only part of the journey. In order to get those workloads on those aircraft, or in order to get those workloads in a factory, you also need to consider the full lifecycle of the device itself. And you don't handle a device that is inside of a UAV or inside of a factory the same way you handle a server. You have to deal with those devices in a way that is much more akin to a setup box. So we had to modify how the OS was behaving to deal with devices. And we reused what we had built in rail for each edge aspect and combined it with MicroShift. And that's what became with that Red Hat Device Edge. We're in a low swap environment. Space, weight, and power, right? Or very limited around a small UAS in this demonstration. So the ability to spool up and spool down containers and to save computing power and to do that on demand and orchestrate that with MicroShift is paramount to what we're doing. We wouldn't be able to do it without that capability. It's awesome. I want to get both of your opinions. Nick, we'll start with you and then Gray, we'll go to you. In terms of MicroShift, what is its super power? What differentiates it from other computing solutions in the market? So MicroShift is Kubernetes, but reduced to the strict minimum of a runtime version of Kubernetes so that it takes a minimal footprint so that we maximize the space available for the workload of induced very constraints in the environment. On a board where you have eight or 16 gig of RAM, if you use only two gig of that to run the infrastructure component, you leave the rest for the AI workload that you need on the drone. And that's what is really important. And these AI payloads, the inference that we're doing at the edge is very compute intensive. So again, the ability to manage that and orchestrate that is paramount to running on these very small board computers, right? These are small drones that don't have a lot of space. It's going to be efficient. It'd be efficient with it. How are you guys involved? Talk about the relationship. Still you guys were tightly involved. Talk about the roles you guys played together. Was it co-development? Was it customer partner? Talk about the relationship. Yeah, so we started actually with Satellite. So you can think of small CubeSats in a very similar environment to a low-powered UAV. And it started there and then in the last, I would say a year or so, Nick, we have worked together to develop MicroShift. We work closely on Slack channels together like we're part of the same team. That's great. And hey, Red Hat, this is what we need. This is what we're looking for. These are the constraints that we have. And this team has been amazing and just delivered on everything that we've asked for. I mean, this is really an example of the innovation at the edge, industrial edge specifically. You got an operating system. You got form factor challenges. You got operating parameters. And just having that flexibility, you can't just take this and put it over there. But it's what really is a community applied to an industrial context. So what happened there is we worked as part of the MicroShift community together with a real-time communication channel, the same Slack that anybody developing Kubernetes uses. We've been using to identify where the problems were, how to solve them, bring new ideas. And that's how we tackle this problem. Yeah, a true open source model. I mean, the Red Hat and the Lockheed teams were in it together on a daily basis, communicating like we were part of the same company. And that's really how you move these things forward. Yeah, and of course, open source is great, but also you got to lock down the security. How did you guys handle that? What's going on with the security? Because you got to make sure no takeover of the devices. So the funny thing is that even though what we produce is highly inclusive of security concern, our development model is completely open. So it's not security by obfuscation. It's security because we apply the best practices. You see everything. Absolutely, yes. And then you harden it in the joint development. There it is. Yeah, but what we support, what we offer as a product is the same for Lockheed or for any other customer because there is no domain where security is not important. When you control the recognition on the drone or when you control the behavior of a robot in a factory, security is paramount because you can immobilize a country by infecting a robot the same way you could immobilize a military operation by infecting a UAV. Not to change the subject, but I got to go on a tangent here because it popped into my head. You mentioned CubeSat, not related to theCUBE, of course, or theCUBE for the video. CubeSats are very popular. People can launch space right now, very inexpensively. So it's a highly contested and congested environment. Any space activity going on around the corner with you guys? So remember the edge is not, the world's not brown. It's edge is now in space. It's Mars is the edge now. Our first prototype for MicroShift was actually a CubeSat, an IBM product, the project called Endurance. That's the first time we actually put MicroShift into use and that was a very interesting project. Very early version of MicroShift and now we have talks with many other people on reproducing that at a more industrial level. This was more like a cool high school project. But to your point, the scalability across different platforms is there, right? Right, we're running on top of MicroShift on this common OS. It just eases the development. Behind the scenes, we have a whole AI factory at Lockheed Martin where we have a common ecosystem for how we actually develop and deploy these algorithms to the edge and now we've got a common ecosystem at the edge. So it helps that whole process, right? To be able to do that in automated ways, repeatable ways where we can instill trust in our DOD customer that, you know, the validation and verification of this is a really important aspect. It must be a fun place to work. So it is, it's exciting. There's endless opportunities. And let's get a lot of young kids applying for those jobs. They're barely into the hole. I mean, this AI is a hot feel and people want to get their hands on real application. I was, I was serious about space. Is there space activity going on with you guys? Or is it just now military edge, not yet military space? Or is that classified? Yeah, so we're working across multiple fronts. Absolutely. That's awesome. What excite, oh, sorry, John. So what excites you most as you said, never a dull moment with what you're doing, but just the potential to enable a safer and more secure world? What excites you most about this partnership and the direction and the, we'll say the trajectory it's going on? Yeah, I think, you know, for me, the safer and secure world is paramount to what we're doing. We're here for national defense and for our allies. And that's really critical to what we're doing. That's what motivates me. That's what gets me up in the morning to know that, you know, there is a soldier on the ground who will be using this technology and we will be giving that person the situational awareness to make the right decisions at the right time. So we can go from small UAVs to larger aircraft, we can do it in a small confined edge device like a stalker UAV. We can scale this up to two different products, different platforms. They didn't even have to be Lockheed Martin. And more devices that are going to be imagined. More devices that we haven't even imagined yet. Right, right, that aren't even on the frontier yet. Nick, what's next from your perspective? You know, in the domain we are in, next year is always plenty of things. Sustainability is a huge domain right now on which we're working. We have lots of things going on in the AI space. Stuff going on with Lockheed Martin. We have things going on in the radio network domain. We've been very heavily involved in telecommunication and this is constantly evolving. There is not one domain that, in terms of infrastructure, Red Hat is not touching. Well, this is the first of multiple demonstrations. The scenarios will get more complex with multiple aircraft and in the future we're also looking at bringing a lot of the 5G work if Lockheed has put a large focus on 5G.mil for military applications and running some of those workloads on top of MicroShift as well as things to come in the future that we are already planning and looking at. It's needed in theater to have connectivity. You've got to have your own connectivity. Paramount, right? Absolutely, it's Paramount. It's game-changing. Guys, thank you so much for joining John and me on The Keep talking about how Red Hat and Lockheed Martin are working together to leverage AI to really improve decision-making and save more lives. It was a wonderful conversation. We're going to have to have you back because we've got to follow this. This was great. Thank you so much. Thank you very much for having us. Our pleasure, thank you. We really appreciate it. For our guests and John Furrier, I'm Lisa Martin. You're watching theCUBE live from KubeCon CloudNativeCon 22 from Detroit. Stick around. Next guest is going to join John and Savannah in just a minute.