 Live from the campus of MIT in Cambridge, Massachusetts, it's theCUBE, covering the MIT Chief Data Officer and the Information Quality Symposium. Now, here are your hosts, George Gilbert and Paul Gillan. Welcome back, this is theCUBE. Paul Gillan here with George Gilbert at the MIT CDO IQ Symposium. We are coming to the end of our first day of live streaming coverage. And our guests, two guests joining us now, Steve Oren, Chief Technologist at Intel Federal LLC and Dr. Joseph Salvo, Director of Industrial Internet Consortium and Manager of Complex Systems Engineering Laboratory at the mouthful at GE Global Research. This is an industrial topic that they were addressing today in their session, talking about value chain, the value of data across the chain and particularly in industrial context. I want to ask you, Steve, first about your role at Intel Federal, what is the role of a technologist at Intel Federal? So basically my role is to help drive technology adoption and to help advise the federal government on the use of technology and to help build solutions with the federal government that allows them to execute their mission goals, leveraging technologies from Intel as well as the full solution stack on top. So working with the ecosystem, both on the system integrator side as well as the OEMs and the solution providers to build solution stacks that meet the government needs. How do people think of Intel if they think of microprocessors or what are you doing at a data conference? So at the end of the day, you need microprocessors to process that data and really the only way to be successful in the data environment is to look at the pain points of that data processing. So Intel is focused on analytics and both from the hardware side all the way up through networking, all the way to the analytics, platforms and the algorithms because we want to optimize the data experience. We want you to be able to get your answers quicker, you want to be able to consume the large amounts of data faster and so working together with the ecosystem and with the actual in customers that need to be able to process the data understanding the workloads is the only way we all are going to be successful. So Dr. Salvo, tell us, we were talking earlier that GE uses these examples of if we could just save 1% of fuel cost or resource cost and it's a very, very large number. Give us two examples, one where you think it's a reasonable example of what you might save in resource cost and then a sort of a business transformation example. Yeah, so I think one of the things that comes along with becoming a digital industrial is the recognition of the value in the network and GE has actually always been a network company. If you go back to our founder Thomas Edison built the first electrical grid in the United States in Lower Manhattan, late 1800s, we have huge transportation networks, power networks, healthcare networks. Today, because of Metcalfe's law and the increasing computing access that people have, we can take advantage of that like never before. So GE is now focusing and getting more of the value from the network to the customer. So for example, wind farm that buys many wind turbines rather than looking them as a single energy asset, we can optimize the wind farm as a whole and extract 5, 10% more value over the length of those assets because they're networked and because we can optimize them in real time. The other thing that we do currently within GE and for some of our partners we optimize the supply chain networks. GE works with thousands and thousands of suppliers and partners, so we can use these same networking technologies to optimize our supply chain networks, our logistics networks to deliver new products and services much, much faster than ever before. Would just to continue on that thread, thinking of an analogy where some of the autonomous driving examples are fleets of trucks or convoys where one slips strings behind the other. Would it be that you're more densely packing what moves along the supply chain or is the coordination a difference in quality, something that you could not do before? I'll give you a very concrete example, very similar to what you just described. GE builds some of the best locomotives in the world that haul all kinds of freight across the United States. If we just increase the average speed by a few miles per hour, that returns tens of millions of dollars in value across the network. And we do that by knowing where the trains are, knowing how fast they can accelerate, how fast they can slow down. We can automate the actual driving conditions for the locomotive. So that's just one example of how we can use automation and network thinking to create enormous amounts of value. One of the things, I'd like to ask you both this question about changing mindset. Now GE is often thought of as kind of the poster child of a company transitioning into a more digital business with great success. Intel certainly interested in getting its customers to think differently about how they manage data. How do you get people past this operational focus, this operational mindset of using data, just to streamline the business, get them into the thinking about data to actually expand the business and find new business opportunities? Have you learned any tricks for getting people to think differently? I think some of the tricks we've seen in the industry is where you help them, number one, you help them solve problems that they're not able to solve with their normal operations today. You pick a problem to go after. You don't want to try to change their entire IT infrastructure overnight. That's just too hard to do. So you pick a problem that's near and dear to the company. And then you show them the power of what analytics can do, whether it be predictive models, being able to do predictive maintenance is one of the really good ones that a lot of people see, the value orchestration, resource optimization logistics is another area. And one of the areas we've had a lot of success in is cyber threat intelligence. So using analytics to solve a key problem that the current infrastructure isn't able to solve, be able to show the value, show the return on investment for using the analytics solution, helps them get them in the door, get the infrastructure in place, and then they can start applying that the know-how that they have, whether it be the internal on-prem infrastructure, a cloud-based architecture to other problems. That's sort of the gateway into getting into analytics if they haven't bought in already. How does that process work to GE? Well, in a network economy, in a global world, everything is not deterministic. We can't predict everything perfectly. So you have to be able to make decisions in real time. So using the analytics on top of our platform with real-time data feeds, you can start to optimize in real-time and not necessarily automatically, but you give options to the owner of the asset pool. And he can make strategic decisions or tactical decisions based on his need at the moment and his competitive circle. So whether it's a train or a wind farm or a hospital with all their healthcare assets, they can make moment-to-moment decisions on the best way to deploy those assets. Have you, actually, have either if you started to think about empowering your customers to be B2B or B2C information companies? In other words, where it's not just the product, but it's the information and the analytics so that they develop new product lines that are information-based. Well, I think once our customers have a better handle on how they can modify the operation and the units, for example, the rail industry lost a lot of business to trucking. Because trucking was able to more clearly define delivery points, so perishable goods, goods that had to be received in a specific window, a lot of them moved to the trucking industry. The more we can make the rail industry predictable, faster, visible, transparent, it's gonna open up more business opportunities for the rail industry. Wind farms, the same thing, they wanna maximize their return on investment so anything they can do to anticipate load needs, distributed load to other types of systems, all the benefits. So the bigger the networks you can make, the more value opportunities you're gonna get. And I think if you look at the federal government, in some respects, they've always been, in many agencies, you've always been about the data, whether it be IRS or Census, as well as on the intelligence side. And the key is being able to turn those data assets into serviceable, whether it be services to be able to consume them throughout the organization. In the consumer's world, you say, you wanna be able to monetize your data. In the government, you wanna be able to turn to something actionable. And so we're seeing a lot of government agencies using analytics and using these models and capabilities. And the next step is to try to turn that into a service, whether it be to enable their own mission or to help enable other missions. And on some of the standouts, NASA's been a really good example of doing that because they have the charter to share data. The IRS is another one because of the fraud detection capabilities that they have. So when you can turn your data assets into services, it's really the corollary for the government on how do you monetize and do the B to C and B to B for data in the federal space. You actually segwayed into something I wanted to ask about, which was data has greater value when you can join it with other data for richer and richer context up in the cloud, particularly as opposed to the edge-based data. Under what circumstances can you encourage customers or can you do it on their behalf? To bring in more data to have a richer context or maybe in the case of the federal government, often you can't share or industrial customers might be sensitive. What we have not had much trouble in getting people to share once the value story is clear. There's all types of data sets that create value at different hierarchical levels. There's value created at the object itself, groups of objects and system level value. We can look at the manufacturing sector, for example. There's been a focus in GE, we have over 400 factories to create brilliant factories. Well, we can take the data that we get in the factory, pin it to each component, every product that we make, rather than just looking at statistical norms of collections of products. Then you can do upgrades, system designs. You can redesign products based on your knowledge of every single component and the characteristics that went with that component rather than the overall average. It changes the way you look at your customers' operations, how you look at servicing the equipment that's out in the field. Remember, we have a trillion dollars worth of assets out in the field that we have sold or are monitoring or are servicing. So that is a transformational way to look at each product as an individual, not just as one of a collection. I think if you look at it from the federal perspective, first of all, there's different kinds of cloud. It doesn't mean that they have to share all their data into some public cloud. There are lots of clouds that are built for federal, that are FedRAMP certified to the security level. And so that's where you're going to see the data collection and the aggregation and these data analytics pools come together is in these government specific clouds. But part of it really comes down to the use case where it makes sense to share whether it be something like threat intelligence, logistics, resource, I mean they will share. And especially across agency, they're getting over those hurdles. A lot of government agencies have sort of spent the last couple of years trying to figure out the mechanisms to enable the sharing, whether it be for cyber threat intelligence sharing, whether it be just for resource optimization. When you look at sort of the VA, the defense health, the broader agencies, HHS of how do you manage healthcare across the DOD and into the VA, they need to be able to do data sharing across that. So they're figuring this out and they've got the cloud infrastructures that are helping them do it without having to necessarily take everything all the way out to a true public cloud. Of course, we're out of time. We'd love to get into the Industrial Internet Consortium, Internet of Things, all the data issues that that presents, but we are out of time. Thank you so much for spending a few minutes with us here. Steve Oren, Chief Technologist Intel Federal and Dr. Joseph Salvo, Director of the Internet, Industrial Internet Consortium at GE. This is Paul Gillan, this is theCUBE. We're going to wrap up our first day's streaming coverage of the MIT CDOIQ Forum next. We'll see you back here in a minute. Hi, this is Christa Vanny from...