The IoT Show EP02
ALAN BEHRENS [00:00:28] Welcome to another edition of the IoT show. My name's Alan Behrens from Taxal and I'm joined by three notable guests. Bill Boswell from Siemens. Josef Waltl from AWS Amazon and Diego Tamburini from Microsoft. So welcome, gentlemen.
DIEGO TAMBURINI [00:00:48] Thank you.
ALAN BEHRENS [00:00:49] I think first off, let's just let our guests introduce themselves. Bill?
BILL BOSWELL [00:00:53] Sure. My name is Bill Boswell. I run marketing for Siemens MindSphere and our cloud application solutions and we talk about MindSphere is the cloud based open IoT operating system. So I'm excited to be here Allan.
ALAN BEHRENS [00:01:04] Great, thank you. Diego.
DIEGO TAMBURINI [00:01:07] Thanks Allan. My name is Diego Tamburini. I'm speaking to you from Seattle, Washington, US. I'm a principal industry lead for manufacturing in our Asher Engineering Division, our cloud team. And my team work the Asher pro development team providing the industry point of view. We help them understand the opportunities and challenges of the industry so that they can develop the best platform for manufacturing and specifically for IoT.
ALAN BEHRENS [00:01:42] Great. Thank you. Josef.
JOSEF WALTL [00:01:45] Hello, Joseph Waltl. I'm the Global Segment Lead for industrial software at Amazon Web Services. With industrial software, we mean all types of applications and services that support manufacturing companies in product design, production design and production. And our goal is to have the widest variety of industry leading applications for our customers in collaboration with our own technology to solve their industry specific problems.
ALAN BEHRENS [00:02:13] Great. Thank you. So I think, you know, just to set the set the scene. Why don't we just ask each of our guests to give their view of what IoT and the industrial IoT means to them and their companies?
BILL BOSWELL [00:02:25] Okay. So when I think of industrial IoT, what really jumps out at me, Alan, is is the complexity and the the need to connect into very brownfield kind of environments that are out there in most factories and most plants today. Not that we're not adding state of the art, new technology as well that has IoT capabilities embedded into it. But there's really this mix of of environments out there that where the context of how a machine operates, how a plant operates, how in other industry settings like mobility and building technologies, how the control systems work, that really brings a whole new dimension to an IoT problem. That context is what what sets it apart for me.
ALAN BEHRENS [00:03:08] Right. Interesting. Yeah Diego, how about yourself?
DIEGO TAMBURINI [00:03:11] Yeah IoT is strictly speaking about collecting sensor data from your devices equipment, but all those data of the health of your devices, data on how your customers are using your products. So data that ultimately is going to provide you insight to to reduce cost or to get new sources of revenue. The topic of IoT is relevant to our customers because with it they can pursue new revenue opportunities, new efficiencies, new services, and that can ultimately give them a competitive differentiation and a for my company. Of course, we we are essentially a software platform company. So we develop the the the cloud infrastructure for customers and all their software developers to develop their IoT solutions. So so for us, we think very, very intently on what are they, IT ioT capabilities and services that we can, we can provide in the cloud platforms to become building blocks for people to develop their software solutions.
ALAN BEHRENS [00:04:28] Great. Thank you. And Josef,.
JOSEF WALTL [00:04:29] Adding on top of this. What we definitely see in IoT it is the whole industrial space for industrial IoT is really complicated as a whole lot of technology out there. So this is referred to as the automation pyramid where you have the PLC on the controller level. You have these cadre systems on the line level and you have these manufacturing executions systems on the factory or shop floor level. And with IoT, you can see that there is new sense of data coming in that hasn't been there before by retrofitting by new equipment and shipped by automation providers that can naturally plug into industrial IoT applications. But there is also data fed out of all types of dimensions of the pyramid, and new applications are built in with lots more data that, uh, that needs totally different types of computation and technology and mechanisms to generate insights and analytics.
ALAN BEHRENS [00:05:27] Right. Excellent. So as I said earlier today we want to talk about platforms and, you know, the term platforms is a slightly amorphous. You know, we hear the word platform in many different disciplines. And as an analyst, you know, I hear across many companies in different contexts. So what I'd like to understand from our guests here is their view of what a platform means to in the world of industrial Internet. And then we'll go on talking about some of the attributes and some of the decision making factors in selecting them. So, Diego, why don't you start us off and give us your view of what is a platform to to you?
DIEGO TAMBURINI [00:06:07] Yeah. And of course, Allan. And I agree that is a it's kind of over used terms and an every every single person probably provides a different a different interpretation. But in my view, a platform, a software platform is a set of services or capabilities generally exposed as APIs and host it increasingly in the cloud that can be used to build or stitch together applications so that the basic the basic characteristic of a platform is that you can build stuff on top of it. Why are platform important to do IoT specifically? Because it will be very difficult prohibitive for for most to to do IoT without the cloud, without platforms because you are dealing with potentially thousands of devices located all over the planet streaming huge amounts of data that needs to be stored, processed and recent order. All the while you need to ensure security across firewalls, availability and reliability of the messages. So you could conceivably do that without a platform, but it will be extremely difficult to do it right.
ALAN BEHRENS [00:07:27] Right. Excellent, Josef.
JOSEF WALTL [00:07:30] Adding to that, I'd like to come to a different, let's say, dimension of a platform for me that the platform is a common set of services. And with all the technology and that has been developed by our companies, it's easier for customers to build just a little bit on top they need for their applications. For the point of commonality. I think there is exactly the difference between an IoT platform and an industrial IoT platform, because an IoT platform is common across all types of verticals. So it provides security, provides streaming of data, provides analytics, which is true for different verticals when it comes to industrial IoT. There are specific connectors to legacy brownfield hardware, specific protocols like OPCUA that needs to be supported and all types of other certifications. There is sometimes also when we look at, for example, smartware there is differences in the legal legal structure per country. This is what our common sets for for industry, for manufacturing, but also food and beverage and so on. And there it makes sense to add another layer of common functionalities that can then help customers in a certain vertically or even a sub vertical like aerospace and defence to build on top and only did the part they really need where they differentiate towards their customers.
ALAN BEHRENS [00:08:57] Great. Interesting. And Bill.
BILL BOSWELL [00:08:59] I think that the book Diego and Josef have brought out great points and it really does get into the kind of complexity of those platforms and the ability to securely and easily connect to them. So, you know, with Siemens, with 30 million automation systems out there, 75 million contracted smart meters, a million connected products mean that that level of global scalability is another aspect that I think that we need to think about when we talk about a platform, because you need to be able to not only work on that one particular shop floor, on that one particular line, but you need to be able to use the cloud to be able to access manufacturing information that comes from all over the world in that in near real time as possible. I think the other thing from the platform that we'll probably talk about today is that notion of a hybrid environment as well, where we're talking about analytics that are done at the edge and the connection to the pyramid with things like MES, PLC systems as well as what you can do in the cloud and the platform needs to be able to seamlessly support you building applications that can work at all levels that that interaction.
ALAN BEHRENS [00:10:05] So, I mean, just take that point a little bit further. Just give us an idea as to what people need to think about when they're deciding on. So what is what sort of platform am I going to require for my application or for my my implementation of industrial Internet?
BILL BOSWELL [00:10:20] Sure. Well, I'll take a first stab at that and then pass it around. So so, you know, first of all, when you think about that connectivity layer, you need to be able to support protocols like OPCUA, but there are also dozens of other protocols that are out there from from kind of legacy devices and other specialized manufacturing kind of systems. And in my experience with the customers that we deal with every day, you are going to have a complete mix of Siemens systems and other vendors and old equipment and new equipment and and so you need to be able to be able to support the platform has to be able to support getting data from any of those systems. Oh, and by the way, including in data from the back office systems as well. So you can do the data analytics on things like ERP and PLM and supply chain management. So the platform has to have the ability to connect data adjust data for many of those sources and do that in a way that you can provide value to your users without having your company become expert programmers. And so we leverage APIs from the cloud providers like Amazon and Microsoft, and they're they're awesome. But we need to be able in this industrial IoT to bring a level of granularity to the APIs that allow you to do industry specific things. Right. So a little bit different than than low level programming. So the combination of the our domain knowledge, if you will, in those areas, along with the strength of the cloud infrastructure providers makes for a very powerful combination.
JOSEF WALTL [00:11:51] Maybe to add on that. What we also see is when we talk with customers, they either have the opportunity to work with a cloud provider and directly build on top. Then they need to get softer companies build up their that competence forever, probably. Then there is partners where you can go to a company like Siemens and say, hey, I would like to build on a higher level of expertise in that industry domain. And what we really often see is we see a hybrid scenario, meaning that there is equipment there. Industrial IoT platform has all the connectivity and there is a lot of innovation that happens in cloud providers like AWS, where customers just want to use that new type of analytics service that makes it easier and to maybe also connect to connect equipment they only have in the factory. And therefore, it's also important that the platform has a lot of interest, but it also is compatible to other technology to be combined with to generate a very specific solution for a customer easily, because ultimately customers want the solution that fills up the whole spectrum of demands and is built up quick and can profit from the velocity of innovation from different providers they are working with.
ALAN BEHRENS [00:13:12] Diego, do you want to add to that?
DIEGO TAMBURINI [00:13:14] Yeah. At the top, It depends on whether you are an end user or a software vendor. So so let me start first with what's common. Whether you are an end user or software vendor. You want to consider availability of data centres globally, latency regulations. You want to consider security, regulatory compliance standards and protocols supported. The SLAs is the service level agreements are extremely important, of course. And can you estimate the cost of running your solution? So that's kind of the considerations that are common. Now if you are an end user that who is planning to develop their own IoT solution, it's fundamentally a build versus buy decision. Is your case so unique that you need to develop a custom solution or should you buy a solution from a vendor? Do you have the development skills or the resources to hide them? If you want to develop your own solutions from scratch. In this case, you want to go the pass way the platform as a service way in which you stitch together your own solution with the building blocks that the cloud vendor provides. So in that case, you need to consider what are the building blocks available. Do they provide the stuff that you need to do? Or you may want to buy a solution from a vendor. In that case, you actually have two options as a software as a service. In which case the vendor probably made the platform decision for you. But you still care, right about the cloud vendor underneath their reputation, the availability, security, regulatory compliance. And then there is also kind of an intermediate option if you want to buy a solution froma a vendor, which is infrastructure as a service. So probably means that you are running a vendor solution such as MindSphere that runs on on on VMs on the cloud vendor platform. In this case, you have to make sure that these solutions are properly certified in the vendors VMs. And that they auctions the cloud vendor provides for the different levels of VMS. And frankly, a portability and vendor lock in. Can you can you offer it depending your customers may want that ability to go from one cloud vendor to the other.
ALAN BEHRENS [00:15:47] Right.
DIEGO TAMBURINI [00:15:48] So that's and that's if you're an end user, if you're a software vendor that you are developing software on top of the platform to sell to your customers. Of course you want to look at what are the building blocks available and what is my core competency? For example, if AI or machine learning is not your core competency and is a competencies that is very hard to do to ramp back. You may want to pick a vendor with strong AI and machine learning capabilities. Will my customers care about which vendor I use like I mentioned before and which provides me with the most flexibility. For example, can I leverage say containers to offer my customers a choice of cloud vendors. So even though we're talking about platforms and the assumption assumption of a platform is that you are going to build a software solution on top of it, the platform also impacts the end solution even if you are just subscribing to a solution from a vendor.
JOSEF WALTL [00:16:55] I think you're absolutely right. There is, um first of all the functionality and then there is also at least four big customers when we talk about industry. Also the interest. What is underneath a platform like a MindSphere? I think especially for for software vendors. There is always the decision of, do I want to go fast do I want to use maximum building blocks where I then would give up for lets say parts of what we call micro services, the complete independence. Whereas whether I want to cater to the customer request, that would have a strong opinion. I think there is no right or wrong. There is just it needs to be made up and there is opportunity cost. If you stay as independent and there is a possible speed to market and maybe a discussion with your end customers for the other route, and I think as long as these considerations are taken into place, I think vendors will be in good shape to take the best of the technology that it's provided from from the larger platform vendors.
BILL BOSWELL [00:18:07] And I'll add onto that. I think that, you know, your definition of what's the right platform for you depends a little bit on maybe your company's size as well. And so, you know, the ability of a large kind of enterprise customer to want to have a platform that they can go in and maybe in conjunction with their partners or even with their own IT groups, they can customize and extend with building their own applications on top of it is one thing when you get into the small and medium sized companies that need a platform that they can get out of the box applications or work with a global partner ecosystem to be able to develop applications for them that can just be deployed through a store and be able to start operating. That's a little bit different characteristic of a platform as well. What we've been able to do, working with the companies that we've talked to here is hopefully provide that balance, of kind of managed platform as a service or managed service, right with out of box capabilities while still providing that ability to have native accessibility to their infrastructure environments to find that right balance so that you don't have to pick if you're an enterprise customer, you don't have to completely roll your own. You can start with a platform allows you to take advantage of the native capabilities of cloud to extend it. If you're a small or medium sized company, you have access to those out of the box application and a global partner ecosystem that can can support you on it. So.
ALAN BEHRENS [00:20:31] I mean, talking on this topic of the small versus large companies. Let's go to the small, because I know that they're the ones that, you know, a lot of this type of technology is very new to them. They have very few software engineers sometimes they have none. What are the specific considerations that perhaps your words of wisdom you can offer to the sort of smaller companies?
BILL BOSWELL [00:20:53] Well, so so I guess words of wisdom are regard to get started. You really need to start with an end in mind, have a small project well-defined that you want to take on. Maybe it's in simple condition monitoring, maybe it's in a bit of a smart maintenance or a predictive maintenance kind of environment, but really pick a project in mind. You don't have to invest with the cloud infrastructures, a lot of capital in terms of trying it out. Many of the devices you want to connect already put out data in the format that you can just feed it into the cloud with easy gateway capability as well. So, you know, get started with the applications that are there and prove business value to yourself. And then if you need to go on and and add additional applications or extend the applications, you have the freedom to do that. But it's get started and try it because you can bet if you if you're waiting, your competitors aren't. So with the cloud and the ability to kind of just jump in there and try it out, go.
ALAN BEHRENS [00:21:52] Right, Diego. Any any added comments to that?
DIEGO TAMBURINI [00:21:58] I second the Bill's advice to to start small with something that is not probably business critical, but but proves the value. Identify a small team, motivated team, agile, a multidisciplinary team and pick a problem to solve. And and start with a with a value in mind. So do proof of values instead of proof of concept.
ALAN BEHRENS [00:22:28] Right, Josef? Any anything to add?
JOSEF WALTL [00:22:31] I think totally agree to what you said and what we very often see is that as soon as even long term employers of manufacturing companies get exposed to the to the technology, they get really, really excited. So my advice would be of a trust in the innovativeness of your own people. We have great stories where people really jump on it and really want to do something totally different. And the other thing is, once you have your proof of value customers jump on, you might find something really interesting. So design in a way that is scalable, that it's capable, that it's secure right from the beginning. Our companies will help the customers there. It's no rocket science. It's done before. But as soon as you have a customer that is global and this is true for almost all customers in the manufacturing space, many of them, then you need to make sure that you do not have additional cost is scaling it up. And that's easy, especially with the cloud technology.
ALAN BEHRENS [00:23:36] Right, right. So sort of on a sort of final point. I'm really keen to understand if there are companies that are sort of on the edge and making decisions on whether or not to invest in the IoT, and they they're considering platforms. I'm not interested in why they would choose yours, but what sort of decision making process do you think they'd need to go through to to decide what's right for them? Especially smaller companies.
BILL BOSWELL [00:24:00] Sure. So so I think that I'll kick it off. I think the connectivity is the base commodity. Right. So does it connect to the devices that you need to support? So I think openness in those environments to be able to support mall protocols, to be able to support open format is is key to that. I think doing that in a secure manner is is critically important to that, because if you're going to be opening, you know, being able to send information up and take advantage of the business value you have of being able to see that from your locations all over the world, you need to make sure that you're doing that in a totally secure environment. I think that's one of the things that that working with our partners that, you know, manufacturing companies tend to be slow to adopt to new technologies. And so there is often a few years ago I had to go in and explain to companies why the move to the cloud was important. And I would often get asked the question, well, I don't want to do it, because inside my firewall is more secure. And my answer to that is in working with companies like we have here today is that they have thousands of people that are dedicated to working on security, monitoring security every day. How many people do you have in your in your IT department or in your OT group?
ALAN BEHRENS [00:25:09] Good question.
BILL BOSWELL [00:25:10] And and so I think that the ability to do those kinds of things in a secure environment in the cloud is just it's a given. It's been proven. And but that's a key platform decision. You need to make sure that you go with somebody that has that proven technology and then, you know, choose that platform that you can get started on and add applications to easily to go. But I think that connectivity, security, openness is are the key fundamental pieces.
ALAN BEHRENS [00:25:38] Josef.
JOSEF WALTL [00:25:39] I think adding to those definitely basic requirements. I would add to two things. One thing is speed of innovation, because every time of platform IoT platform or industrial IoT platform, they will evolve rapidly, especially if we've seen that the industry will tend towards several larger platforms. I think the speed of innovation will increase because it's just a number of different applications. That's built on top of it. And the other thing is look at not only the technology and the innovation, but also your access to that innovation. That is that will ability of that platform around the world. That is number of partners. Numbers of applications, because that ultimately will define the individual value a company can take out of that innovative platform that it wants to decide for.
ALAN BEHRENS [00:26:39] Thank you. Diego.
DIEGO TAMBURINI [00:26:41] Yeah, I will only add that, that when you are considering a platform to develop your solutions, that the main the main question you have to ask yourself is that do you have the development skills to do it yourself? Is it is it worth it to do it yourself? And consider that the closer you get to the to the iron if you will, that the lower you go into the solution, the platform stack you have more control. You can do whatever you want, basically. But but it requires more knowledge you you have to make architectural decisions from one storage system to another from one data ingestion mechanism to another. Whereas as you go up in the stack more towards the solution as a service, or infrastructure as a service solutions the vendor would have made those decisions for you. It it it removes control from you, but it also removes the concern. So you really have, I don't think it's a matter of being small or large. It's more a matter of being having the development skills and the appetite to develop your own solutions. And then you have to manage and support them. So and see exactly in which point of their of solutions spectrum, you want to be all the way down to the iron or all the way up where you just subscribe to an application.
ALAN BEHRENS [00:28:20] Great. And I think, you know, we'll bring it to a close. I think that we've had some really interesting insights from all of our three guests. I hope that you found this very useful. We're going to have another four episodes within the series. So keep your eye on the website. And I'd like to thank you, Bill, Diego and yourself, Josef.
JOSEF WALTL [00:28:42] Thank you.
ALAN BEHRENS [00:28:43] For coming along and being with us and thanks for your time.
DIEGO TAMBURINI [00:28:44] Thank you for having us.
JOSEF WALTL [00:28:46] Thank you.
DIEGO TAMBURINI [00:28:47] Thank you for having us, Allan.