 Live from London, England, it's theCUBE. Covering Discover 2016 London. Brought to you by Hewlett Packard Enterprise. Now, here's your host, Dave Vellante and Paul Gillis. Chris, Chile, London, it's like a February day in Boston here, Paul Gillin, my co-host. And I am really pleased to welcome Richard Phillips, who's a global technology leader at GE Digital and Trippartane, who's the HP CTO looking after GE within Hewlett Packard Enterprise. Gentlemen, welcome to theCUBE. Thank you. So, Bill Rue had a split. We've had him on theCUBE many, many times, and we know the GE story. We covered the announcement of the initial industrial internet, Beth Comstock on. We were just doing mines and machines. Very exciting times. So, and you spent some time in Silicon Valley. I did, yeah, I spent two great years out there. And so, tell us what it's like to be this industrial giant, trying to build a software company, a multi-thousand-person software company in the heart of Silicon Valley. It's exciting, it's complicated. It's breathing a new breath of fresh air into an industrial company and reinvigorating the way we do business. It's completely changing mindsets and perceptions as GE, a 125-plus-year-old company into something, a digital, relevant organization. It's great to be part of that spearhead. So, Tripp Partain, you get the award for best name on theCUBE this week. You won out over Antonio Neary, sorry, Antonio. So, tell us about your titles, really interesting. You got GE in your title. So, obviously, that shows a commitment to the partnership. Tell us more about the partnership in your role. Yeah, absolutely. So, my main role is to work on key technology innovation projects between the two companies, which sounds really cool because it is really cool, right? And we've had a history of doing these types of projects with GE even before GE Digital. A lot of folks may not realize when in hospitals, when you've got a GE CT machine or the mobile X-ray machines, there's a lot of HPE technology embedded inside of those that takes care of handling, you know, in the case of the CT machines, all the different images that are being collected and digitized and then processed very quickly there on the spot. And so, we've been doing those types of projects with GE for a number of years. And so, extending that into GE Digital, being there at Mobile World Congress when they had the official announcement of GE Predix going live, and then really developing the partnership further from there has been very exciting. Great, Richard, where are you seeing bottom-line benefits right now from your industrial IoT initiatives? What are some examples? So, for us personally, we're actually laying out a lot of our brilliant manufacturing technologies into our own manufacturing shops. So, GE operates some 400 manufacturing lines globally. And by the end of this year, we'll have about 75 of those introduced with Predix and the digital technology that we're bringing to bear. That's going to save us not only production time, it'll improve our quality of manufacturing, but it's going to reduce costs enormously for the whole of GE, which is going to keep accelerating that expansion throughout our production lines. All of our business units are seeing huge productivity increases through using our digital technology, being able to monitor the assets that we actually manage on behalf of our customers. We're able to do so more effectively. We're able to predict when failures are going to occur and proactively take care of them before failures occur. And that's really important, especially in things like the airline industry. Unscheduled ground time is a huge problem. And if we can do anything to help reduce or remove that entirely, that can only benefit our customer base. Well, I hear this figure, $6 billion in revenues and digital revenues for GE this year or last year. Where are those revenues coming from? So from the combination of different pieces we have, when we actually formed GE Digital itself and we moved our IT portion of our group under the GE Digital organization, we were a top 25 software company just by doing that. And we had around about a $4 billion order book just from doing that now. Now once we start introducing predicts and start making the software sales and subscriptions to those kind of models, we're looking around about $6, $6.2 billion this year. So that is truly selling predicts and our IRT solutions to our customers who are subscribing to them and making them use in their own enterprises. Now we also have enterprise customers building capabilities in our predicts solution and reselling to their own customers. And you've said, I think, both Jeffrey and Mel and Beth Kompstock have said on theCUBE that you plan on being a top 10 software company by the end of the decade. Yep. What's, tell us about how we should be thinking about the difference between the consumer internet and the internet of things or the industrial internet of things. So where we differentiate is, you know, for the consumer internet or kind of the IOT sphere, we're not the difference between the industrial internet of things is the type and nature of the data that we're dealing with. We're not trying to connect toasters and fridge freezers to the internet. I'm connecting jet engines, steam turbines, nuclear power plants and collecting the data and dealing with that. The type of data and the nature which we have to deal with it is so entirely different. The scale of the data is different. And the way in which you have to manipulate the data and manage it to be able to surface the right information out of that to be able to make use of it in a timely manner is so different to that was, you know, when I need a switch on and off the toaster. It's just the scale is so different. There's been a debate and I think both of you can speak to this over sort of where does the processing happen? I know HPE is very big on processing at the edge. What are you learning about? What's the optimal balance of centralized versus edge? We're seeing now that we're kind of been rethinking about this over the past couple of years or so. You know, everybody focuses around the cloud platform and they think that's going to solve world hunger. And realistically it isn't. You need to spread the processing out to the right places at the right locations to make sense of the type of data that you have. And so if you have very high frequency kind of vibration data coming from a steam turbine that's coming in so thick and fast you cannot have that data being sent to a cloud system and process. It has to be done at the edge. And so we're absolutely looking at edge solutions to be able to do some initial analysis where it makes sense to be able to deal with the control systems to make changes live as things occur and then to send the right type of data back to the cloud to do the bigger value propositions and type things. The integration with bigger data sets like your weather data, like seismic data, because you really need big horsepower. The historical analytics. Exactly. We size at Wikibon. We size the industrial internet of things. Just around the time you guys launched, you know, the company, it's enormous. The numbers are just mind boggling. We don't even have to review that. But the other piece that we just released and this really relates to HP strategy of the intelligent edges, the cost of doing industrial internet with edge plus cloud versus just cloud is it's astronomically lower. So the economics of leaving the data there, doing the analysis and maybe moving 5% of the data back in most use cases versus 95% or 100% of the cloud, the economics is so much more compelling. Can you talk to that, Tripp? Yeah, and so we've been having a lot of joint customer conversations. So historically, GE and HPE are in a lot of the same customers. We've traditionally been working with, say the CIO side of the house in the data center and they've been out on the OT side or the operational technology side of the house in the operations and as IoT use cases are coming up, these groups are coming together a lot of times for the first time. Right. Not always even happily so. Indeed. So what happens is depending on the use case, there's an oil producer that's got a rig in the middle of the Gulf of Mexico. So if you think about the cost of trying to have connectivity to support a cloud infrastructure, that's not going to work, but we can leverage the compute power that HPE brings to the table, have rig IT literally running the Predix platform on site and then synchronize back to the cloud how and when it's necessary or it may not even be a cloud synchronization. It may be a true up synchronization that takes place depending on when people travel back and forth to the rig, to your point on the cost, thinking about a live satellite cost, which would be the only aura running a cable all the way across the ocean floor versus being able to use local IT as needed and then synchronize as necessary. Huge difference and it helps actually with the use case that's required to get the data processing done timely enough. Working with GE, which clearly is the leader in evangelizing industrial IoT right now, how is that informing HPE strategy? What have you learned from working with GE that it's going back into your product strategy? Yeah, and really looking, and we're even getting to the levels of working with the infrastructure designers at GE to actually do some customizations or create either chip level or board level capabilities to match up with the way the platform works. So we're getting a lot of insight into, as Richard was talking about, sort of the fidelity of the data, the type of processing that's necessary. You know, usually when you launch a product, we try to make a generic product that can serve a lot of different needs. But when you start to look at a use case where maybe I'm outfitting half a million elevators, every dollar that goes into that box means something. Because if I'm going to have half a million, I mean, excuse me, half a million devices on each elevator, I don't want a generic device. I want one that does exactly what I need it to do, but no more, no less because every dollar I add to that is an instant burden to the bottom line. So, you start to understand the use cases that are involved in how to support these different industries, and in taking that feedback back to our teams and making sure that we really co-innovate together and have the right products in the right place makes a big difference for the customer success. So my business partner, John Furrier, often co-host of theCUBE with me, hosted a panel with Jeffrey Immelt, and he loves to quote Jeffrey Immelt saying, I went to bed one night as the head of an industrial giant, and I woke up the head of a software company. And then I got Bill Rue, obviously couldn't be here. He has another great quote that I want to share with our audience. He said, we've already learned some powerful lessons. The first is that we must commit to changing GE's business model, to be driven to the business outcomes we aim to achieve. Second speed is everything, including the speed at which we learn, the speed at which we execute, and the speed at which we pivot. Sounds like a startup. For you guys to succeed in some of your ambitions of being a software company, do you not have to extend the Predix Cloud beyond just GE's products? Can you talk to that? Yeah, you absolutely do. So the way we've been building our product line has been very much based around the assets, the people, and the processes, regardless of who makes the equipment in the first place. So all of the analytics, all of the capabilities to ingest the data, are absolutely OEM agnostic. We'll take whatever device is out there, we'll connect to it, and we'll make sense out of it, we'll provide insights to it, to make sure that we are addressing the market needs, which is people don't want 10 different systems to interact with, they want one system that gives them the holistic picture for their entire enterprise. That's what Predix is about, that's what we're delivering. What are the tough, you're dealing with so many variables, harsh conditions, just so many things that you have to work with. What are the toughest problems you're working on right now? Some of the biggest problems we have is connectivity. So one of my potential customers is looking at how do we get connectivity to power plants in Siberia? These guys, they have a maintenance engineer go visit their site once a year. The connectivity solution to that, much the same as the oil fields, is satellite and maybe you lay a landline, but I mean the economics of doing that are so massive. How does that compare to the benefits of applying the solution in the first place? And so you've got to make some really good decisions about how much data you want to send back, coming back around to the edge processing. How much information and how much processing do I put at the edge to be able to solve those problems real time at the site, versus actually get it back to somewhere else and not deal with the economics of communication? Interesting business case, right? Is a once a year truck roll good enough or is the advantage of actually having the analytics in place offset that cost of connectivity? Rubik can help in a lot of cases, but maybe SpaceX can help in this case. There we go, we're just going to have a Tesla. But if you look at some of the strategy we've taken around our edge line products, actually having them harden to the point that they don't require to sit in a data center, you're able to have them out in the environment, you've got the multiple connectivity, communication capabilities available on the device, and the way they've been set up from a power and a heating standpoint, you can actually put them up on the wall in the factory because a lot of times these locations aren't going to have a data center, and if they do, it's probably a closet, right? So it's not, exactly. So it's not going to have the capabilities that you would want in a data center anyway. So having the ability to have the right level of compute at the edge, but also hardened for these environments is pretty critical. Yeah, exactly. Big DDoS attack in October, a lot of people freaked out, worried that their thermostats are going to be compromised. An even bigger issue I would imagine in an industrial corporation. How worried should we be? I think we need to be cognizant. I would steer away from worried because worried infers that we don't necessarily have a plan. We need to be cognizant of the realities of the world that we live in today and to take appropriate measures to be able to protect ourselves. So we invested in a company called WorldTech, their ace, an industrial cybersecurity leader. We've got cybersecurity products from our partners and our strategic technology partners that we're going to be integrating into our technology base to ensure we make sure that the infrastructure we have in place is protected, that the OT assets on site are protected, not just from the data transfer that's happening, but from intrusion into those sites. There are not many people in the market that can really offer those kind of capabilities because you have the combination of the domain experience and the IT know-how from the way that the intrusions are happening these days. I mean, potentially your physical security becomes more important because you have more at stake with that data on site. And just simple things like USB keys, okay? You know, on an oil rig, we had one example where the use of USB keys is very prevalent that the systems, the computer systems are very disparate around the setup itself. And so the guys, they would take data off of one system and transfer it to another, but they were finding they were getting viruses, malware, all these kind of things. So what we actually put in place was a sanitizing box which allowed the guys to introduce a new process. Next time you go to a PC, put the USB key in, hit the go button, it's sanitize the USB key and then you can do your data transfer. And the next time you want to do it, you repeat the process. Very, very simple security measure, but it's saved, you know, shut downs on rigs, it's saved virus, the decontamination you have to go through for all that. It's huge business to make sure that you get that right the first time. Let's talk about the organizational implications. We've got IT and we've got OT. What's, give us the update on those two worlds we really don't speak within organizations that much. How is that changing or is it changing? I see it is definitely changing when we know we're seeing the two worlds or they're kind of colliding together. We're starting to understand each other's language a little better and we're finding they're having kind of domain experts from either side kind of swap sides for a while to learn the different patches, helps that alliance, helps the communication paths. So we need to actually, you know, work out what that common language pattern is for the integration between those two worlds and we're making great strides towards that and we're absolutely relying on our technology partners to help us get there and, you know, we've been in this space a long while anyway so we're really trying to help that. One of the interesting things in the area, again, I like to get into the cool co-innovation area. So you mentioned WorldTech which really focuses on OT security. Can't say too much because it's still in the works but we've got some really cool stuff with the WorldTech team and our Aruba teams working together. If you think about WorldTech understanding the OT connectivity, those protocols and security, obviously with Aruba being a leader in wireless and the IT security that goes on top of that, if you had all of that working together with Predix at the same time, you create a converged device that allows the IT and the OT to come together much more easily. So again, bringing these together and then helping leverage our best capabilities to go ahead and solve these problems for the customers ahead of time also helps with that transition because we've already worked on bringing those worlds together. Is OT parlance harder for the IT guys to learn or is IT parlance harder for the OT guys to learn? You know, when you've been sort of armed deep in like Greece and something to try and change out, I don't know, a HPT turbine blade and understanding the problems that come with the fatigue and the sort of the metal and downstream distress, trying to explain that to somebody who understands servers and computers can be tricky. But it's just as hard the other way around. I've seen the struggle in both directions. So it is not easy either way. Any advice to companies who, I mean, every industrial company is going to have to wrestle with this with integrating these cultures, 30-year veteran engineers and software, upstarts, any advice on how to make that happen? The biggest advice I think is be open, be collaborative. You know, learn from your peers, understand how they're changing their game and don't try to be in that closed environment. The traditional industrial clothes, it's in my shop, not in my backyard, those kind of philosophies, that their day is done. We need to be in this open collaboration space working with partners to ensure that we are delivering the best technical solutions and understanding each other's domain. Only then are we going to succeed. Top takeaways. Let's see, you guys are three years in now since the launch? We're, yeah, three years in here. I mean, we've been working on it for about five years all up. It's amazing how fast things have progressed. Takeaways from the show and the context of this discussion. I think it's been sort of an eye-opener for some of the guys who've come around. They haven't necessarily realized quite what GE were doing and how we've been changing ourselves. So being able to impart some of that on the passers-by and the guys who've been learning a little bit about what we're about, that's been great. And also to hear in reverse the kind of problem spaces that maybe some of the traditional IT teams or the IT customers have been on site. The use cases they have are so similar to some of the use cases we're solving in an industrial space. The technology can only be a similar solution to some of that stuff. Tripp, final word. Yeah, I think we got a lot of good feedback on both sides. A little surprising for some folks to have an OT company here at HPE Discover. So I think that was a big eye-opener in a lot of ways. And then really talking to them about the fact that in this space there's really no one company that's going to be able to do it. So kind of going back to your other question about how do we make this happen? How do we keep it moving forward? I think companies looking at partners of leaders. So you've got leaders in the IT space, you've got leaders in the OT space who recognize neither one of us are going to be able to do this all together. So looking to those partnerships with folks in a leadership position to be able to help these companies traverse the challenges. And I think that's the feedback we've seen at the booth and when people have been coming back. I agree, absolutely. Excellent gentlemen, thanks very much for coming on theCUBE. Love to bring this topic to our audience. It's an exciting one, huge market opportunities, a lot of innovations. I really appreciate the insights. All right, keep it right there everybody. This theCUBE will be backed from Excel London Discover 2016, right after this pre-port.