 Hi everybody, Jeff Frick here with theCUBE. We are on the ground in San Ramon, California at the ever-growing GE software now, GE Digital Complex. I think they have over 1,200 people here. Every year we come out and visit Bill and the team and every year it just grows by hundreds and hundreds of people. So an impressive, impressive show for sure and we're really excited in this segment to be joined by John McGee, the CMO of GE Digital. Welcome John. Thank you, John, good to be here. Absolutely. So as this evolution of the GE software was just doing kind of software for internal things and then bigger things, and now you guys are really taking it up a notch to be, like you said before, really a significant piece of GE's business. Yeah, that's right. I think GE's investments over the last five years in digital technologies have really proven to bear fruit both for GE and for our customers. And this facility was set up to really bring together some of the best minds from Silicon Valley in software and cloud and analytics and accelerate our own digital industrial transformation. And now with our predicts platform and the formation of the GE digital business at GE, we're now working with customers on their own digital industrial transformation. I love it. So it's basically GE's take on all things that we see on the other side of the Bay in terms of cloud and big data and what's going on all the time in the IT. But GE kind of came at it from the OT, the operational side with the experience in locomotives and power and all those types of things. But we're seeing them come together and there's probably no better kind of vision of that than what you guys are doing with predicts. Yeah, that's right. I one of the things we learned early on is that general purpose business computing tools would only take you so far. And the industrial companies like GE and others have a set of unique industrial requirements in terms of how they, the types of data they work with, how they work with that data, the distributed nature of manufacturing plants and fields and assets that drive all kinds of productivity for our customers. And that collectively led us to say, we need to develop a unique platform for industrial internet of things or IoT solutions. And that's what led to the creation of the predicts platform. So predicts is great, cloud is great, but when you're talking about industrial things, you know, they're out in the field, there's horrific environmental climate that they're involved in, whether they're in a bottom of the ocean or in a big field. And then you also have latency issues because let's just face it, speed of light is too slow. So how do you apply kind of the benefits of cloud to the industrial internet and all those vast number of machines, devices that are out there on the edge? Right, I think the answer lies in the edge technologies that we and others are developing, some of our partners working with them. And the cloud is an essential piece of industrial computing. We've been clear about that from day one. When you think about the massive amounts of data coming off of these machines, this equipment, these assets, the number crunching horsepower that the cloud gives you, the economies of scale, the secure environment are essential to successful industrial IoT. But it's not enough. And we've seen that over and over where the requirements of industrial companies where you wanna make equipment smart, you wanna put compute right on a wind turbine or a jet engine or a locomotive. You wanna be able to aggregate data in a manufacturing plant and do things with it at that level without always having to send everything to the cloud. And what we've done with Predicts is make it possible to support those kinds of unique industrial requirements with what we call an edge to cloud continuum where wherever the Predicts system is, whether it's on the tiniest medical device or the largest locomotive, you've got the ability to run analytics at the edge as it were and to benefit from the cloud and to make that all work together for application development and security and so on. So how does the process go in terms of figuring out what should be computed at the edge? What should go back to the cloud? You've got transportation issues in terms of moving data. How much compute can you support? You have power issues. Again, if you're out in the middle of nowhere with something that does need to drive compute, the classic compute, store and networking always been a challenge to get those three things together. Cloud is interesting, but again, at the industrial internet, it's a whole different kind of challenge. It is, and we really see that you need an industrial internet operating system that can live at all these different nodes, and you touched on many of the scenarios that are driving these kinds of application architectures. Sometimes it's you're talking about an asset like a locomotive or a jet engine that is moving, that you wanna be able to do compute on board as it were and then send the data from those operations to the cloud to the rest of the system later or at a point in time when it makes more sense, or you wanna be able to do an analytic on a piece of equipment and figure out if there's a problem, 99% of the time, there is no problem, but that 1%, maybe that's the data that gets forwarded to do something about it operationally. And then you mentioned latency. A lot of industrial control systems operate as real time and near real time systems, and you can't be to your point about the speed of light. You have to be close enough with the compute solutions to interoperate and integrate with those types of control systems. And so what we've really got is an end to end cloud based operating system that also supports the edge nodes. So you say end to end, but there's a lot of stuff that's already out there, right? There's a lot of companies that have infrastructure that's already placed in to play. How do you play with the bigger ecosystem? Clearly in the software world, ecosystems are really, really important. No one company can do everything. So how is GE engaging the ecosystem? How are you bringing other players in, both on the IT side as well as the OT side to help make this work for the customer? Yeah, we've got a large and growing partner ecosystem that includes ISVs, technology partners, systems integrators, and they're all essential to helping our customers achieve successful outcomes with these new technologies. And in particular, when you're thinking about the edge, there's a lot of existing industrial systems out there in the same way that enterprise IT and CIOs have had to deal with legacy software systems going back to mainframes and client server and all the generations of technology that they're still stuck with and have to integrate. We have to think about all those situations in the industrial world where you've got control systems, you've got other types of industrial processing software that's been there for many, many years, still working successfully. But what's happening now is companies don't just want to be able to turn things on and off and automate robots and so on. They now want to get data out of those systems so they can understand better, use analytics to be more predictive, be more efficient about their operations. And so that requires that interoperability between systems like predicts and those traditional industrial legacy systems. And where do things like, traditionally, it was often standard bodies that would define the way that these systems potentially would communicate. And now we've seen, obviously you've been in the software business a long time, open source is really driving a lot of kind of standards. How do you see that kind of evolving for the industrial internet and the predicts cloud specifically? Sure, I think there's opportunity for more standardization. GE was one of the founding members of the Industrial Internet Consortium, or IC, as a lot of people call it now, along with IBM, Cisco, AT&T, and a couple others. And through that group, they're doing a lot of test cases and real world scenarios so that we can identify collectively as an industry what are the standards that can help support interoperability and so on. It's still early stages for IoT in general and industrial IoT, but there's a lot of promise there to get more consistency across those different systems. Okay, and then how do people get started? Like I just imagine, I've got a big giant plant, I'm making a bunch of widgets, it's been running widgets since my grandfather started the business, and I want to get it up to speed and take advantage of some of these new analytics and tools, how does somebody get started? Because it's a big operating system that's running and kicking stuff out the back door, loading it on pallets. How do people get started and take advantage of this? Yeah, that's right, it's a great question. In fact, it's probably the question when we work with customers and they like the idea of this digital industrial transformation journey that GE's been on, they want to learn more about what we've learned, they want to understand the technology, but it always comes back to, okay, where do I start? Where do I get the ROI for my investment in technology? Where do I get the best bang for my buck? And that tends to look different, obviously, depending on the industry, the customer and their requirements. Sometimes it's being more predictive about spotting problems in your assets or critical infrastructure so that you can fix them before they occur, as it were. So the idea of predictive maintenance and analytics around asset uptime and improving uptime, because for many of our customers, the uptime of these systems, whether you're talking about hospital medical equipment or you're talking about transportation systems, airlines and so on, uptime and the availability of those systems is critical and downtime is enormously expensive. So in those kinds of investments are often a starting point. Sometimes it's operational efficiencies like getting the right information to decision makers who are either technicians that are doing maintenance or adjusting the way a manufacturing plan is operating, making changes that way. I think in general, when you think about all the productivity gains that software technologies have given all of us who sit in front of desks most of the day, you think about workers in an industrial sector, they've arguably been underserved by gains in those kinds of productivity tools and there's a lot of opportunity through mobile computing solutions, through new kinds of applications to deliver just in time information in the context of what a nurse in a hospital is doing or in the context of what a maintenance worker on a jet engine is doing and getting that kind of information in real time, not at a desk they have to go sit at but in new ways because they're performing a job while they're doing this, there's a lot of opportunity for those kinds of gains. It's great you bring that up because there's a lot of conversation not so much with the industrial net but just the machines are coming, right? They're gonna come take all our jobs, they're gonna drive us to work but the scenario that you just described is really giving people the benefit of a lot more data information to make better decisions. It's not a substitute for, it's really enhancement too in ways that they couldn't do it before. Yeah, absolutely and so many customers with industrial operations that they are starting to collect data but that data is being warehouse stored essentially and the insights that are possible to glean from that data are not getting back to the people who need it fast enough to make it actionable and so shortening that cycle time of getting the data, getting the analytics to analyze that data, identifying the right piece of information that can help someone make a better decision and that someone could be an executive responsible for a whole set of power generation facilities who's making the call on whether they wanna open another facility or whether they wanna push their infrastructure harder to be able to sell electricity at the spot market price that next day these kinds of big decisions or they might be as simple as helping an operator run a particular piece of equipment more effectively because you give them the right information about how to do that based on best practices that the data knows from all the different types of instances of that asset. It's a great point and we hear about this a lot come in the insurance business where the insurance business has been based on an aggregated risk pool and then you take averages but now with FitBits and this and that people can start to optimize at the individual level you've talked about and G's got the digital the digital was the digital twin such an interesting concept that yes all those jet engines came off the line at the same time but one got shipped to Southwest operating out of Seattle and the other one got shipped to Dubai and it's operating out of the desert to be able to really start to get a different set of data and perspective at the individual machine total game changer. Yeah, that's right. The digital twin is a powerful construct and it's at the heart of the predicts platform in the sense that the twin is the digital representation of all the physical assets in your environment that you're trying to optimize and just as you said, there's opportunities to optimize an individual piece of equipment like an individual jet engine or locomotive but it gets really interesting when you start to collect historical data across all of the jet engines all the locomotives then you can start to spot patterns your the answers to the questions you're asking can get much better. So the digital twin and the ability to collect all that data not just about what is the manufactured spec for that piece of equipment look like but what have been its performance requirements? How has it been run in a real world environment? How hot did it get? How cold was it subjected to? And all of that data becomes part of this the unique digital fingerprint of a digital twin. So now you can do things like optimize just for that asset it's like you and I both drive our cars very differently but if we look in our manuals it's gonna tell us to change the oil every certain amount of miles. Industrial equipment much more complex much more critical for our global infrastructure and we need to be a lot more sophisticated and the digital twin lets us get to that level of sophistication where each piece of equipment that we're relying on gets treated by like the unique snowflake that it is. All right so I'm gonna shift gears for you let you go and again I appreciate you taking your time out of your busy day. You've been in the software business a while you've seen a lot of change and now we're at this kind of this new thing this whole internet of things which you know the industrial internet is a piece of and G and predicts is a piece of that. In terms of your perspective kind of sitting back a little bit have a glass of wine on a Friday night you know how do you see this opportunity and specifically the way that GE's been able to execute on it because I think as you mentioned in an old interview they're one of the original Dow Jones 18 stocks and here we are in a 1200, 1500 people in the middle of Silicon Valley basically re-envigorating this company and we see the ads on TV the guys were gonna work for GE no I'm not working on fact I'm gonna make software well that's not a software company so I'd just love to get kind of your perspective. Sure yep so I joined GE as part of this industrial internet initiative a few years back probably the most exciting thing that I had heard about in my career and I was thrilled to be able to get in at the beginning of the process and I think as you mentioned GE as a history of innovation as a history of entering new markets at the right time and for us right now this whole area of the industrial internet is that big shift in focus so we've been building teams out of some of the best minds in analytics and machine learning in cloud and architecture and development and really trying to bring the best practices in the general purpose computing world into the industrial world optimize it for the unique requirements of these industrial customers and you know there's been a lot of hype around internet of things and I think when you think about the industrial subset of the broader internet of things that's really where the value it's very apparent where you can get value right away whether it's improved uptime new business models new ways of operating giving information to the right people for faster decision-making so I think this guy's the limit in terms of the opportunity and the potential for the industrial internet. Signing times. That's why we like coming up to San Ramon so John thanks again for taking time out of your day I'm Jeff Frick you're watching theCUBE we'll catch you next time, thanks.