 Live from Barcelona, Spain, it's theCUBE. Covering Cisco Live 2020. Brought to you by Cisco and its ecosystem partners. Welcome back, this is theCUBE's live coverage of Cisco Live 2020 here in Barcelona, Spain. I'm Stu Miniman, my co-host for the segment is Dave Vellante. John Furrier is also in the house. We're doing about three and a half days wall-to-wall coverage. The surface area that we are covering here is rather broad and I use that term, my guest is laughing, Vikas Bhutani, who is the vice president of IoT, of course, extending the network to the edge, to the devices and beyond with Cisco. Thanks so much for joining us. Oh, it's great to be here. All right, IoT thing, I've worked with Cisco my entire career. I watched through the fog computing era for a couple of years. Edge, of course, one of the hottest conversations, something that I've brought up, many of the conversations across the portfolio, but listen, Tony, was up on the main stage for the day one keynote, talking a lot about IoT and IT and OT and your customers and the like. So let's start there as, what's new? How does IoT fit into the overall Cisco story? Absolutely, so as Liz was on the main stage and David talked about the cross-domain and the multi-domain architecture, now IoT and our operational environment is one of the key domains within that environment. And what Liz announced yesterday are two pieces of new that we are releasing at Cisco Live. First of them is an IoT security architecture, which ties together the capabilities with cyber vision and then integrates it within our rest of the IoT security portfolio. And the second part that I'm also excited to talk about is edge intelligence. It's about how we are helping our customers extract the data at the edge and then deploy it and move it to wherever the applications are in the multi-cloud environment. We definitely want to dig into those pieces, but IoT is such a diverse solution set. So it's often helpful to talk about specific industries, any customer examples, so what can you share with us there to help illuminate where Cisco's helping the customers? Love the security angles and edge. Yeah, just a level set. Like when we think about industrial IoT, we're really talking about the heavier industries, plant environments for a manufacturing company. We're thinking about roadways for a public sector customer. We're thinking the grid for the utility environments. We're thinking refineries and oil extraction upstream environments. So this is the kind of spectrum in which we are working in where customers have real businesses, real assets, where the operations is the heart of the enterprise that they're running in. And the technology can really be a revolutionary change for them to help them connect and then extract the data and then be able to make sense of the data to improve their business practice. So industrial IoT, whether you're a roadway in Austria, like Afsena, you're a utility in Germany like Energy, or a EDF in France as an example, NL and Turner in Italy. All of these industries and all of these customers are using industrial IoT technologies in running their businesses better today. Where are we in terms of that critical infrastructure being both connected and instrumented? Where are we in the adoption curve? Sure, look, in many of these industries we have talked about SCADA systems, right? That have been here for 30 plus years for our customers. In most of those environments it's really a one-way flow of information, right? And typically customers stood up separate siloed networks which weren't really connected to the rest of the enterprise. So Rockwell has a saying from the shop floor to the top floor, right? Like how the digital enterprise where all of these environments are coming together is where customers are. Critical infrastructure, as you said, in this day and age with security and other kind of threats, customers are a little hesitant about how they connect it all together. But Cisco is working with these customers in helping them think through the benefits they can get but also make sure from a cybersecurity point of view that you're helping protect the assets, manage these environments because you can't just arbitrarily connect them because IT tool sets just are not ready to manage these environments. Yeah, I love that all the examples you gave were European, of course, being here in Europe. Curious, there's some technologies where North America might take the lead or Asia might take the lead. Is IoT relatively distributed? Is Europe kind of on par with the rest of the world when it comes to general adoption? What we have found in Europe, because of many countries like Germany leading in the renewable energy effort and the climate is a big focus here. Data privacy and concerns around data sharing are much more top of mind in Europe. So we find those kind of use cases getting adopted much, much faster. So in Germany, as an example, Energy, which is one of our customers and they were here with us last year at Cisco Live and we launched a capability with them. They are trying to manage the real time flow of energy in their grid environment such that make sure that there are no outages, no brownouts in these environments. So utilities and customers like that across Europe are adopting technology faster. Manufacturing as always is a leading use case. There we see some of the automotives in US are leading a little bit more in getting environments connected to their environment. But overall, IoT is a global market. We work, we have over 70,000 enterprise IoT customers today at Cisco, right? So we are fortunate to be able to serve these customers on a global basis across the range of industries I talked about earlier. And in a lot of respects, too, I would say the US is behind, right? When you look at public policy from a federal standpoint, US really doesn't have a sort of a digital strategy from an overall perspective, whereas you're certainly India does in certain countries in Europe. You look at the railway systems in Europe. Much more advanced. Beautiful and shiny and advanced. And so, you know, I would say the US has a little bit of work to do here. That's right. My perspective, I wanted to be sure. No, that's right. Like in India, Prime Minister Modi started the effort around 100 smart cities, right? And like Cisco is working with many of those smart cities with our Cisco Kinetic for Cities to kind of create this, connect all of the sensor networks, video surveillance, safety, environmental sensors, managing the flow of that data and digitizing those environments, right? And in Europe, we've been working in France, Germany, Italy, UK, right? I mean, I think we're seeing much more like adoption in these specific industries, but it's a global market. And you know, again, like I said, 70,000 customers, we get to see quite a bit of the landscape around the globe. What should we know about the architecture? Can you give us kind of a high level summary? What are the basics? So in the comprehensive IoT security architecture that we released this week, it really starts with, you know, you have to be able to identify the devices, right? In IT environments, you know, to your laptop and to your PC, like they have been managed by MDM technologies for years, but in the industrial environment, I might have a programmable logic controller that I deployed 15 years ago. It's not ready for modern capabilities. So what you have to really start with is identifying all of these assets and the communication baselines that are happening there. That's step one. Step number two is really, now that I know that this is a PLC or that's a controller, I need to come up with a policy, a security policy, which says this cell in a plant environment can only talk to the other cell, but doesn't need to talk to a paint zone. So I'll give you an automotive. If I'm welding a car, I'm building a car, the welding robots need to be communicating with each other. There's no real reason that the welding robot needs to talk to the paint shop as an example. So you can come up with a set of policies like that to keep these environments separate, because if you don't, then if there is one infection, one malware, one security, then it just traverses your whole factory. And we know customers in Europe that have the networks have gone down and they've impacted $150 to $200 million of downtime impact. Well, we have a real world use case 10 years ago or so with Stuxnet, with Steven's PLC and boom all over the world. I mean, it's amazing. Exactly right. So again, back to identification, then I create the policy, then I implement the policy within our switching or a firewall network, but you're never done. So you have to keep monitoring on a real time basis as the landscape changes, what's happening, how do I keep up with it? And that's where things like anomaly detection are super important, right? So those are the four steps off the architecture that I want to talk about. Yeah, so it sounds like something like cybersecurity is both a threat and an opportunity of bringing together IT and OT. Bring us inside a little bit those dynamics. We know it's one of the bigger challenges in the IoT space. Yeah, I mean, I think like, look, both parties, whether I'm an operational person or an IT person, both of us, both the audiences have their own care about. If I'm a plant manager, I'm measured on number of units I'm producing, the quality, the reliability of my products. If I'm an IT, I really am measured on downtime of the network or the cybersecurity threat. Like there aren't really common measurable capabilities, but cyber and security, it kind of brings both the parties together, right? So when we use our cyber vision product, we're able to provide to that plant manager visibility to what's happening, how are their PLCs performing? Did anybody change my program? Is my recipe for my given product I'm making secure and safe, right? So you have to appeal to the operational user with what they care about. IT really cares about to manage the threat surface, don't let that threat kind of propagate. Now at the board level, because the board sees both sides of it, they're asking these teams to work together because they have complimentary skillset. Well, I think it's critical because, rhetorical question, who's bigger control freaks? Network engineers or operation technology engineers, right? They both, you know, they want to keep that, you know, operation going and they're very protective of their infrastructure. And so it's got to come from top down and it is a board level discussion, right? I mean. Yeah, we have customers where, you know, the board is kind of the CEO has mandated to say, listen, whether it's for the national threat actors or other corporate espionage, right? Like I need to protect the corporate intellectual property, right? And because it's not just a process, it's also about safety of employees and safety of their assets that comes into play, right? So when some of the customers we are working with, where the CEO has kind of dictated that the IT teams help the operational environments, but it is a two-way street. Like there has to be value for both parties to come together to solve these challenges. Okay, so we talked a little bit about the threat. Also when we're talking IoT, there's all that data involved. What's the opportunity there for customers with data? How is this going involved? Absolutely. Look, I mean, I think one of the reasons that customers are doing digitization projects is because they're trying to use the data to make better business decisions. It has to improve yield and meet their KPIs of their industry. So far what we have seen is that all of the data is really trapped in all of these distributed environments. Gartner tells you that 75% of the data will be produced at the IoT edge. But our customers to date have not had the tool set to be able to get access to the data, cleanse the data at the edge of the network, bring the right data that they can create insights with and improve their businesses. So it's been a heterogeneous environment, lots of protocols, lots of legacy. So that's kind of what our customers are struggling with today. Yeah, absolutely. And most of that data is going to stay at the edge. So we need to be able to process the edge. Heck, I even went to a conference last year, talked about satellites that are collecting all of the data. I need to be able to have the storage, the processing, the compute there because I can't send all the data back as fast as it is. So it's a changing architecture as to where I collect data, where I process data. We think it is very much additive to the traditional cloud and data center environments today. It's just yet another challenge that enterprises need to deal with. That's right. So the work that Cisco is doing in the IoT edge environment is, we are enabling these customers to connect their remote terminal units, their machines and their robots and providing them the tool set with four capabilities. First, extract the data. And so we have a set of protocols like Modbus, like OPC UA, where they can extract the data from their machine. So step number one. Second, the data is to transform the data. As you said, over a LTE circuit or over a connection, I'm not going to be able to send all of the data back. So how do I transform the circuit, transform the data, where I maybe take an average over the last five minutes or I kind of put some functions and we are providing, as we're in the DevNet zone, we're providing developers the capabilities such that they can use Visual Studio, they can use JavaScript to write logic that can run right at the edge of the network. So now you have extracted the data, you have transformed the data. Governance is a key topic. Who should have access to my data, especially here in Europe, where we're concerned about privacy, we're concerned about data governance, we are enabling our customers to come up with the right logic by which if there's a machine data and you are the supplier, I'm only going to give you the data, the temperature, the vibration, the pressure that you need to support the machine, but I'm not going to give you the number of units I produce. I'm not going to give you the data about my intellectual property and then you have to integrate to where the data is going, right? So what we are doing is we are working with the public cloud providers, we are working with software ISVs and we are giving them the integration capability and the benefit of this for the customer is we have done pre-integration on the extraction part and we have done pre-integrations on the delivery part which allows them to projects to go faster and they can deliver their IoT efforts. So how do you envision the compute model at the edge? I mean, probably not going to throw a zillion cores, you know, so maybe lighter weight components and I got some follow-up on that as well. Sure, absolutely. Look, Moore's Law is a friend of ours here, right? Like with every cycle, every generation of CPU technology, you get more and more compute capabilities. So the IoT gateways that we provide to our customers today have four ARM cores in them. We are using two of those ARM cores for the networking function, but those cores are available for our customers. We have designed an extra memory for them to be able to process these applications and we give them SSD and some storage at that so we can provide up to 60 gigs or 100 gigs of storage such that now that gateway, that communication device, a router, a switch that's at the edge of the network can kind of do this dual purpose. It can not only process and provide you security for the, and the communications but is now an edge processing node. So we call them IoT gateways and I can tell you like, we are deploying these kind of products on buses, you know, in a mass transit bus. We all ride the buses. There are over six systems that are on that bus, a video surveillance system. I want to monitor the tire pressure. I want to monitor if the driver going over the speed limit. Like we have now connected all of these systems and we are running logic at the edge such that the riders have a safer experience and then they can get real time visibility to wear the buses as well. Yeah, and my follow up was on persisting. So you mentioned storage, you know, flash storage at the edge. And then, you know, you also referred to earlier the challenges, this data today is locked in silos or maybe it's not even persisted. You know, it's analog data all the time. So do you envision, I mean, think about successful, you know, digital companies, kind of born and born digital. Data's at the core. It's traditionally big manufacturing firms, large infrastructure. It's the manufacturing plant or is the center of the universe and data sort of sits around it. Do you envision a period where that data is somehow virtualized and have access to it? We can really build digital businesses around that data. What are your thoughts? Absolutely. So we have been working with a customer. It's like a steel manufacturer in Austria, right? Like the heartland of Europe as an example. And they make high quality steel, right? And when they're building the high quality steel, they have 200 different machine types. And like you're saying, the data is trapped in there. This customer is trying to digitize and trying to do that. But they have been struggling for the last two years or so to be able to get the data with this variety of machines. And they want to use Azure IoT services, but they haven't been able to pipeline the data all the way to their cloud environments. So that was one of our lighthouse customers. And we work with them, like, you know, roll up your sleeves and kind of design the system with them. And we worked to get that data, such that now they're not quite a born digital company, but they are a heart manufacturing company. They can get the best of the tool sets and analytics and all of the things that contemporary companies use. And they can bridge them into this digital environment. And this is how the incumbents can compete with the sort of cloud or digital natives, right? I mean, this is the equilibrium that occurs. That's right. I mean, look, I think we love the digital companies, but they're not really, they don't have physical assets out there working. And I'm more physical or the more of the real economy, whether if you are an oil company and you're extracting oil from a pump jack, right? Like, well, you need to still have the capability to do that better, right? So that's what we are doing is whether you're a transportation like the bus example I gave you, an oil and gas company who's trying to extract oil from the ground or you are a manufacturer or you're a utility, if we improve, use our digital technologies and operate, improve the efficiency in that business, a 0.1%, a 1% that has got much, much bigger implication for us as a society and the world at large, but just making them better and more efficient. Huge productivity gains. Massive. That's right. I think that technology and IoT technologies can benefit all of these industries and Cisco is going to invest it in kind of helping our 70,000 customers to get better with all of these capabilities. Awesome. Congratulations, 70,000 customers. Big number rolling out IoT solutions. Look forward to keeping track of Cisco's IoT solution. I'm super excited to be here. Thanks again. For Dave Vellante, I'm Stu Miniman. Back with lots more Waterwall coverage here at Cisco Live 2020 in Barcelona. Thanks for watching theCUBE.