October 2017 - IoT | Digital Transformation | Data Protection & Privacy

How to Choose the Right IoT Platform for Your Needs - Transcript

IoT specialist and Director of IoT at the eco Association, Dr. Bettina Horster, from VIVAI Software, explains the questions companies need to ask when choosing an IoT platform, and offers tips and criteria for decision-makers.

IoT Platform

© chombosan | istockphoto.com


DOTMAGAZINE: Bettina, let's talk first about IoT platforms in general. How do they function and what is their purpose? 

DR. BETTINA HORSTER: You can think about an IoT platform like a big birthday cake with several layers. You have a layer for the security and then on the next layer, you have the devices and gateways. For example, if you want to connect to your car, there must be something in the car to record what's going on. The next layer is connectivity. You need to connect your devices to the Internet somehow – over the air or by wire.

And the next thing is that all the data has to get into a database. The database can be for a special purpose like logistics or medical things or it can be completely neutral, and you have to implement the applications by themselves. Mostly you have these generic databases that you can apply. 

Then, although you have all the data in your database, that doesn't do you any good. You need to analyze the content and you have to draw your conclusions out of the database. On top of that, you have applications – that can be big applications, micro applications, whatever. But this is where you need the data. The IoT platform most often offers all these layers and that's basically the function

DOT: What in your opinion are the most important criteria for a quality IoT platform? 

HORSTER: Well, first of all, you have to decide what you want. But there are a couple of things which are really important. The onboarding must be really easy. It doesn't do you any good if you need a year or two to do the basic steps. So, it must be easy to do the basic steps, and then we have an agile process to do more and more with your IoT platform. 

And it should be fairly easy also to onboard the partners of your ecosystems onto the platform. Then, it should be scalable. That means that it must be very easy if you need something just for a little while, so, you need lots of capacities that you don't need all the time. 

The integration of different systems, for example devices and standards and gateways should be fairly easy. So, the better platforms have libraries where they already support various devices and standards. Then, you have to see what other specialties of the IoT platform exist. For example if you are into Industry 4.0, maybe you take somebody who is really close to that kind of topic.

DOT: Thinking further about specializations, are different platforms better suited to different application scenarios, and if so, what are the differences?

HORSTER: Well, I think the best thing is if you look at what they did before they went into this topic of IoT platforms. For example, if you have Microsoft with the Azure platform, you can be pretty sure that the identity management is pretty well done, because that's the thing that they can always do very well. That's their specialty to begin with. 

And this is also true for the IoT platform – if you take, for example, SAP, you have somebody who's really good regarding your ERP data and, for example, combining your invoicing with the IoT platform or, if you order something, your order systems. Or if you look at Siemens Mindsphere, for example, they really come from a background where they control machines all the time, so they have the specialty in Industry 4.0. Whether they are the best, I don't know, but it's always good to look at it. 

Or for example, take Bosch. Bosch a-SI comes from sensors. Bosch is one of the biggest manufacturers of sensors, and of course, they are really good at connecting various kinds of sensors to the platform. So, to my mind, it's a very good idea to think about what you need the most, and then you look at the platforms. But of course, there are various other factors that need to be checked. 

DOT: How can a company choose the best IoT platform for their requirements?

HORSTER: Oh gosh. This is not an easy task to do because there are so many things to think about. We, as VIVAI Software AG, boiled it down to only 27 criteria that are important to check. And once you check them all, you have an idea of what's good for you. And then when you have all your criteria, you have to get a good deal with the IoT platform manufacturer – with the business model of the IoT platform manufacturer. 

If the business model of the manufacturer is not good for you, because, for example, you pay for every device you connect and you do smart home services, you have a problem. If every device costs you fifty bucks or so, that can't be a solution for you. So, these are all those things that you have to take into account. 

One other thing is the flexibility of the solution. How flexible is it? Is the architecture open for other programming systems? For example, if you want to use an analytic tool from another manufacturer, is that even possible? Or if I have a solution from an industry sector that you want to involve, is that possible? Or how difficult is it? Is it closed? Is it open? 

And one thing that most people don't think about is what happens if I'm not happy with that IoT platform anymore? Do they help you? Do they have a change process to help you to migrate? I mean, I haven't seen this yet, but I would apply hard pressure to get something. Or security. Is there an end-to-end security? And I'll come to this a little bit later, but it's really tough to get good data security, although they all have security-by-design. 

And then localization – for many, many German manufacturers, at least, it's important to have the data protection "Made in Germany," because it's really strict. So, a lot of them don't want their data to go somewhere where they don't have control. This is especially true if there's a lot of things going on, and if your data centers are working at their limits, they divert the data stream and maybe send it to the United States, or whatever. Or China. Who knows? And this is also something you have to think about. If you say “I don't mind,” fine, but there's a lot of people who do. 

And how is the support structure in these kinds of things? Is there local support that you can visit if you really are in big trouble and it can help you? Or face-to-face? Or do you have somebody that you have to chase until you get an answer from somebody? I would really check that. These are all new platforms. None of them are that old and there can always be problems and you need somebody to help you if there's a problem. 

And then we have, on a broad picture, what's the market position? If you have a good brand and if you're big and you have a lot of resources, you get a lot of marketplaces and there will be interesting partners who write their own software. And you can have access to the libraries, which is a very interesting point – because you don't want to do everything by yourself. So this is why that's a big company business to my mind; it's very, very difficult for smaller companies to get into this. 

And this comes to the next point. How strong is the company? Because you hope that you will have a long term relationship with this company and it's good if they are financially really healthy, and if they have data centers all around the world, and development all around the world that you can follow with your globalization, for example. 

So that's also a thing to look at. And, as I said, do they have marketplaces where they offer really generic solutions that all other people can use? 

So next, a really hard thing, do they support you with your business models? Because most people are not aware that they are mostly working in ecosystems, and with ecosystems, business models are getting really complicated. If they support you, this is a very good thing. 

Next thing – I already talked about this – is the licenses. How is the license model and how stable is the license model? I mean, we had one example where the license model was completely changed and it made all the business models from their partners really difficult, in the end, and then they had to drop back. 

And then, who are the reference customers? Do you know anybody? Are they in the same area as you are? And then, when was the launch of this platform? They are all pretty young, but there are some platforms that already have some experience from cloud services, for example, which makes it a lot easier to draw from their experience. And then, how fast are they if they have to give new features or updates? So these are all the questions I would ask somebody if I were to consider this company for my club platform.

DOT: So what should platform providers do to support companies better?

HORSTER: First of all, it's very important that if you have a question they are here for you, and that you don't have to wait for days, and sometimes for weeks, to get an answer. You need somebody that you can address and this should be ideally one person who's responsible for you. And if you get this, this is already very helpful. They can help you if they have good managed libraries and marketplaces. I think these are the things that really help the companies and if they help you with the business model, where they make business model workshops or something with you, that's even better. 

DOT: Lastly, who has or should have control of the data on an IoT platform? 

HORSTER: One thing I think is really important, and that is something that nobody really takes seriously, and this is the data usage control. Data usage control means that you give the data autonomy back to the customers. Especially for B2C customers, it is really tough sometimes because they don't know what will be done with their data. And also, if you are in ecosystems, sometimes your partners don't know what is done with all the data. And the problem is that there's no trust. 

So, in order to gain trust, which for me is vital to creating a good ecosystem, you have to manage the data properly. So, you have to inform people where the data goes, what data is transferred, in what granularity. You don't need it always bit by bit – it's also good if you anonymize it a little. So, you have to give the user more control. If you give the user the control of what is done with the data, there will be more trust and then it makes the business much, much easier. 

My criticism is, we looked at all the platforms, and this data usage control isn't taken very seriously by the platforms that I've been looking at. And this is something they really have to take care of. In our project that we are doing right now, we need it so badly, and everybody just shrugs his shoulders and says, "We don't have it, sorry. You have to do it yourself". And I think this is a very, very important point. If you think about the new EU laws about data protection, you really should come up with a solution to solve this.

Dr. Bettina Horster is Head of Business Development at VIVAI Software AG and is responsible for the area IoT (Internet of Things, Industry 4.0, M2M), and Director of IoT in eco – Association of the Internet Industry. Since 2009 she has been involved with the Internet of Things and is a pioneer in the field. She is currently also project leader for the IoT- EU project “Smart Service Power – technikgestütztes, altersgerechtes Wohnen im Quartier” (for IoT-assisted aged care in the own home), and recently won the UN NGOs Diplomatic Council “Information Society Award”, and was finalist in the UN  “World Summit on the Information Society” prize.

Please note
: The opinions expressed in Industry Insights published by dotmagazine are the author’s own and do not reflect the view of the publisher, eco – Association of the Internet Industry.