 Hi everybody we're back, this is Dave Vellante of Wikibon.org, we're here in San Francisco with the GE Industrial Internet Internet of Things, an industrial cloud event here in San Francisco, I'm here with David Floyer who's the CTO of Wikibon. Earlier this week Wikibon released under David Floyer's direction along with Jeff Kelly a new study defining and sizing this so-called industrial internet, so David thanks for stopping by the cube and sharing your thoughts with us. You're very welcome. So tell me about this study, sort of what prompted it, what's it all about and we'll get into what you found. Well we were interested in the Internet of Things, what was going to happen after the commercial, sorry the consumer internet. That was driven entirely, pretty well entirely by advertisements. Google's making all its money from advertisements in that space and we were looking ahead to see what were the other big data areas and the other big data areas are things like surveillance obviously and the industrial internet and the industrial internet is pretty small at the moment but it appeared to us that it really, the quality of the data that came out of these machines was much much higher than the quality of data that's coming out of the consumer internet. So is that the big defining difference between the internet of things slash consumer internet and the industrial internet? I guess the internet of things and the industrial internet are more closely aligned than the consumer internet but it's really the difference in terms of the use case one being sort of ad-driven is the revenue model. What's the model for the industrial internet? What's driving value there? What's driving the value there is the ability to improve efficiency and reduce risk. So you can improve efficiency on a local level with lots of data and get feedback about that particular wind turbine if you've got local processing power but you can improve efficiency across a set of wind turbines for example. How hard you drive them? There's a cost to driving things hard but it may be worth it if you're going to make a lot of money because you have a power outage in some other plant. These sort of trade-offs are done at the moment by seat of the pants, you know rough calculations. Gut feel? Absolutely gut feel very very valuable but if you can do the math and if you can do the analytics in real time and adjust things on the fly you have tremendous opportunity for efficiency across things and then even more than that you can have efficiency across a country. If you can avoid the brownouts that we've had in California and New York recently, those to society are worth a huge amount, absolutely huge amount of reduced risk and inconvenience. So bringing that up to a higher level to another network again is huge potential opportunities for ways of making money. So you said it's relatively small today but the numbers that you put forth for 2012 or about 20 billion for the industrial internet is that right? That's right. That's sound pretty sizable. It's not must not. But when you look at the total cost of all the equipment out there, what's included in that is the sensors, the equipment, local equipment, the network, any cloud activity. There's not a lot of that at the moment. And also in the services and the software and the people are running it. So it's the total cost of... Total spend. Spend, sorry, on the processes there. And the return on it at the moment because like anything new, the returns aren't going to be fantastic at the beginning. Those are about 23 billion we've estimated. Well so let's unpack that a little bit. You talk about the spend on the infrastructure for the industrial internet is about 20 billion. You're talking about the value created as a result of that spend is 23 billion. In other words, the savings in productivity and risk reduction and elimination of downtime and the like is 23 billion on 20. So relatively small value proposition today. We're not excited too much. We're not excited too much. So what's that look like by the end of the decade? Well by the end of the decade, we think that the spend is going to be around half a trillion, $500 billion. That's a pretty enormous increase. And that's going to be fueled by the establishment of platforms, the establishment of protocols, collaboration agreements between people. That will fuel the potential of a huge amount of added value. And that added value again, we think by the year 2020 will be about $1.3 trillion. Okay, so substantially higher ROI. Much higher, 150% ROI. So what's holding things back today? What are the headwinds? Is it technology? Is it people? Is it process? Is it just maturity? Is it cultural? It's all of those things that are holding it back. And if you look at how industry in general rides on the initial consumer-led, any sort of process, if you look how many years it took for Flash to come out into the enterprise, it was two years, more than that, more like three years before it was at the equivalent level. And the same thing is happening with the industrial internet. It takes even longer because you've got very well-established ways of doing things. You've got paper procedures and replacing paper procedures which people understand, they know they tick the boxes, they do the checklists, replacing those with a computer that's in real time going to decide what to do. That's a big step. Especially for an engineer who's got a system that works really well. Absolutely. If it ain't broke. Don't fix it. Yes. And that's a big step, cultural step. It's a big reliability step. And there are some huge issues in how do you test systems like that? How do you avoid the 0.001% probability that instead of making a smart decision is going to make a catastrophic error? The machine. The machine. Yes. And those sort of issues about developing the software, testing the software, getting people comfortable with the software are going to be some of the big challenges over the next decade. That's a maturity issue. We had Beth Comstock on earlier, the CMO of Gene. She talked about what's next for us as proof points. Yeah. And so we need to see some of those. What about data standards and the like? Those are in some industries, those are quite plentiful. But in other industries, they're not. So it depends. They've got the pack system, for example, which are extremely good standards in healthcare. But in the rest of the system, they're still to be worked on. Excellent. All right, David Wilson, thanks for stopping by. We'll pick this up a little later. And Bill Rue is here. He's going to come on next and share with you the vision that GE has for the industrial internet and the industrial cloud. So keep it right there. Thanks, David. Appreciate it. Appreciate it. Very welcome. Keep it right there. Everybody right back with our next guest right after this. This is theCUBE. We're live in San Francisco.