[ad_1]
Azure and IoT ads mimic Microsoft Build 2019, including Azure IoT Edge, Azure SQL Database Edge, an AI and robotic toolbox that uses Azure tools, Kubernetes updates related to Azure, and more. VentureBeat spoke with Sam George, head of Azure IoT at Microsoft, to provide a more complete picture of the company's IoT landscape.
We began by asking George to give us an overview of the current situation of Microsoft and Azure IoT.
Sam George: In the early years of the Internet of Things, we spent a lot of time making the Internet of Things possible, especially for early users, because it's the market phase. We find that the market is changing between early adopters and the early majority. And that begins to move through the chasm.
Much of what we have focused on over the past two years has been to facilitate the internet of things. You already have a wide range of IoT features – tiny microcontrollers on the Azure stack in between. We have a wide range of platforms offering for connection and management of data devices and analysis, IoT data business process integration.
Among the most interesting things we have done in the past two years, we have taken more time to bring value to partners and customers. We started with solution accelerators. So you can set up an operational IoT solution in your Azure subscription in minutes.
We now offer a software offering as a service to partners and customers. And that builds on what we've learned in the early years of IoT, that's a large percentage of these IoT solutions, sort of – you know, they do the same thing , no? They monitor devices to learn more about what's happening in a physical world business, they store data, they get insight, and they trigger business processes.
And so that this information really allows us to offer a SaaS offer really meeting the common needs of IoT. Today, we call this SaaS solution offering IoT Central. In IoT Central, you can press a button [and] provide an application – [it] takes about 15 seconds. It's a configuration system.
And it only takes about an hour to customize it for a device or to configure it for IoT Central. And then the advantage is that when you connect devices, we automatically adapt the system. We carry the pager, we keep the whole process running – both the application and all the underlying services.
At the beginning of the IoT, you spent a lot of time creating an operational dashboard and setting up monitoring rules, and so on. And then we made it easy.
What we announced this week is called IoT Plug and Play. Because we made the cloud very fast, the device is still slow. And what I mean by that, even though I can access a cloud-based configuration system with IoT Central, I'm still going to write a lot of code on the device, right? And, indeed, there is a close link between this code and what the solution is waiting to receive.
So, for example, if the solution is waiting for vibrations, humidity, accelerometer values, etc., I have to make sure that the device sends the exact data in the same format. And so there is this close coupling between the two.
You've seen a lot, many years ago, before Windows was plug-and-play; device providers should build a lot of code, a lot of software on the device. It also had to provide Windows software, and both had to match. And that sounds a lot like the IoT today.
The biggest difference is that, first, it's open source, multi-platform, it works on any operating system, [it’s] multi-language, and all that. So it's a big step forward. And so I showed a demonstration in my presentation yesterday, in which you take a device that Azure IoT Central has never seen [and] turn on that. It connects IoT Central. It seems. He begins to send. He starts collecting data. That's all. Then you only define a few monitoring rules and you use IoT. So, you know, this is the speed at which we help our customers adopt a production-level IoT solution.
VentureBeat: How has the rise of intelligent intelligence affected what you need to do as an IoT group?
George: We have seen this turn on the road several years back. And that's why we introduced our Edge Support with Azure IoT Edge. It's already widely available, customers are already in production – it's going very well.
I'm going to tell the story from the point of view of a customer who started with IoT, then moved on to Edge. It's quite interesting, because it says a lot about the evolution we've seen over the past two years. A few years ago, when we announced IoT Edge, we talked about a customer who was using Azure IoT – Sandvik Coromant, which manufactures metal cutting machines. They had developed – in the cloud – a machine learning model to predict whether or not there was an imminent failure of the machine, because it would be damaged. And when it worked, the part that worried them was, well, what happens if there is a LAN failure. [Now] the machine is vulnerable because we do not know when to close it.
And the funny thing is that they did not know it for decades, but now that they knew it, they wanted it to be repeated all the time. Initially, they built a second implementation on the device itself. When we looked at this (you know, we started talking about advanced computing), we thought, "Why can not our customers use Azure services or their own services anywhere they want?" , whether physically world or azure? ". And for us, it was really part of the genesis of IoT edge.
And so fast today, they can use this same machine learning model, run it on a device and make it work – we just add a definitive offline support. To connect, download a deployment. Cut the [internet] link? He keeps running.
VentureBeat: You can update it, but it has been built in such a way that you almost do not need to update it.
George: But that's your case. It's a fascinating thing – a kind of second part of the story. So, the model used looked like this: an IoT device sends data to the cloud, learns from it, creates a machine learning model, foresees something, does not it? Now, what it looks like is, it picks up where this machine learning model stopped.
Now […] I deploy it on the device, which runs at very low latency, with network fault tolerance and all the rest. Then I periodically send more data to the cloud and reformulate models because the machines change over time. So, something that predicts a machine failure today will not be in two years because the machine's characteristics are changing. This is a loop between the smart edge and the smart cloud.
VentureBeat: A virtuous cycle.
George: That's right, yes. It really is. And in fact, the second part of what I demonstrated is an IoT edge device connected to IoT Central. And we have also enabled plug-and-play. So, when it connects, it streams video with artificial intelligence overlays.
I've shown a worker safety scenario, where you have a camera in an OSHA environment, where you want to make sure people are wearing a safety helmet, and who can detect if anyone is doing it or no, then help to remind them that it's time to don the helmet.
VentureBeat: Is it easy to [implement IoT solutions]? Because some things require coding …
George: Cognitive services are – somehow [the level of the] average knowledge worker can do [it].
We did a demo – I think it was the last version – with Scott Guthrie in our speech. We had a small IoT device capable of managing one of these cognitive services. And we built a machine learning model using cognitive services that we called "Scott or not?" So there are a lot of pictures of Scott, then pictures of other people. And then we labeled them. On a website, you click on "Create a machine learning model". And it took a minute or two. And then we downloaded that as a container and we deployed it.
VentureBeat: Its basic function is therefore quite simple. But what about the creation of useful things? How complicated is it? How long does it get complicated?
George: There is a big gap between, you know, "I'm building a cognitive service" and "I've built something that can win chess." Because there is a spectrum of AI, but that's been fascinating to see how much we can do things like the cognitive service benefits. And it's not just a personalized vision, it's also about understanding speech and language, text and, you know, optical character recognition … you can do a lot with that.
VentureBeat: I think I often thought that intelligent intelligence and AI were to some extent identical. But they are a little different.
GeorgeWhen I think of these new waves of IT that help customers transform themselves, a customer wonders, "How can I improve my business?" Then there is a set of techniques that allow them to do it – IoT, Edge, AI, they are all very connected. So, I do not see them as separate domains, I see them all to help companies transform themselves. I really do.
For example, IoT is a prerequisite, in many cases, for edge computing. You must be able to manage the device and talk to sensors. Then you can start to put workloads on it. So I can start to understand this data locally. And then, once I work on the periphery, I want to deploy the AI and take advantage of the features that it provides.
It's a stack of layers, composed. There are these waves of computers that all occur at about the same time and are compensated for only a little. And that provides really important game changes for businesses.
VentureBeat: In terms of penetrating the potential of these layered technologies, where are we?
George: What I can say is that we really see the market moving between these early users and the early majority. The types of clients we talked to today are remarkably different from those we were discussing with just two years ago. And they are looking for very turnkey solutions, or a weak entry barrier to do it themselves, whereas many of the previous clients were sort of pioneers and early innovators.
I would say that it is definitely the innovators and the first users who are currently in production. And there is a very large number of them who understood that this would have a significant advantage [to] my company. And that has.
VentureBeat: Is it mainly business customers at the moment?
George: The early years of the Internet of Things, we met many business customers. But now that we are making things easier, we are starting to see more and more small and medium-sized businesses appear. And also consumer.
[ad_2]
Source link