 Live from London, England, it's the Cube. Covering Discover 2016 London. Brought to you by Hewlett Packard Enterprise. Now, here's your host, Dave Vellante and Paul Gillis. We are here at Excel London. This is the Cube, the worldwide leader in live tech coverage and we're covering three days wall-to-wall coverage of HPE Discover 2016, extracting the signal from the noise. Dave Slider is here as the vice president and general manager of the Communications, Media and Entertainment Solutions Group at HPE. I'm going to help the telcos clean up that hairball and doing a good job of that. And Paul Godonis, the president of Inmersat Enterprise, the division of Inmersat. Welcome to the Cube, good to see you guys again. So let's start with you, Paul. What is Inmersat? So we're a global mobile satellite communications network provider. We basically provide satellite communications to a number of different verticals, including shipping, which is our history and heritage, aviation, governments, and also enterprise businesses. So Dave, this is a good IoT use case. We're going to have a reference before trying to help the telcos, which is a whole different discussion. We're going to focus on IoT today. Talk about your role in your division in IoT. Sure, so from an HPE perspective, IoT is all about explosion at the edge. From millions and billions of devices to trillions and quadrillions, so we're adding a couple of zeros there. And of course, a company like HPE, having compute and storage and networking at the edge has a role to play. But the interesting part from my part of the world, expanding out of the telco world, is the connectivity and the device management. So being able to auto discover these devices when they say hello world and get turned on, got to provision them, got to fault manage them, et cetera, and it's all got to be automated. There can't be any cost. A lot of these use cases are very low revenue. So you've got to be touchless, humanless, automated, very high volume. Paul, I remember the explosion of interest. There was a temporary lip of interest in satellite phones back about 20 years ago in the early days of cell phones. Never really took off, very expensive. Has something changed about the technology now that has made your type of technology more affordable? Yeah, I think it's the same with any technology really. When it's initially introduced, it's always a lot more expensive. But then as time goes on, components become cheaper, launch systems become cheaper. For example, actually the economics change vastly. And so we're getting to a position now where actually a voice call on a satellite system is cheaper in some cases than global roaming on a cellular system. And how much of your business is voice versus data? Interesting question, I don't have the numbers exactly, but it's changing vastly. So it used to be very much down to voice, but we're seeing the change become much more pronounced now. So we're well over the balance, the tipping point of data being the major part of what we do. What kind of performance are you able to get over your network for data? So performance in terms of speeds, then really you're talking anything from half a megabit a second for a B-GAN system, which is our real global mobile system. Or up to 10 megabits per second to 50 megabits per second for a global express K-A-BAN. So very different. So Dave, what are the use cases you're seeing for edge computing and IoT? I don't know, I want to dig into Paul specifically. So they're all over the place and that's the interesting thing as part of this explosion is on multiple accesses, right? It's consumer, it's enterprise, it's smart metering, smart cities, smart trash cans all the way through to the stuff that Paul and team are doing. The neat thing that we find though is that the innovation is occurring with these very specialized, broad ecosystem of companies. They're very, very good at making very specialized devices and very good at making specialized applications in the cloud. They don't want to think about or worry about anything in between. So they come to an HP and they go, look, if you can do the connectivity and the device management as a service and just take care of everything in the middle, we'll do what we're really good at at the edge and in the cloud. So that ecosystem just explodes because we can take care of that middle for them. And so how are you, Paul, using this edge cloud combination? We've done some research at Wikibon that shows sort of the cloud only, I think you've seen this, Paul, the cloud only cost versus if you do edge plus cloud, you reduce data volumes, you lower costs significantly. Is that what you're seeing and how are you applying this concept? Well, Dave actually makes a really key point for us and that's the point that mostly the part in the middle where you have the devices and actually the outcome is a very little interest to anybody. It's just something that anybody expects just to be there. So, and that's where we sit. That's where it must always been. So we can't rely on those kind of connectivity sales anymore. We have to really start to develop that whole ecosystem as Dave put it that we can actually sell an outcome to a customer. So it's really not just owning that middle ground, but starting to develop the links into both ends. And that's really what is very important to in our SAP with this relationship with HP Enterprise. Do you see any industries now that are ahead of the curve in terms of adoption of IoT using your technology? IoT in transportation and logistics has been around for quite some time, but it's still a very simplistic use case. What we are finding actually is that there's a huge amount of take up in the energy market, in mining in particular, and also in agriculture. And agriculture is one of the most interesting ones. It's surprised us really. You'd expect the energy market and mining where you're pressured by commodity prices and a dollar per barrel, trying to use technologies to drive efficiencies in your operations is to be expected. But agriculture is actually becoming something that shows a lot of interest for us. And there's an area I think that we, HP Enterprise and InMOSAP, will collaborate on a lot in the future. Why is that? I'm just curious, why agriculture? Well, you think about the challenges that we've got with growing populations, problems with water shortages. It's all about driving the productivity of food production. So if we can use water in a smarter manner without wasting it, if we can actually produce food at a lower cost, then we can actually start to drive a more sustainable ecosystem for food production. Can we go through a practical example in agriculture? Take us through, walk us through. Sure. So one actually that we're working on in particular is a plantation. This plantation has an issue with being able to drive productivity of the plantation itself. So the issue is that they have mail trees and female trees. So basically the mail tree really reproduces and drives the growth of the plantation and the sustainability, but the female tree produces the actual crop. And at the moment it's been difficult for the farmer to be able to understand exactly what is going to drive the productivity. It's a bit of a hit and miss gamble about that mixture of male and female. What we've actually done is we've realized that you can really see and develop and change that mixture of male and female to drive productivity by controlling the irrigation of that area which has an impact on the soil type and what it's created. So by using IMSAT to provide the connectivity and the sense of technology, HP Enterprise to actually do the data analytics and provide the interface, then we can allow the farmer to be able to control the irrigation to drive productivity in the plantation. And you're instrumenting what? The soil, the trees, the whole system? We have sensors in the soil around across the plantation, hundreds and hundreds of sensors throughout that whole area which are then checking the level of irrigation and then feeding that information back to a data aggregation source which at the moment is then taking that data back into HP Enterprise's data analytics capabilities and then punching the numbers and bringing back the output. The interesting thing actually is in the future which may actually sound a bit strange coming from the connectivity provider but it's actually doing some of that analytics at the edge which is going to be really interesting for a lot of our customers. So what I want to ask you is, isn't that kind of where you ultimately, well it depends on how much data you have to move, right? But if it's a lot of data, you don't want to have to move it. Yeah, well and you think of this use case in particular is so intriguing especially as a relative newcomer from my perspective, right? So you're talking about irrigation ditches on a palm tree plantation where the people are under a lot of pressure sort of geopolitically, right? Not to be knocking down more rain forest and planting larger. So increasing the yield on your existing usage, you got drainage ditches with little dams in them with sensors. So we do the analytics and say, hey, section grid 42 or whatever, send more water or send less and a little gate opens or closes. All gets uplifted over satellite and analyzed in the cloud. Now if you put an edge line server out and because there's hundreds and thousands of these things you put a little edge line server out you can optimize what you do on plantation versus in the cloud. When you get, when you look at customers who are sort of on the leading edge with remote IoT, particularly occasionally connected, what balance are they trying to strike between use of broadband and use of compute resources? Are they tending toward minimal connect time and maximum local processing or always connected? What does the trend seem to be there? It varies hugely based on the use case. If you think of things that involve video and video surveillance and face recognition, et cetera, obviously the implications are you don't want to be backhauling huge, huge amounts. Whereas a smart meter, 50, 60, 70 bytes once a month send it to the crowd. If you need to retry it a couple times, not a big deal. So the use cases, and this is back to the point I made earlier that I think is really, really key. Unleashing the power of that ecosystem, as HPE is saying we can do the connectivity and we can do the device management. You do the edges, so suddenly there's an explosion and the variance is huge so there is no single answer at all. And you're doing the analytics as well, is that as a managed service? And you learn, right, over time and the palm plantation example is a really good one. We're going to learn over time what we should be doing at the edge, what's the optimal use of the connectivity satellite because a lot of these places don't have good coverage of anything but. But you're selling analytics as a service? Yes. Is that, what is that, a haven service? Yeah, it's all part of our universal IoT platform that does this device management and analytics that has northbound APIs in so the partners like Inmersat can do analytics on their own. We do some base level use cases and we're finding every use case we do we add to the capability. And the plantation, they're paying for this as a service or as a combination of, they put in infrastructure obviously they're paying for the sensors and who do they pay, does that all work? Well it's still early days and a lot of what we're doing so we're in a proof of concept stage with a lot of these kind of things and we're working out the business models but it will be an as a service type of system rather than. And the customer sees Inmersat or they see Inmersat and HP, how do you guys go to market? Yeah, it's very much a joint effort so they see us both together. It's actually been my team that's developed this proof of concept initially and then engaged with HP with this one particular plantation. What's been really interesting from our perspective is that actually the HP Enterprise team particularly Gary Wood and his guys have come back and found a number of other use cases so not just barn plantations but actually how can you control irrigation to conserve water in a walnut plantation in California where actually there's a huge issue with how you use your water. So it's that partnership where we're starting to see an initial concept developed by Inmersat that actually HP Enterprise is just exploding exponentially into other areas and that's the power of the partnership. Well I wonder how you see your business developing from here as IoT takes off on this huge ramp up right now will you be selling complete solutions with compute and sensors or will you be sticking to your broadband knitting? Well we have a very loyal and strong distribution channel ourselves and so where we are is at the moment we're developing these proof of concepts and the start of working with HP Enterprise and the goal is to enable our channel partners to be able to start selling to that into their existing customer bases. So for us it's about how much value can we drive through that distribution channel and what can we Inmersat provide in partnership with HP Enterprise to help them engage with their customers and drive more value. So Dave the strategy if I understand it is to take advantage of this edge explosion by putting resources on the edge actually doing the analytics locally reducing the amount of traffic that goes back which is ironic that you're actually supportive of that but the volumes of data are so high the opportunity is to provide the service. Yeah and I think the key is that actually it's about going back to that customer and finding more and more use cases so actually how do you track the individuals that are out loan workers in the farm? How do you actually start to put sensors onto vehicles tractors for example help them with zero and predicted downtime? Can we put because a lot of these are near the coast put salinity monitors in to see if there's any seawater coming into the plantation and so as we start to drive one use case to the edge and that means less data goes over the Inmersat connection we start to find more use cases actually that we can then enable with Sattler. And HPE's differentiation is you actually have systems that you can put at the edge and do that. Well I think our differentiation is the end to end right? The fact that we have the edge is fairly obvious I think what surprises a lot of people is this connectivity and device management and analytics the platform in the cloud. So our ability to go in the flexible business model where we can go to the Inmersats of the world and work together iterally and our view is look we'll provide it on-premise, we'll provide it in the cloud we'll provide it hybrid, we'll provide it as a service so we can make it economical 100 units PLC up to billions. And by cloud you mean whatever cloud you want? Yeah, private, public whatever the variations are huge around the world. Google and Facebook have both floated rather phantasmical concepts of making broadband universal involving balloons and giant mylar kites and such, do you see any alternative to the satellite emerging in the foreseeable future? I think that the use cases are different. So the Google Loon type projects and the Facebook projects are about bringing connectivity to the unconnected. So it's almost about consumer level connectivity whereas we're about industrial grade enterprise connectivity which you need to rely on a much different level of system. So I see the two is different and I believe that the satellite technology will change but I still think it will be there. Excellent. All right gentlemen we have to wrap. So last question for you Dave just give us the vibe of the show specifically as it relates to IoT we're kind of in the IoT zone right around here you see people playing with different devices and you know liquid pumps. It's fantastic the energy level is every time I think it can't go any higher and it does and it's really exciting for me and my team particularly as we've come out of the telecommunications domain to be able to work with Paul and the other this ecosystem that exploding. Like it's just phenomenal to think of the sort of human impact that we're all going to have in the next decade or so a lot of fun. Gentlemen thank you very much. Appreciate it. Good luck with the initiative. All right keep it right there Paul and I will be back with our next guest is the HPE Discover 2016. You're watching theCUBE.