 Live from Las Vegas, Nevada, it's The Cube at IBM Interconnect 2015. Brought to you by headline sponsor, IBM. Okay, welcome back everyone. We are live in Las Vegas for The Cube here at Mandalay Bay for IBM Interconnect. This is part of the Go Social Experience, part of the big conference we're bringing social media, social business, social content. This is The Cube, our flagship program. We go out to the events and extract the civil noise. I'm John Furrier, the founder of SiliconANGLE. Show my co is Dave Vellante, founder of wikibon.org. Our next guest is Stefan Berber, VP of Engineering at Bosch Software Innovations. Welcome to The Cube. Hello, thanks for inviting me. Thanks for having us. We're live, we're pumping it out all around the world. So first, tell the folks out there before we get into some of the conversations around the internet of things, cloud, what all this is happening. What's Bosch software about? What do you do there? Obviously, VP of Engineering, you manage coders and all that good stuff. What are you guys doing? Talk about Bosch and then we can get into the conversation. Yeah, we founded the software and system house within Bosch in 2008 to enable us in the internet of things to conquer this world. The software innovations is providing a software platform for Bosch business units to connect their things, their product to the market. But we also offer this for third parties, for partner business. And as we believe in an open platform approach, we also believe that IBM and us will play an important role in the future. So talk about the company global in size, presence. So all around the world, certain geographies. Is the consumption model regional and global? Can you explain some of that? Yeah, we are growing by acquisitions. Now actually we are in the third phase of another acquisition we just announced on Monday. So it's just a week ago. We started in 2008 with a company in Germany, Lake Constance, very nice place by the way. If you go vacationing there, it should be there. Then another Java company in Berlin. Now we bought a company located in Cologne in Sofia. And we outreach now to Chicago here in the US. In Singapore is our Asian hub. We have also an office in Shanghai. Though we are still a startup with around 600, 700 people. We have a small startup. Yeah, for Bosch it's really a startup amongst these nearly 300,000 employees is just a small entity in the Bosch family. Well Uber's a startup and they just raised a billion dollars worth of trillions. So this is the question that everyone's never mind. I think big companies act nimble and agile like small companies and like small companies feel big is kind of the theme of this new way. Cloud powers that. But one of the things that we talk about in theCUBE is this global consumption. The need to be global from day one. Global first is a term we coined here in theCUBE. Mobile first kicked around in cloud first. But global first, what's the global view from an engineering standpoint? Is it challenging? What are you guys doing? Well IOT if you look at the markets are strong in the U.S. it's strong in Europe and strong in Asia. So if you miss one of these markets you have trouble from the early beginning. So you need some footprint of course and with our brand we are of course strong in Germany and Europe. But we also see a lot of traction we get here in the U.S. and then Asia is also easy win as we are very strongly growing there. So IOT, I mean it's so big. I mean it touches so many parts of our lives. So in that sense we can relate to it. On the other hand there's just so much to it. So how does Bosch look at the internet of things? Maybe take us back to the sort of beginning of when you guys started. Making it probably wasn't called IOT back then. I mean pretty early called it that way. Internet of Things and Service was the first term. We skip the service. So what time frame is this? The initial discussion internally we had started in 2005. The question was what would happen if all our Bosch products are connected to the internet? How this will change our business? How we have to adapt technology-wise? And then we had a long discussion. Are we going to do this ourselves with the software? We have a lot of embedded software out of these 40,000 engineers where we have 15,000 software developers in the Bosch group. But most of them do embedded work. And IOT is embedded, but it's much more. It's also distributed networks. It's Java development, JavaScript. It's web pages, it's databases, it's M2M. So we thought, okay, we have to have a footprint in this market and that's why we started our own company understanding this business. So the early discussions back in the mid-2000s, what was the impetus? Was it predictive maintenance? Was it consumer behavior? At what point did privacy inject into the discussion? Yeah, yeah, we have a very generic view on that. Each product is associated with the service in the future. It can be predictive maintenance. It can be improving the product. It can be even linking this product to others with new functionality behind. So we see ourselves to be also a service company in the future and it's software as a service aligned with the product. And predictive maintenance is I think the no-brainer. You can see particular production and expensive assets on the road. But there are many more, making things easier to use, understanding how people using your product is also motivation. Insights. Yeah, insights from that. And if you look at the market of insights, how people are really using things is a bigger market than advertisement. And so there's a lot of opportunities out there if you have the data and the connectivity. So the idea germinated, say, in 2005, how did you get, and then you made the decision let's do much of this in-house. Maybe do some partnerships. And I'm sure you had some open source discussions which you talked about. Yeah, a lot, a lot. So okay, so what came next? Is okay, now let's get a team and let's do some experiments. We started in some verticals to be early mover. Where did you start? We started with a project in Singapore, Electromobility. You see, if you have electric cars, charging station, drivers, the scarcity of charging station, you have to park and charge at the same time, you need reservation, you need billing and payment. We thought, okay, that's a good thing for a startup. Look at smaller projects where connectivity is a must. You cannot do this differently. And we won this project in Singapore to provide the infrastructure. And from there we gained. So we did the first installation of our platform, we learned from that. Then we acquired another company that brought BPM along with that. And because we know that a lot of processes being run automatically. And if you do, if you collect sensor data, and we produce three and a half million sensors every day, if you consume this manually with people, there's no way. You need really automated processes in the backyard to take over the data and make business sense out of that. And this learning, yeah, I think we'll continue with each project we do. But the good thing is we have product out. So there's a heating system being connected since 2009 with an iPhone app. It's also a robot lawnmower coming to the market just these days. We have electric bicycles being connected. So there are many products already in the market where you can see IoT make a change already. So at some point somebody said, well, how are we going to make money at this? What's the business model? It's a business model, yeah. So we'll talk about this. It's an interesting part. For companies who use to build products, it's a change in business model because the service alone, you cannot really pay upfront. You have performance based pricing or fee or subscription. These are the business model usually suitable for these kinds of products. Traditional companies have their problems with that. And that's why we started a lab at the Hochschule St. Gallen in Switzerland. This is a real expert university on business models and IoT. And there we have now 10 PhD students, they write a thesis and their goal is to start a company with this business model. So we're combining this, it's also that pushing our organization to a new business model with these expertise. So specifically, how do you make money in the IoT business? Is it a subscription base? Yeah, the description is something that usually always have in there but we see that we have to grow with our customers. So that is risk sharing also behind this. So we have business model on the numbers of things being connected, the numbers of transactions you have. So once the growth is coming and our customers also making more money, we also then taking more revenue out of that. This is, I think the general pattern. So it's kind of an elastic pricing and revenue model. And the traditional model that I get all the money when I sell the product, I think this will die sooner or later. Now what about open source? Where does open source fit in? I wanna talk about anything you do might be doing with IBM there but let's start with the open source question. Open source plays an important role. You believe in an open platform? Doesn't make sense to build an internet of things platform for Bosch. It should be for all things around the planet. But we want to have some ways of steering the development of this platform. We want to co-evolve this with partners. And of course in strategic points we want to know what's going on. Like privacy, we want to know where the data is. We want to know how the policies of data combination and distribution is. That's why we believe in open source as a business model also. But also we have a lot of open source technology in our software anyhow. We are a member of Eclipse. Of course, most of our gateways are Linux operated. We are a Java company and this I always assume is part of the community based programming language. And this is a lot of similarities with what I see with also with IBM. We are strongly founded on OSGI technology. Also something IBM has been doing for many years. All of our tooling is basically heading to Eclipse and automotive but also for IoT tooling. And Eclipse plugins are just the way to go. So talk more about your relationship with IBM. I mean, I know Bosch is a big company. I'll show you by a lot of mainframes from IBM, but specifically as it relates to the software side of the organization, what are you doing with IBM? Yeah, so one of the areas that is interesting to look at in the embedded world is that if you build embedded systems there are a lot of variants of each other. Usually we build platforms and out of these platforms we derive software pieces and engine control software is an excellent example. There we produce 20 million engine controlled units per year and they go to different customers. Each customer have different car. So we have 400 projects in parallel running being supported by three and a half thousand Bosch engineers and this to manage this with software configuration management, collaboration and this 1,500 variation points behind this. This is done with IBM product line engineering solution, PLE. And this is really co-evolving activity. Our work in product line and IBM's work there, we co-innovate basically this product. And we're happy to have IBM to bring this to other companies in the automotive domain as we see that if this effort is spreading around this also helps us cooperating with our customers and partners. So when you say co-innovate, I mean you bring engineering resources, IBM brings engineering resources and you create solutions together. Actually all the new features in PLE are derived together. I wouldn't say that all the features coming from Bosch but I think the most innovative one we're providing and knowing what's important in that business I think it's also key value for you folks at IBM. So how does this all translate into actual activities in the business? So Stefan if I had to say, look at Bosch's enormous company, huge product portfolio, how much of this portfolio is Internet of Things enabled? Okay, as of today. Yeah, and where's that going? I will see, oh Mr. Denner is very clear on that. He wants to see a plan for each product when it's online. It's not a question if, it's just a question one at the right time. And if you look at some very tiny or very inexpensive products, the time is a little bit later but if you see at the important assets, it's much earlier like our robot lawn mower, it's around $1,000. Of course you can make this an online experience. Connect this to iPhone, do remote, remotely start your lawn mowering, setting up calendar times when it should mower and it shouldn't depending on weather and all these things you like, as you don't like to spend the time on your grass. So this is there already here and some other products I think maybe five more years but you don't know. We hit the inflection point and internet of things and we are prepared, we're just waiting for when it's the right time for a product business-wise to go online and then we go out. And what about, we talked about privacy a little bit before. A couple things, well first of all you're in Germany which is very strict about where data can go. So presumably you have to leave all the data in Germany. But how are you handling privacy? Is it an opt-in model for the consumer for example, your lawn mower example? How does that all work? We want to make this very transparent. We believe that trust is an important ingredient in terms of things or this will not unfold if people do not trust these systems and that we handle their data in a trustful way. Then there's no innovation, there's no data and then there's no value behind. So we allow each customer to decide which kind of data they want to share with us and which they don't. So this is very transparent to them. Of course then some business model are not possible like sending advertisement to them is not available anymore, which is okay for us because we believe that the product and the service itself has its value and it's basically price. And if you do something on the side, okay you can do that, but it's not the main purpose of our business. Our main purpose is invented for life and this is improving lives of people and not sending advertisement to people that they didn't want to see. So are you doing anything with, talk about data a little bit. Talk about we were just at the Hadoop world last week. We've been doing Hadoop, John and I and theCUBE for a long time focusing on an area. What do you do with all that data? Are you using things like Hadoop? Where does that fit into the whole strategy? Yeah, storing the data is an issue by itself. We use all kind of technology. So if I look around I'm always amazed that I find something new. So we have of course Hadoop, we have MongoDBs running around. We have IBM databases, Oracle databases. Basically it's a variety of data storage and each of them has its own value. Like if you go to NoSQL databases, MongoDB is a good example. Of course we have that. But one thing is to store the data. This is actually the easier part. Taking all the data and making sense out of that it's more difficult part. We have a partnership with NIME that's a small company located in Germany and Switzerland and the US for an open analytics platform to make more value out of the data because in the end it's not sufficient that we only have these highly skilled data analytic guys who can make sense out of the data. We need about five million IoT developers to do all these applications and they must be able to do analytics in the daily base and that's why we are looking for this Swiss army knife of big data analytics so that 80 or 90% of the big questions can be answered by a typical engineer and although they're very tough and more, the last maybe one or two percent we go to data analytics people. So John and I are always talking about the developers as the linchpin really. How are you, what's the message to developers? How are you attracting them? You got, you published an API, maybe you could talk about that a little bit. We just published in December an Eclipse IoT project called Wato that's the interface to things. That's an approach where you generate specific platform drivers for a refrigerator, security camera or car based on the common abstraction language. We see there's some traction behind. We just announced this in the eclipse conference in Ludwigsburg last year and we see that a lot of company jumping on board and I hope also IBM is joining us. I had a lot of discussion here on the conference about that. I haven't seen this yet so it needs a lot of education but I think this is the way we go that we have one common API to all things. This is what openness really means and then all the developers, they don't have to care whether this is a Bosch device or a GE device or somebody else. They can start doing the application. So choice for customers is a big thing. That comes up. So cloud is great but it's happening in the open. Glue Mix is developing fast. What do you see for cloud? I mean, open stacks out there. Can you comment on infrastructure as a service, platform as a service? Where are the work areas that you think need to get done to move the ball down faster towards getting to the goals of the customer outcomes? The infrastructure as a service, we rather source from somebody. So actually we are agnostic there. We just need the resources of course. We need the network. We need the storage and also the compute. Platform as a service, we are adding some ingredients that are related to managing the devices out there, having security there. And also you have some nice technology about Roots Engine. We bring this in there. And also we have BPM technology we think we can leverage there. And then on top, the real innovation happens in software as a service. Something a B2C consumer is really interested in to buy as a fee or performance based pricing. Or even B2B customers like predictive maintenance for a specific machine integrated into the business process. This is our main revenue source on that stack. What does systems of engagement mean to you, that term? Especially internet of things, could be a probe, a sensor, a human being to a mobile device, real time. What's some of these concepts? Has that translate down to what's under the hood? Yeah, if you look at the whole user experience world, it's so interesting now because you have a smartphone, maybe a variable. You have a computer screen, laptops, PCs. You have an operator and you have the thing itself. So there are so many ways you can interact with the service and the thing. It's a little bit indistinguishable in the end. And the whole industry design, computer, UX, UI stuff, this all comes to a new arena. And making this a seamless up to somebody you call the operator and the operator is able, yeah, I know your devices here. It's in this situation getting this as one smooth user experience. This is tremendous task for engineers and also designers. So this is something we are working on. We want to push our brand in front because this would be seamless. How do you handle the cultural shift that's happening with companies and then employees and engineers where the new theme is open source is all outdone in the open, but you got a business that has some legacy and also getting engineers closer to the customer action. It used to be keep them in the back room and go away. Now design, iteration, agile design, not just being graphic design, but like design and customer experience with real time. There's real sensitivity towards relevancy and fast, it's fast time to value. So how do you deal with that? Like you give them training, you bring the engineers out in front of customers. Is it a mindset? How and if do engineers get close to the customers? User experience is a lot about mindset. So we have a sort of own academy for this topic. We're going through trainings, this is something. But I think it's rather the awareness training that it's important. The way you do it, you rather do this in the concrete projects. And having success in a concrete example like in our e-bike situation, we have a very nice navigation computer now. It's called Neon. And if you see this, everybody within Bosch understands, yeah, this is the way we should go. And this is all the lighthouse projects that steer the whole company forward. So what about BlueMix? What does BlueMix mean to you in terms of their evolution? Where are they in their journey? A lot of times the developer's like, well, I'm comfortable with Amazon if you're born in the cloud or if you're in the data center, nah, I don't want to do the cloud security. Is there new things that you see that are really forcing developers saying, okay, I can take that step and put my toe in the water? I think I was very early on seeing BlueMix. There was the day before the CBIT opened in an overfare two and a half years ago. I saw the first demonstration and I thought that's the way we should go. And if you look at the concept we have for our IoT platform, it's very much this idea that you have microservices, you mesh them up to an application, you have multiple companies providing their service on one platform. I think that's exactly the right way to go. Do you think there's an opportunity to actually monetize the data that you're collecting? Is that something that you guys talked about? Yeah, of course, yeah. There are interesting models now behind. You're always... Are you doing that already? Yeah, we... Data as a service? We offer data services, analytic service, this is something. So we know, we have domain expertise, we know how to do it work and also we have analytics capabilities. So if you look for production, for example, we have ourselves 250 plants worldwide. We know exactly how to analyze data from production shop floor. This we can offer to other companies and we do this. Why do they buy it? For benchmark purposes? It can be benchmark, it can be predictive maintenance, it can be, please give me a model that I understand how much testing I need at the end of the line for specific product depending on the parameters of production. It can be on, should I ship this product or not? So it can be multiple things. The data itself is more interesting because today we don't have really a data market. If you look at the business model, the internet company currently have, I always refer to it as a black market. You give your data and you get software for free and you don't pay taxes. That's the definition for a black market. You don't pay taxes for something you get. You have to get out of that. You should be as transparent as buying a car. What is the value of data? Who has it? Is it really anonymously? Is it aggregated? This is the kind of information you need the transparency behind. Do you feel Stefan Bosch is a cutting-edge player in IoT? Relative to your competition. Where do you, how do you guys size up your own sort of internal competitive? Well, I can claim we started very early thinking about this. I think also from the traditional companies, we are early mover. I hear in the US, I see GE also in a similar situation. And we are pushing for this. And if I see all the support I get from our board, our CEO, to move forward there, it's pretty cool. So I'm very confident there. But internet is a very fast business. And it's always a question whether a company that is more than 125 years old is able to take up the pace and remove that. You have resources. But sometimes you have, it's slower to make decisions. But I mean Bosch as a global player, you mentioned GE, obviously IBM with Smarter Planet. I mean, in a lot of respects, they are the ones that we're going to drive internet of things. Of course then you've got the developer ecosystem that you're trying to appeal to. And some new startups. Let's see what's coming out there. Yeah. All right. It's exciting. Yeah, it's exciting. So we're at Open Source. How are you guys looking at Open Source? We were coming here earlier. Actually someone picked up our tweet, Open Source is probably the movement of the century that's changed the society, the world. I'm a strong believer in Open Source. In particular, the internet wouldn't happen without Open Source. And I'm pretty sure Open Source wouldn't be there without internet also. Because otherwise you cannot really cooperate that easily across the globe. Yeah, but Open Source used to be renegades, fighting for freedom. Now it's standard. Open Source is now tier one first class citizen. Is it first class citizen? Yes. It truly is. Awesome, right? It's generations of layers of improvement standing on the shoulders of others before you is always the phrase. But now it's affecting business models, the federations of federated groups. Does this concern you? You worries you at all? Are you happy with it? I mean, it's affecting business. I mean, we're Open Source media. We heard Power has an open power systems group here. IBM has had great success. Over 100 partners and they're a new open source kind of movement towards hardware, open compute in the data center we've seen. Everywhere. Everywhere. So I mean, does it stop? No, I don't think it would stop. First of all, the IOT will also be only possible with Open Source, I'm pretty sure. And of course, for a traditional player like Bosch, it's a learning curve until we have all the lawyers or the business people and controlling people understand what value capturing is all about. It's a completely new for them. It's the educational part. But on the other hand, we wouldn't be here with our software we have today without Open Source. It's pretty clear. And we're pushing this farther. And the question is, how you build up ecosystem that do the right thing? And this is, you cannot really control. You just can ask and ask people to join and then. Where's the next wave? Obviously, Internet of Things is a big wave. People are saying that's the wave. It's very bubbly right now. Also, you've certainly, the consumer side, you see Uber and whatnot, it comes going public. But in the enterprise, there's a lot of waves coming. We were, we've been speculating on the cube here that folks watching know that we believe that it might be high valuations, but innovations here. Usually, that waves come in sets. Is Internet of Things the next wave? And what other waves might be out there that people might not see? We're hearing all kinds of things about machine learning getting easier. Talking about in-memory kind of transactional things. Coders doing better things. More library automation orchestration. This is all buzzwords, but that's reality. What's next? What do you see the next waves? I think the waves will be a little bit anticipated in specific markets, like Smart Home, believing is catching up. CES here in Las Vegas was a clear indicator that this is going to happen now. People now are so used to their smartphones and even youngsters, they don't believe if some software is in a dishwasher, why I can use my magic wand, my smartphone, to access the dishwasher. It's also a generation topic, maybe for us. There's a lot of data exhaust, too, coming out of it. I'm in the market for a dishwasher, and this conversation is really swaying me. I'm actually in the data. You get garbage disposal, you get rid of the trash, the bad data, the good data. So the exhaust of data coming from this is off the charts. So I've got to ask you, people talk about data lake, it's kind of like a data warehousing concept. Yeah, of course you can take all the data out and store it somewhere, but I think it makes sense to analyze the data locally first and really aggregate it in a way that you make sense out of that. Like sending out temperature information from this building, everyone second, does it really have value? No, it doesn't. So you really look at what I'm interested in. Temperature peak, might be interesting, somebody open the window, or something. Wild variations, something like that. So this is something, you have to put more brains behind the data, and this has to be locally, not only in the cloud. But it's diverse, too, there's different use cases, different architectures, especially indiditive things. Doug Baylog was saying, one of the key conversations that they're creating, that they're joining and they're enabling, is one of them is drowning in data, this insights model. People drown, people can drown in many things, companies can drown with too much data, different not being prepared for data. So, will there be more deaths for the comparations out there in the drowning of big data, in the lake or the ocean? What do you think? Which one will have more casualties? Lake or the ocean? I think it's just a business question in the end, how much money you spend on the lake or the ocean? I mean, statistically, more people die in the ocean. The lakes, but lake is a term that's been kicked around. We're talking about a new term called the data ocean, because it's a little bit, there's a lot of different currents, different diversities, internet of things, there's no data types, you've got to be prepared for anything. Anything, yeah, like in our situation, our M2M is prepared to take any information model from a device that's coming up, dynamically. If you're not prepared for that, you're busted, because every day there's coming out a new thing, every day you learn there should be a new API to a thing, a lot of new innovations and creativity people out there, and if you do not support them, oh, you're just stuck with old APIs, inflexible structures, and it's not the way. And the weather forecast for that kind of data ocean is to use predictive analytics and technologies, right? Because internet of things, these are probes, right? Yeah. And regarding the lake and the ocean, I believe, if you really look how much value I get of the data, then you also question yourself, shall I store it? Right now, I have the feeling we store everything, and we burn a lot of money, energy, and compute, and the storage capability is that. I think in the future, we will be more sensitive, what makes sense. I don't know about you, but I can go back to my emails from first emails I wrote at the university time, I still have it in my backup. But it's a small format, how do you get it? How do you get it? Yeah, it's a text email that's easy. That's where you put the lake. You put it all in the lake, right? You put it all in the lake. You put the new tape, they say. Well, you can surf waves in the ocean, and that's a good thing. You surf the waves in the ocean. The lake is a little bit more reliable. You put everything in the lake and surf the ocean. It's fresh water, you can drink it. You drink value from the lake, and you surf the waves in the ocean. Okay, data ocean, data lakes, we're here, breaking it down. I'm John Furrier with Step On. Thanks for coming on VP of Engineering, Bosch. Great to have you. Again, data ocean is trending on Twitter. We'll be tracking this. We'll be right back after this short break.