With intelligence migrating further and further into devices and systems all around us, what does an endlessly connected world look like? Can the intelligent connection of people, processes, data and things create its own economy? Will it be open and accessible to innovators? How does connecting everything change the role of people, and of existing institutions?
Read the full transcript below. (Transcript by Realtime Transcriptions.)
Kirkpatrick: This next session is an all-star discussion on what happens when intelligence spreads much more broadly in the world and everything gets much more interconnected. It will be led by Jon Bruner of O’Reilly Media. So, Jon, take it away.
Bruner: Thanks, David. I’d like to thank David and Simone for putting together such an outstanding panel. The people in front of you have a lot of really interesting thoughts, and they’re doing some fantastic work that has to do with bringing software intelligence into the physical world. The lines between real and virtual are becoming blurred, and I think we’ll see them get even more blurred over the next few years as the sort of work that the panelists are doing gets extended and made commonplace.
To my left here, we have Carlos Dominguez, who is from Cisco; then Dmitry Grishin, who is the head of Mail.Ru and also has an investment firm called Grishin Robotics; Alex Hawkinson, SmartThings; and Bill Ruh from General Electric.
So I will have each of you’ve say a few words about what you do and also answer the question why now.
Dominguez: I’m Carlos Dominguez with Cisco, as Jon mentioned. I get to come to a lot of conferences and talk. So I do that primarily.
Why now. I think it’s a perfect storm, Jon, the convergence of a lot of things. Number one, is the processing powers available, the pricing is available now; and I also believe that the networks are there, to be able to start supporting some of the connections. We worked through a lot of the protocol issues. So we are starting to see the evolution of these things, and still with a lot of deterministic things that we need to tackle, but everything is available. It is a perfect storm.
Bruner: So the groundwork has been laid?
Dominguez: Yes, groundwork is laid and now it’s taking it to the next level.
Grishin: Yeah, I’m Dmitry. I’m one of the founders of Mail.Ru, the biggest Russian company, which are now public. And also, one year ago I launch investment company Grishin Robotics. We have planned to invest $95 million to consumer robotics companies and Internet of things.
Right now I think I agree with Carlos. It’s—price is going significantly down. And I think key reason for it is it’s funded by evolution, because most of the companies which you have right now in smartphone, I also use it for Internet of things and for robotics companies. So you can buy right now HG camera for $3. Ten years ago, this camera cost I don’t know how much. $10,000. So prices is going down significantly.
The second important thing is that introducing of technology like 3D printers allow to do prototyping much, much quicker. And I think especially for hardware companies, prototyping is part of the job which are really tough and very long. Right now with 3D printers, you can do much, much configurations.
And of course creating platforms like Kickstarter, which helps companies and startups to enter, especially for hardware, raise funding. But I think, more important, to understand what the people think and how people react to your products and make some market evaluation.
So I think these three key components make me believe that for us to—we make huge evolution in software, in Internet. And next will be huge revolution in hardware, where you’ll use all power of Internet and extend it to real world, which my feeling have not too much innovation for last several decades and we need to change it.
Bruner: You were telling me yesterday that adding something like a sensor to something that’s never had a sensor on it is just a couple of dollars, right?
Grishin: Yeah, right. For example, if you can look on this room, we can put sensor to each chair, which will give the information each person sitting in the chair or not. And for example, you can get realtime information how many people here in this room, how much power you need to put here for air-condition, for example. You can also understand does session interesting on not, if people leaving or not leaving. So it’s—a few dollars, but you change price of the chair too much, just a few dollars, but you really create a lot of application to understand what’s happening. This is very powerful.
Alex, who are you and why now?
Hawkinson: Alex Hawkinson, founder and CEO of SmartThings. We’re trying to bring the Internet of things and your home to the regular consumers and put your home in the palm of your hand on your smartphone.
So notable for a few different reasons. Trying to make it easy and accessible and also we’re an open platform, so some number of thousands of developers and device-makers are building new interconnected things that work with us and what we are doing.
I think I would agree with all the other comments on underlying trends that are supporting it. I think the number-one factor in why this is happening now is the smartphone revolution, is my own hypothesis. I think that even with the underlying enabling technologies, really it’s a psychology change in consumers and humans, where you expect the world to be hyper-connected and with you at all points.
And of course, the underlying benefits of connecting up the physical world, as Dmitry was beginning to say, are profound and low cost and work because of the wireless networking technologies and so on.
But it’s the hyper-connected consumer that’s start of demanding it now.
Bruner: And you are building a lot of modules that people can put together with kind of moderate expertise, right?
Hawkinson: Yeah, so we try to make it easy for anybody with a smartphone. So we sort of jokingly internally say $200 and ten minutes is what you need to begin bringing your everyday things in your home online, like your front door, let’s say. And then once they are connected, simple applications, you can really change life to make it more secure, safer, more fun and smart in different ways.
Ruh: I’m Bill Ruh. I’m responsible for GE’s Global Software Center. I think at the core of GE, we make our money and are known for large machines like gas turbines, wind turbines, jet aircraft engines, CT scanners, and we sell the products and services around that.
Well, what we see happening is everything that’s been talked about, low-cost sensors, low-cost communication. In fact, we look at the consumer Internet and that has just driven such change that we think now the technology prices are such a point we can drive it into the industrial world.
So we’re really—where Alex talks about the hyper-connected consumer, we think about the hyper-connected plant or the hyper-connected machine and we see how is that going to change our world. And when we look at our opportunity versus the consumer, I saw a Bank of America report recently that said about $32 trillion a year is spent in the operations in the industrial businesses we serve.
So we look at the idea, if we could make a 1% change, just in that it’s a huge opportunity for us and for others to participate and change the industrial world the way the consumer world has changed.
Bruner: A lot of people have noticed software is a terrific business to be in. Does this mean that GE is getting out of the industrial business and switching into the software business?
Ruh: You know, we get this question all the time. I think the way our chairman and CEO, Jeff Immelt, says, he’s says he’s come to the realization that every business has to embrace software and analytics in a way that they haven’t before. They have to make it a core competency.
However, that doesn’t mean you’re a software business. I think to me the greatest software businesses in the world don’t sell software. I think of Google and advertising and Amazon with retail and so on.
So what we see is not so much the idea that we are a software business, but we realize that our services business—and by the way, I think this is all services businesses, are going to be foundationally changed through this idea of sensors, data, analytics, and how we enable people to be more productive.
So for us, the cornerstone is really not being a software company, but extending our services to go to software-enabled services, and we think that you have to make it a core competency.
Hawkinson: One comment for this audience is that I think this is very natural at some level. I think of this as sort of the third phase of the Internet. You had 15 years ago, sort of information started coming online, it predated Google, and now the knowledge graph is there, what is the impact of that on society, from how you discover answers to problems to finding businesses that you want to intersect with, and it’s just changed so much. And of course offshoots to that is the cash flow changing the software landscape as well.
People are online. You have the social graph. What does that mean for elections, as an example, the topic just before this? And now what’s the next thing to be engulfed by the Internet? There’s the information, there’s the people and now it’s the things. And the physical graph has equal implications, if not more profound ones, about digitization and software and what it’s going to do to the everyday world.
Dominguez: The powerful thing, Alex, I agree with you—I have a very good friend Rick Smalley who wrote a book. What he said as he was describing the fact that things are connecting, they are all generating data and we are gaining insight, he was trying to describe the benefits of that to like a 10-year-old child. And what he said was—and I love this description. He says, “Imagine if all of us, we were walking through the world with one eye open.” That’s the way we all see the world. “And all of sudden something’s going to come to us that’s going to give us some insight or wisdom that for the first time all of us are going to be able to see the world with two eyes.”
And when you think about what—this Internet of everything, Internet of things, the industrial Internet, or whatever we want to call it, the fact that things are connecting, they’re sensing stuff, they’re generating data, that data is being processed, analysis are being done, and whoever is the recipient of that is going to have some knowledge or insight that they didn’t have before and ultimately change an outcome. That’s why this thing is the here and the now.
Bruner: The way I have often thought of this is as you think of something like Amazon.com and how heavily instrumented Amazon.com’s web site is and how much value they get from that instrumentation, how much intelligence they can use to guide their business and you take that sort of instrumentation and intelligence and apply it to the physical world instead and you get a simple kind of intelligence out of it, similar kind of ability to make really good decisions.
Hawkinson: Maybe some classical examples might be helpful for people. But it’s also happening because the benefits—the intersection and the technology point is the benefits far outweigh the cost of doing it now at this moment, from friction and the economic costs and so on, but if you—it’s not just—starts—there’s individual problems, but it adds value in the consumer space and you can give great industrial.
You mentioned kids as an example. Two kids die in the U.S. every day, toddlers, that consume household chemicals that they shouldn’t have, right? And a tiny sensor on a cabinet can make a huge difference in the lives of the emergency room every day and so on as well.
So these little life problems can be profoundly impacted with a free solution effectively, but then you magnify it up and you say, okay, well, what about turning things off automatically.
You mentioned the air-conditioning in this room, when people aren’t present. In households, that’s 30% of energy use can be impacted by lighting appliances, HVAC systems, just by invisibly make it operate better when you are not there, not impacting the way you feel about it in the near term.
So these changes—and I could go through security safety insurance, crime, health, other stuff—they are really profound in so many facets of life that that’s the sort of Cambrian explosion we are about to see, the enablement of all those benefits.
Grishin: Just thinking about where we can go in the direction, and my vision is that you have two stages. First stage you have right now, we start to sensor things in the physical world. Sensors will be everywhere. You can make video. You can get some sensor information of temperature. And you collect this data and understand this and try to make some application based on this data.
And of course the number of sensors and places we can put it is growing exponentially, but I think much more exciting for me, like next stage, when you can not only sensor the world, you can change the world.
So to me you have not only like understand if somebody are sitting on the chair, you can move this chair, so you really can change physical things. And this is the direction I think is just create absolutely new amount of applications, which it’s very difficult to imagine right now. And I’m really excited when you can apply all this power to changing physical world.
And the other thing which I want to mention, I think that for pretty much for all entrepreneurs, it’s very obvious that you can go and start Internet company right now. You can have three friends or cofounders and start an Internet company. But it is not very obvious that you can go and start a robotics company, which now is real situation. You can have like three or five people, like $100,000 initial funding, and with a lot of building blocks, a lot of this hardware available, I think we’ll have big new wave of hardware startup. And I hope it will be bigger and bigger. In my belief, the biggest innovation will come not from big companies; it will come from new startups, which will try to leverage this new Internet of things.
Bruner: One of the core tenets of this discussion is it is getting easier and easier to start building hardware. The prototyping tools are available. The Chinese supply chain is available. Even the highly automated American supply chain is coming back and is really competitive again for developing things in small runs and starting to develop new products.
I want to hear Alex and Bill both comment on what sort of expertise do you need to start building hardware and intelligent physical devices? You come from very different backgrounds on this.
Ruh: We break the world in the consumer and the industrial world, and I think the enabling technologies are much the same. And by the way, we see this idea of free, open-source hardware and software as sort of—and the cost of things coming down as the enablers, but the real value is in the applications that you build on it, and not the technology. Technology’s interesting, but just the enabler.
When we look at the industrial, we see a lot of people building applications in the consumer world. I would say everybody has an experience there. Everyone has domain knowledge, so anybody can get into it. But when you look at the industrial side, what’s been the key barrier is that domain knowledge about how an industrial power plant works or a jet aircraft engine. And the regulatory environment creates a domain knowledge that’s much more barrier that fewer people have. So we think that in the industrial side of things, the technology is the enabler, but it’s nowhere near enough, the domain knowledge. And those who have that domain knowledge are going to be at the center point of driving those things.
Because in the end, there’s really two things customers want in this. One, is almost all of them are driven by fuel and carbon. If I can reduce fuel consumption and carbon, it’s almost always the number-one and number-two costs in these industries.
And number two, is if I can stop things from breaking—and we all know what it’s like to have a flight canceled. About 41% of all cancellations are mechanical. If I can get rid of that in these industries, that’s huge. But you really have to understand the domain, understand the physics, understand the operations.
So we think the barrier to entry in the industrial side to domain knowledge is huge, where the barrier on the consumer side is a much lower bar, but the technology, for both of us, I think is pretty much the same enablers.
Hawkinson: I wake up some mornings—the idea for SmartThings was over Christmas 2011. It’s 22 months ago or something. I was totally a cloud software guy before that. And there’s many mornings I wake up and I’m thinking, What the hell? How did I get here?
But the reality is, again, a lot of these underlying feeder trends, mature networking standards, to add connectivity, robotics-based manufacturing—if you go into one of these plants and you haven’t been, you see the—you realize the rise of the machines like lasering the pieces together and stuff, and you’re just like, Wow. You load in your CAD design and out pops your prototype device. And 3D printing and things like that to get the enclosures right.
So we actually did all of our first 10,000 households, all the manufacturing in Michigan, in the U.S., because the—for agilely moving through the process, it was so quick to iterate our various hardware components from our hub to our devices. We had to find some domain expertise. We had our own secret methodology for going and recruiting some electrical engineers in the business that were great.
But taught from start of looking at it as a technology person to having our first working prototypes four to five months, certified within ten months and shipping within twelve months or so. And these are pretty sophisticated things at some level.
What we realized, though, is the hardware is just such a thing layer. It’s sort of the enabler of the software opportunity, from our standpoint. So we enable it to break down—we further break down the barriers to other—it’s going to accelerate other hardware projects that don’t need to build as many pieces that we did and, of course, developers that can use their regular developer skill sets to build apps for the physical world, now that we have created this underlying enabler.
So I don’t know, there’s a GE device, there’s a light switch, we have given it—every GE light switch an API, right? And who knew that Double Tap was an app that the developer ecosystem would write, that when you double tap your light switch, it actually means something different. You can configure it in your household so that when you double tap your master bedroom light switch at night, it means you are going to bed and it locks the doors and shuts off the lights.
It opens up the physical—it’s both easier to build hardware, but there’s just an avalanche of developers that are just now getting access to this opportunity space that they haven’t been able to go after before.
Think of your mobile app marketplace. There’s an app for everything on your mobile phone. There’s chaos to be sorted out, but that sort of same explosion is coming to the physical.
Ruh: For the lighting world, you want to open up, there’s certain kind of applications. You don’t want to have someone double tap their gas turbine and suddenly you find it could shut off.
We had a very interesting security discussion that dealt with certificates and things like that, and if you don’t have a certificate, you would—the way it works in the consumer world, you would turn off that device. Well, you really don’t want to turn off your gas turbine.
People don’t get that testy when their connectivity or they have to reboot their machine. They get really testy when their electricity runs out at their house.
So there’s a different kind of world; however, the technology is the same. And the question is: How are we going to apply it in that world so we don’t end up with unintended consequences? Very different in industrial and consumer setting.
Hawkinson: I wish I had a picture to show you. A 10-year-old wrote to us last week and we are posting it on our blog in a couple of days. But an inventor started using Arduino Shield, that we built one that works with our platform, and hooked up a set of laundry applications because he was tired of his mom reminding him to do his stuff. And yeah, whipped it up and like literally over the course of the week and has a functioning prototype of the ecosystem.
So again, the barriers are lightweight. There’s certain low-hanging fruit problems you can go after very easily, and then there’s clearly control theory and on up exists for a reason in large-scale industrial processes.
Bruner: Dmitry, you are an investor in a company that runs satellites with Android.
Grishin: Yes, and I think Peter will be somewhere here. Better for him to talk about his company.
But yeah, I think we did several investments, and one of this is like satellite company use new standard called [Indiscernible] where it’s small satellites. The idea is for the companies to build huge group of satellites very cheaply and give access to them to students and further applications.
And I think talking to what Alex said, I really excited how quick he can go with his hardware device to the market. What we see in talking with startups, it’s too huge issue for them to work with hardware stuff. And if you look on their Kickstarted—a lot of Kickstarted companies later have problems to deliver products on time. It’s still huge issue. And I think one of the important things which we try to do, we think we need to—huge ecosystem of people who understand the commitment. Right now, people are sitting in big companies and this technology is not very, I would say, accessible to bigger startup community, but the situation is changing.
For example, we see more and more hardware incubators start to—we invested in one of them called Bolt, based in Boston. It’s like special place, machines, you have engineers who have experience of building the product, making the manufacturing. And of course it can help you to not make a lot of stupid mistakes.
But still, people joke in hardware, it’s called hardware for a reason, because it is hard. And really excited that more and more Internet people, coming to hardware world, and they bring a lot of Internet culture to consumer, but still they found a lot of huge problems.
For example, they need to deliver the physical items to the consumers, right? They need to work with distribution. And sometimes they have a lot of surprises. This size of the package is not working, you need this size. And otherwise, economics is not working.
For a lot of them, a surprise that retail take huge part of the margin. And they usually write the business plan and say okay, we have 20% margin and it’s good, but 40% of your final price can be—they could buy retails. And they also think huge thing which we see very often with hardware companies that usually when you create software or Internet product, you deliver to the customers, and if you see problems, you fixing them. But if you deliver hardware and you have some problems, people send you the hardware back and you need to do with support and understanding how to deal with this. This is—we see this over and over, and hardware startups now educating themselves.
But definitely, it is very important to have whole ecosystem of the people, and also very important to have several ways of startups. Because once you have successful startup, you have people who already know how hard it is and then they can teach and be advisors for the next wave, which we know from Internet. And not successful startups, can learn a lot and their next startup will be much more successful.
Bruner: One of the key points in this new hardware and connective device conference that we’re putting together which is called Solid, is that hardware is getting easier, but it’s still incredibly hard. And you have all of these challenges that people who have worked in software, which is incredibly scalable, incredible return on investment, doesn’t translate immediately into hardware in the same ways that you are used to.
Grishin: And I think in terms of hardware we need to do pretty much the same thing which we did in software. We need to build a lot of building blocks. We already are doing the raspberry pie, other things, but still it’s much more building blocks should be done. It is very important to build—much easier to set up any kind of hardware device.
Hawkinson: Becomes easier for each wave. You mentioned Kickstarter a couple times. We had a big Kickstarter campaign. And it was interesting for us. We had had enough history and founding capital in to make it work. The 1.2 million we raised on Kickstarter, just the hardware shipping cost $1.5 million, right? And so it’s not like no product development or other thing. We just announced our A round this morning in terms of continued economic support for the company with Greylock leading.
But it is hard. There’s enough moving parts. The barrier’s not nothing, of course, and the risks are you only get a couple shots. If you ship a lot of hardware and it has got a fundamental problem, you are done, unless you’re a giant company. But the barrier is coming down to the zone where in the active startup and investor community, it is possible to build some real winners out of it.
Bruner: I want to raise a point Dmitry made earlier when you said you think the next round of innovation is going to come from startups and not from the big companies. So I want to hear Carlos and Bill respond to this. Both of you are making a lot of investments in this world and taking innovative approaches to some of the problems of software in the physical world.
How are you building the right sort of innovation structures inside Cisco and GE?
Dominguez: It’s kind of interesting. I think the work that Dmitry and Alex and others are doing is actually great. We want innovation to happen everywhere. But when you look at it through the lens of Cisco, we see a couple little different things.
So number one, is we are moving from a world that there are billions of devices connected, and with all these very inexpensive technologies and being very easy to reproduce, there’s going to be trillions of sensors and devices out there. That is not a minute task. It is a scale, function, growth issue.
So how do they connect? How do we connect them? How do we make sure they are authenticated, that you say who you are?
And kind of very interesting, there’s a manufacturer that makes light posts, LED light posts. And it’s got a mesh network on the top. It’s got these biological sensors, water sensors. It’s measuring pollution, temperature and all the things you want. It’s got LED panels.
So imagine this is deployed in a city during some sort of major festival or fun thing that’s going on, somebody hacks into the height pole and I can say on one of the LED displays, “biological agent detected, emergency, run this way,” and I hack the other side and say, “Run this way.” You could actually harm a lot of people by just hacking into a light post.
So these are issues we never had to confront, but now we do. So from our lens and our point of view is making sure that things are secure, they are what they say they are, and we are running all these things behind it really makes a big difference.
The second thing is everything’s generating data, but there’s a new wave. When we think of big data, we think of it being shipped out somewhere, being stored and processed in the cloud and all that other stuff. There’s a whole new wave of data in motion, which is the sensors and because of the processing power and everything else, things are occurring on a local basis.
For example, back to the lighting, if I’m walking around, the lights are going on as I’m walking, if it’s going back to the cloud to some central site to turn it on, it may not make it in time for me to do there. So you want to basically make sure that you are activating and processing and bringing some insight and knowledge at a local level, but you also want to take that and combine it with some historical stuff to give you the best outcome.
So we are worried about the security, we’re worried about how to connect everything, because this is going to put a tremendous amount of burden on the network, and how do you do it efficiently.
Bruner: So security and robustness.
Dominguez: Robustness and then mixing it with the analytics process. Sometimes you need to reconfigure.
I mean, you’re in a stadium, just an example, you’re watching the World Cup, somebody scores a goal. If you look at the peak of what’s happening on the network at that moment in time, everybody’s doing a video. If you’re a service provider, the demands that are coming to you are incredible.
So how do you do it if you’re doing it on a manual basis? You can’t. So we’re looking at it for those peak events to have enough intelligence on the network with automation and some controls and business rules to be able to reconfigure things on an on-demand basis, maybe get more servers and open up more storage, open up more bandwidth. That’s kind of how we are thinking about it.
Bruner: Bill, how are you approaching this problem of innovation inside a big company?
Ruh: Look, I think that’s mostly true in the consumer side. There’s a very low barrier of entry here, failure and pivoting and fail-fast. You can do that in a consumer world, and everybody has that knowledge. Now, on the industrial side, it’s quite different. Fail-fast pivoting on jet aircraft engine business is not really going to happen, at least I believe in our—my lifetime.
I think that the technology in complexity, the regulatory environments and so on—and by the way, simple thing like the law of physics has a lot to do with what you can and can’t do.
So the fact is, I think that’s going to continue to be the domain of big companies who have that kind of skill set.
Now, where innovation is going to occur is around that operational number, because when I talk about the size of the opportunity in the industrial world and operations, it dwarfs that consumer side, but the problem is, how do we innovate around that to improve. So can I get insight into how I operate the jet aircraft engine so that I can reduce fuel burn, the number-one cost for an airline? That’s tremendous opportunity. Can I figure out how to reduce and get rid of that unscheduled down time? I think that’s where big companies and small companies will play.
But if you really are going to get into this idea of a gas turbine and a power plant and generating electricity, if you don’t have some domain knowledge, some way to connect in to understand the physics, it’s very hard to get in.
So I think the reality is we’re going to see that companies like GE in the industrial space have to embrace innovation and help to drive it. At the same time, I’ll tell you, we are investing in to small companies who are innovating. We invested in a company called Mana and it does search but for the industrial space.
So I think—because the idea of search and how you could apply that to people in the field in a way that it supports their job, where they find the right person, the right knowledge, and we’ve talked about that for how many years, I think we are just at the cusp of getting there.
So to me the industrial space is a little more complicated than the consumer space. It is going to require partnerships between the big and the small. It is going to require leadership, because in the end, the one thing that happens is people get real testy when the plane doesn’t fly, when the electricity goes off. How many people really are okay with a couple of days without electricity? It’s a totally different environment you deal with.
So when you think about it, I think a partnership is more important here, and as a result, this is going to go a little slower, but I think that’s how it will play out.
Dominguez: There’s actually kind of an interesting dichotomy. On one end, you’ve got an aircraft engine, the other end, you’ve got these sensors in the home. Where I think the real value is going to happen is when we kind of meet in the middle and converge. We just came back from an event in Barcelona we sponsored to try to bring 700 different companies together to talk about standards and data exchange and a few other things.
One of the examples they showed, they had sensors at the Barcelonians were putting on their windows that detected air pollution, temperature, and a variety of other things. The example that was shown is the actual city of Barcelona was bringing it all in, monitoring pollution, and then from there, they were driving some gamification programs down to the users to say, Hey, we’ll give you’ve a bonus if you ride your bike today, or we’ll give you a bonus if you ride share or take this bus and you’ll get some things.
It was kind of interesting, the power of looking at these silos all of a sudden being converged both from an industrial and government side and from the user side.
Bruner: Dynamically priced air pollution.
Dominguez: Yeah. And then the third layer of that, which was quite interesting, is if you took the third view, which is the enterprise account, if there’s—they saw in this graph there was going to be a demand on power because of the temperature and what they wanted to do was actually put incentives out to the enterprise accounts to shut power down, shut lights off, shut things down.
So you are starting to see—where the power really lies is not this or that. It’s really the convergence of both.
Bruner: To me the power in this is that it’s modular. So you can have GE operate the gas turbine very reliably, supply some data about output, which is an industrial firm and very robust and very sophisticated, then supplies data on electricity pricing to any module that can accept data on electricity pricing and do intelligent things with it.
Hawkinson: One comment generalized across this space is even on the consumer side, I think the companies that win will take things like the security privacy control theory underlying pieces that you need to get right very seriously.
So us as an example, even though we are a little startup, 40 engineers, we do ethical hacking audits on the outside on the platform every single month. We have watched with great interest the Belkin WeMo Hackaday post that started a fire in a home by turning something on and off too rapidly that wasn’t on our platform, but realizing that it’s not always nefarious. There can be two apps that are competing to turn something on and off, let’s say, as an example. And you have to build into the platform level that not happening.
So back to Dmitry’s point on building blocks and your point on modular, we view it as the role of some of the early companies, smart things on the consumer side to try to put in place some of those things so it frees up innovation by others, so the app developer shouldn’t have to know that about the device that’s being touched by their application.
Bruner: Now, Dmitry, a comment.
Grishin: Let me not agree with William. I think industrial market need much more innovation, much more competition, and I think right now biggest companies don’t have much motivation to try to make disruption technology.
For example, if you look at car industry, people do very minor improvements until Tesla come into the market, and then they try to work quicker. And I think that we still do not have supersonic jets only because of this. We should have much more competition in the industrial world. And if you can look on Internet, we already see more and more consumer products coming to enterprise, and I hope the same situation will happen with hardware.
Involved in creating consumer market, we will try to bring it to big industrial application, maybe it’s building blocks, maybe something else, but I fundamentally believe that we need much more competition and big—for big industrial companies, and until now, they don’t have too much motivation to make huge bets.
They improve engine by engine. So if you fly maybe 8,000 kilometers per hour, now 8,500, 8,600. But to make huge bets, we need to—much bigger competition.
Bruner: So I’m sorry to cut the conversation off right here. We only have a few minutes left, and we definitely want to take questions from the audience. Over here, we have a couple of questions.
Stikeleather: Hi. Jim Stikeleather with Dell. In one of our scenario game-playing routines, we took the industrial Internet, we took the advances that were taking place, material science, we took this real shift from ownership to rents that are going on.
Theoretically, GE is approaching the point where you could build an engine that literally would last forever with all of the sensors and everything that’s associated with it, or a locomotive or any of a number of things. Ford could build a car that could last—because you are monitoring everything and predicting analytics, telling you repair it before it fails.
That totally disrupts manufacturing business models and has a significant impact on employment, both in manufacturing and repair and maintenance.
So I was wondering if the panel could kind of talk about those sorts of implications, even though everything is going to be better, but there’s some real implications there.
Ruh: Just a couple things to note, with regard to that, first of all, if you just take jet aircraft engines, it takes years to build one, right? So it turns out that our engines last decades. In fact, we sign service contracts to service them for decades, and we could continue to service those over longer periods of time. What is changing, however, is the air frames change, right, size, scope, and you have to adjust, obviously the jet aircraft engines.
The second thing is, innovation is occurring in these markets continuously. If you could, you would innovate every few years, like you do with the Chipmark and others, it’s just the scale and the time frames of these are longer.
So over time, the ability to put new materials—and materials, science, supply chains, everything are changing to the point where it is much more cost-effective to change those out in two decades than it is to try to keep something forever, just because you are able to do things with the physics you couldn’t do before.
So today, these machines run for a very long time. In the utility domain, we have products that have been out there 80 to 100 years in the field that are still running. So the fact is that what does change is the materials, the physics, the designs, the need to deal with other kinds of environmental factors, and you’re always going to have innovation on these. So I think—but today, many of these products are living for a very, very long time.
Hawkinson: One quick comment on that, you know, that I think is in the job space even more so than the—in the hardware universe. I think the implications of connecting up all this data have big implications for other industries.
And I’ll give you two quick examples. One is the insurance base, which is about risk mitigation. You can mitigate a lot of risk, right, with perfect information arising out of the world. One example, which I think that whole industry is ripe for disruption for the next five to ten years.
We heard from Aetna, the healthcare space, in a wonderful way, the implications of putting these sensors in place. And sort of as an example in elder care scenarios, where we already have lots of applications being built on our platform that improve the lives of people incrementally very simply, but there’s a big ecosystem implication of that as well.
Bruner: Let’s just take one more question, very quickly. Yes, sir.
Dhillon: My name is Gaurav Dhillon. I’m with SnapLogic. So a question from a software plumber to the audience. What standard should we pay attention to and protocols? There’s OBD-II in the car industry, but there’s so much of this going on, that for us to track it and help our customers use it, we have to pay attention to this. So just your thoughts on standard and protocols in this area.
Bruner: Carlos, IT.
Dominguez: Well, IT, actually, if you go to our web site, there was a lot of standards discussion in Barcelona. And that’s all being published, so you might want to see what the collective intelligence of about 700 people that are trying to do this market, what they are saying, but TCP/IP is a very foundational one.
Hawkinson: We see there’s a lot of underlying competing standards. It’s very fragmented now at the connectivity layer. So Wi-Fi, VisiWave and a bunch of others that are out there. There’s not yet a set of standardizations above the connectivity layer for applications, right, and common data types that come out of some of these physical connected things. We are trying to do some of that in our platform to make it easier, but it’s one of those places you will just have to see evolve a lot over the coming years. So just encourage you to ping me if you would like to talk about it more.
Ruh: We think on the industrial side there’s plenty of standards and regulations, by the way, more than sometimes you can figure out how to deal with, I think and probably know that.
At the same time, what we see is that—and we are working now to try to create some consortium to deal with the industrial side of this as effectively as the consumer side is already done, but the industry is a step behind. And as a result, I’d say you end up with more proprietary solutions, which don’t lead to this building-block approach, which we actually see as the big inhibiter to being able to move these industries forward.
So the fact is, I think, we’re probably a decade behind where the consumer world is.
On the other hand, I think, it’s a natural progression that the technology pricing had to come to a point where the industrial world could take advantage of it. And consumer world is still going to lead the way, and we’re going to have to figure how to evolve that and special it for the industrial world, but not do something different. And I think that’s been the—that’s the challenge that’s existed in this space.
So I think you’ll see some things in the next coming twelve months that would start to change that.
Bruner: All right, I have let our conversation go well over time, but I know that we will continue this discussion over the course of the next day or two with all of you. I’m looking forward to it. Thank you, again, to the panel. You guys are fantastic.
Kirkpatrick: Great discussion.
Congratulations on the news today. That’s very cool.