Microsoft Azure for Industry Podcast

On the Cutting Edge with Microsoft Azure IoT Featuring Sam George and Jeff Cornwell

Episode Summary

The Internet of Things (IoT) has taken off in a huge way and is impacting industries far beyond the traditional use cases. In this episode we discuss modern IoT solutions and capabilities including managing large scale IoT implementations, Microsoft-specific devices, using Azure IoT services, IoT on the edge, developing IoT solutions, and even running containerized services on the edge. This episode will jumpstart your IoT initiatives and give you a comprehensive picture of Azure IoT.

Episode Notes

The Internet of Things (IoT) has taken off in a huge way and is impacting industries far beyond the traditional use cases. In this episode we discuss modern IoT solutions and capabilities including managing large scale IoT implementations, Microsoft-specific devices, using Azure IoT services, IoT on the edge, developing IoT solutions, and even running containerized services on the edge.This episode will jumpstart your IoT initiatives and give you a comprehensive picture of Azure IoT.

Show Links

Guests

Sam George - Corporate Vice President of Azure IoT at Microsoft

Sam is responsible for Azure IoT Central, Azure Digital Twins, Azure IoT Hub, Azure IoT Edge, Azure IoT Hub Device Provisioning Service, Azure Maps, Azure Time Series Insights, Industrial IoT & Manufacturing, Automotive / Microsoft Connected Vehicle Platform, Smart Buildings, Smart Energy. Additional responsibilities include IoT business strategy and results for IoT in Azure, IoT product offerings and roadmap, developer facing features, integration with other Azure services, partner and field enablement and more.

IoT is an incredibly exciting and innovative area of computing and I'm proud to a be part of this massive transformation.

Follow Sam on LinkedIn

Jeff Cornwell - IoT Business Development, Microsoft Cloud Marketplace

Established business development, partnerships, product and strategy leader with proven success in driving growth and scale for consumer and commercial experiences across emerging platforms. 15+ years of experience structuring and negotiating high-impact deals from source to execution, developing productive and lasting partnerships.

Follow Jeff on LinkedIn

Episode Transcription

David Starr:

Welcome to The Azure for Industry Podcast. We're your hosts, David Starr and Paul Maher. In this podcast, you hear from thought leaders across various industries discussing technology trends and innovation, sharing how Azure is helping transform business. You'll also hear directly from Microsoft thought leaders on how our products and services are meeting industries' continually evolving needs.

 

David Starr:

In today's show, we're talking about the Internet of Things, or IoT, with two preeminent experts in the field and learning how Microsoft is making deep inroads and new plays with IoT offerings. First of all, I'd like to welcome Sam George, who is Corporate Vice President of Azure IoT at Microsoft. Sam, welcome to the show, and I wonder if you could tell us just a little bit more about your team?

 

Sam George:

Thanks, David. It's great to be here, first of all. Yeah, I represent a pretty significantly large team now at Microsoft that builds all of the different Azure IoT products. We started our journey in IoT about six years ago and we've made some great progress and I'm excited to be here and talk about all of that with you today.

 

David Starr:

Then, we're also joined by Jeff Cornwell, who's the Global IoT Category Lead for Microsoft's Commercial Marketplace. Jeff, what's your role and your team's mission?

 

Jeff Cornwell:

Hey, good afternoon, David. Thanks for having me. My role in a nutshell is I lead our IoT Category for our Commercial Marketplace. If you think about a marketplace, think about in the terms of a retail supermarket in simplest terms, we have categories, and one of our categories that is key to our success is IoT. My role is really to lead my team to do a couple of things. One is to ensure that we have the best selection of IoT solutions that are built on Azure available for our customers in the marketplace. Then, secondarily, we want to make sure we're driving the success of those solutions and making sure our customers have access to them, can purchase them, and deploy them in the most automated and efficient way possible.

 

Jeff Cornwell:

Then, ultimately, we want to make sure that our marketplace as a platform has the right features and functionality so that we can fundamentally deliver that marketplace promise of a full kind of end-to-end find, try, ultimately deploy scenario. That's my world and I'm excited to be here, so thanks for the invite.

 

David Starr:

Hey, that's perfect. Thanks, Jeff. That's a good explanation. I'm going to start out with a very broad question because a lot of people wonder what exactly we mean when we talk about internet of things. A lot of people have this impression that it's a bunch of raspberry pies connected together over the internet. Sam, I wonder if you could explain to us your favorite way of explaining the IoT world, and what is the current state of the industry?

 

Sam George:

Absolutely. The way to think about internet of things is, and I did this with a friend of mine once who is in the aerospace industry, and he was asking me a similar question about how would you explain internet of things, and I said, "Well, tell me about some of the things that you do in your line of work?" He was explaining some of the things that were happening in their factories, and I was saying, "Hey, how much visibility do you have into that? Do you know in real time everything that's happening in your factory?" He said, "Well, you know, some places we do, there's some other places we don't." I said, "Imagine if you knew in real time what was happening in the physical world and all the different things that you care about. What could you do with that?"

 

Sam George:

We wind up having this discussion about these optimizations that they could make in cost savings, and then at the end, I said, "That's all the internet of things is. It's using this technique of deploying inexpensive devices to monitor things that are happening in the physical world." So, when I describe the three atoms of the IoT molecule as it were, it really comes down to three different components. There's things, and often those are devices that are again monitoring something in the physical world. In IoT, you're typically not that interested in the device. It's a means to an end, but you're really interested in what the device is monitoring.

 

Sam George:

Then, there's insights that you gain from the monitoring of those things, and then there's informed actions that you drive as a result. It really is that simple. You see those three in every single IoT solution. As far as the state of the industry, IoT's been around a long time. It was called machine to machine back before cloud, and cloud made IoT fundamentally different in two important ways. The first is it meant if I wanted a deployment of a internet of things solution, I no longer had to go build my own data center. That was a particularly onerous part of doing machine to machine. Then, the second part was now that I don't have to build my own data center, I can also take advantage of built in cloud functionality that helps me build my IoT solution faster.

 

Sam George:

For example, back in the day of machine to machine, when I was taking bare metal servers or virtual machines, I'd have to start building things from scratch. Today, you can go to the Azure portal, and you can spin up one of our cloud gateways for connecting to IoT devices, we call it IoT Hub, and you can connect to millions of devices and handle trillions of messages, and you don't have too write a line of code in the cloud to support that. That's something we take care of. That didn't exist back in the day. The industry has come quite a way. I'd say the state of the industry right now is we're just in the middle of the transition. If you look at the technology adoption life cycle of innovators and early adopters, and then you get the early majority and late majority.

 

Sam George:

We're right in the middle of this transition across the chasm between the early adopters and the early majority. We're starting to see customers that are showing up that just want to take advantage of IoT in a simple way. A lot of what we've done over the last five years and being in market was to build out the core capabilities that's required to build these hyper scale million devices send trillions of messages solutions and to be able to manage all of that at scale. What we started the journey a few years back is making that simple, and we've got a whole bunch of products that I'll talk about today that really make it so that almost any business on the planet can take advantage of IoT. We're right on the cusp of that transition from possible to easy.

 

Paul Maher:

Fantastic, Sam. I'm just listening in there. It's super exciting to get your perspectives on the space and the opportunity. Switching gears a little bit, so for our listeners of which I'm sure we'll have industry customers and partners listening in, if you were to ground this in your thinking on Microsoft's role in really thinking about IoT and the opportunity, how would you describe that to really ground people in their thinking?

 

Sam George:

When you really look at an IoT solution, an IoT solution is a business solution. It is a foundational and fundamental part of a business, and as such, we think that IoT is going to touch and transform every single business on the planet. When we think about that, that is not a winner take all market, that is not a Microsoft's going to go do this for the world. That is a market that requires deep ecosystem and deep partnerships to really be successful. We've built a leadership position over the last five years with exactly that approach. So, we look at our role, as we looked at IoT, the lens we looked at it through is we said, "What are the things that are going to be really hard? What is the toil that is going to be undifferentiated heavy lifting for all of the different vertical solutions that are going to be built?"

 

Sam George:

Building cloud services to manage two way connectivity to millions of devices sending trillions of messages, and doing that in a secure way, and enabling device management is hard, is really hard stuff. But, it's also undifferentiated if all I care about is predictive maintenance of my water pumps or knowing when a bridge is at risk and needs to be fixed. That's not what's differentiated, the vertical solution is differentiated. Our fundamental position is that we build out the hyper scale cloud and Edge capabilities, the plumbing. We view ourselves as hyper scale plumbers, and then we have a rich partner ecosystem of ISVs and GSIs and system integrators, and hardware providers that provide these vertical solutions that help you with the real business value. There's this really nice symbiotic relationship between us and our partner ecosystem in IoT.

 

Paul Maher:

I think that's really good to just ground ourselves. We think a lot about... Playing back what you just said, which is great for everyone to hear is hyper scale plumbers, I like it. Think of Microsoft as providing the technology and the services, but key to all of this in terms of innovation and also really IP is our ecosystem and our partners. I think it's important for everyone to hear that. The opportunity that we bring with the platform and services, but as Microsoft, we're a partnering company, and that's all about our ecosystem.

 

Sam George:

Right.

 

Paul Maher:

I think that's a good segue. Jeff, I know you work with a lot of the partner ecosystem and our IoT partners, would you like to share some of your thoughts about the partner ecosystem around IoT?

 

Jeff Cornwell:

You bet. It's a great question and we're seeing some pretty tremendous momentum and it's an exciting time to be in this space. I think we've seen an evolution in the IoT space, significantly, over the course of the last even 12 to 18 months, where customers are increasingly building their own IoT solutions leveraging our technology. But, so too are our partners and our publishers. Our partners are building IoT based solutions leveraging this hyper scale plumbing we're referring to and putting their own secret sauce on top of it, whether it be full verticalized solutions or even horizontalized solutions to support niche use cases.

 

Jeff Cornwell:

We have a great platform for not only our customers to build solutions with our platform, but so too, we give that capability to our partners. When our partners are building these solutions, the marketplace is a great way for them to not only get visibility for their solutions but get those in the hands of their customers really efficiently through purchase and deployment. Again, really exciting times in marketplace. It's the tail end of that process that really enables our customers and partners to connect on these unique solutions.

 

Sam George:

Let me just highlight a little bit as well. Just even recently, when the world was hit with the COVID-19 pandemic, our first reaction as Microsoft was let's go to all of our partners and let's assemble the partner solutions that will help with COVID-19 and will help businesses more safely return to work. We went and worked through our partner network. We've now assembled almost 200 partner based solutions that are everything from one of our partners can actual drop ship a COVID-19 testing station that's fully connected right outside a business. Some of those other partners can do things like, they have an IoT solution that'll help keep hospital workers safe in hospital environments. It's all from the partner network and all of those are up on marketplace.

 

David Starr:

Those are fantastic examples. You have already, Sam, used the term Azure IoT, and I also noticed that you used an example of how Azure might play into an IoT ecosystem with the IoT hub that you mentioned, that service. How does Microsoft's IoT strategy really meld with Azure? What's Azure's big role behind our IoT play and what are Microsoft's big investments in place here?

 

Sam George:

When we started our IoT journey, and again that was six years ago, one of the first things that we did is we understood what does an IoT solution look like. That's where we figured out that it's things, insights, and actions. Then, the next thing we did is we said, what does Azure already have or is about to have, and then what is Azure missing. What we did is we went about building the things that Azure was missing. A lot of what Azure has had and has now are things like great data storage and analytics systems, great business process integration systems, great abilities to build like for example with Power BI to build BI dashboards to take advantage of citizen developers with power automate and power apps. There's a lot of stuff Azure already had.

 

Sam George:

What it didn't have was the part where cloud services or connecting to and managing those millions of devices at scale, of software development kits that run open source cross platform for IoT devices or edge computing offerings. So, we went about building all of those and really focused on that. Our first click stop in IoT was to build out these platform foundational services. Those were services like IoT hub which is that cloud gateway for connecting to and managing all of those millions of devices sending billions of messages, or IoT edge which is our edge computing runtime, runs on devices as small as raspberry pie or as large as the biggest server that you want to run it on, and it runs container based workload of Azure services themselves, or even software development kits for IoT devices themselves, and even offerings like Azure sphere, which is our highly differentiated, highly secure... You can almost think of it as a SaaS offering for an IoT device to monitor in real time the security [inaudible 00:13:59] the device, provide updates for it that Microsoft stands behind.

 

Sam George:

Even now, our newest member which is Azure [inaudible 00:14:07], the tiniest of our edge offerings that runs in two kilobytes of memory on a microcontroller or 50 kilo bytes of memory or above for a connected experience. We did all of that, and the first thing that we noticed, and it's worth calling out is that when we looked at the IoT market, our analysis told us that it was really the enterprise and commercial markets that were going to stand to benefit hugely from IoT. For example, if I'm... Pick one of our earliest customers, [inaudible 00:14:41], who is managing millions of devices worldwide, or another one like [inaudible 00:14:47] who builds these food processing machines that process 70% of the world's food supplies and are always looking in realtime for toxins in that. Those are two examples where they wanted to take advantage of these services and were able to, and build an IoT solution. They're fairly sophisticated companies, but when we looked across these commercial, and they save millions of dollars doing this, but as we looked across the industry, what we saw is that the majority of customers, that large part under the bell curve of the technology adoption life cycle, they don't have these large sophisticated development shops.

 

Sam George:

What we went about doing was simplifying IoT and making it easier and easier for mass market adoption. For that, we got into our next phase of IoT, which is building IoT Central. What IoT Central is is IoT Central assembles all of the different Azure services for you for an IoT service. You don't have to understand how cloud solution development works, you simply provision an application, takes about 15 seconds, you connect devices to it, and you can connect as many as you want. Under the covers, we're orchestrating all of those Azure services for you, super easy to use, super scalable. That's been great, that's been that next click stop up. IoT Central is another important part.

 

Sam George:

Both of these, both the platform that I've been talking about and IoT Central are really geared toward connecting to, managing, finding insights from assets. We talk about how there's this evolution of connected solutions and IoT is really at the phase of connecting assets. What we're starting to see now is this emerging pattern of I don't want to just connect manufacturing equipment in a factory, I want to connect the whole factory itself, the entire physical space. I want to be able to know what are all the processes that are happening in the factory, who are the people that are doing them, what is the supply chain inbound to the factory look like, what do the downstream suppliers look like?

 

Sam George:

As you start to connect entire environments, which is that next click stop, that's what the new service that we just made public preview available of Azure digital twins is for. It's about connecting entire environments. It's built on top of our IoT, you can almost think of it as IoT for assets, services, IoT hub and central. But, it also adds this new ability to track entire environments, to be able to model those environments to track the live state of them, and then to track the past, to be able to simulate possibilities and then to also be able to predict the future. That's what digital twins is about. There's more but those are highlights right now.

 

David Starr:

When you think about predicting the future, are you thinking about things like time series data where I'm looking at predictive maintenance on my manufacturing lines and stuff like this?

 

Sam George:

Absolutely. In fact, that brings us to another one of those core IoT services, which is when you get data from an IoT device, the majority of the time it comes in a time series fashion, right? For example, if you're measuring temperature and humidity and vibration, you might get temperature, humidity and vibration every 10 seconds, and that repeats itself for days, months, years. If I connect to one device and start analyzing time series data, looking for anomalies and patterns, that's one thing. But, when I have a million devices and I have petabytes of IoT data and I want to find insights, then you start getting into very complicated data wrangling techniques, very complicated analytics.

 

Sam George:

So, what we did is we again simplified that at the data and analytics layer with Azure time series insights. We make time series insights as easy as I have an IoT hub, it's connected to a large fleet of devices, I provision Azure time series insights, I connect it to the IoT hub, and it automatically starts collecting data that comes through IoT hub. It makes it so, and it's a platform but it's also got a built in user experience. That user experience is as easy for a knowledge worker to learn as Word or Outlook or PowerPoint or Excel. It's very straightforward. What we've done is we've made it so that IoT data is now as easy to find trends and patterns and anomalies in for an average knowledge worker as it is for that knowledge worker to do something in Excel. It's very, very easy.

 

Paul Maher:

Excellent discussion. I'm going to take us a little bit back. We mentioned earlier on in our conversation about the Microsoft commercial marketplace, and so as I look at the IoT space which is clearly maturing and customers are increasingly leveraging Azure IoT capabilities if we discussed about, either directly or via partner solutions. Jeff, maybe I'll come to you. What role do you see the role of Azure marketplace, which is one of the storefronts under the Microsoft commercial marketplace, what role do you see it playing in terms of from a partner point of view, publishing IoT solutions, and of course from a customer point of view actually looking to go buy and purchase IoT solutions? I'd love to get your perspective on this.

 

Jeff Cornwell:

[inaudible 00:20:09] appreciate it, and I think from the partner or publisher point of view, we make it as easy as possible to take your solution and allow you to sell it and deploy it in the most effective way possible. Lot of tooling there to support that, to make for a great experience. But really, from a customer's perspective, I look at it in two big buckets. The first is marketplace really makes it easier for our customers to, what I referred to earlier, find, try, buy, and ultimately deploy solutions. It's our mission to make sure that partner solutions are available, customers can find them easily, try them where appropriate, purchase them, leverage their existing Azure account or EA to make that purchase, and where possible deploy hose resources directly into the Azure subscription and get going, so really increasing the ease and ultimately get them quite a bit faster time to market, so get them up and running with their solution as quickly as possible.

 

Jeff Cornwell:

There's really two sides to this as well. In a perfect world, we've got fully repeatable solutions where our customers again, buy it, and deploy it. But in many cases, IoT solutions can be highly complicated in terms of the number of components, the customization, et cetera. The marketplace also allows for certain customization, such that large scale enterprise deals can be negotiated, customized, and managed through the marketplace, so that a customer can buy via a pre-negotiated term with the right components, with the right terms through that marketplace engine. So, marketplace is increasing that efficiency and making the experience easier, not only for the sell serve experience but also in support of those larger enterprise, what we call co sale deals in our ecosystem. A lot of important capabilities here that we're giving to both audiences.

 

Paul Maher:

Fabulous, super exciting. Certainly for our listeners, we'll make sure we share some of the links in the show notes if the Microsoft commercial marketplace is new to you, and we'll hear a little bit more as we go along. David, back to you.

 

David Starr:

Sure, one of the things I've been curious about is a term that's already been referenced a couple of times, and that's IoT on the edge. Now, we know that a typical device is if you will on the edge, right? A raspberry pie was given as an example earlier, or an Azure sphere device. But, we're also moving services to the edge. Now, I wonder if, Sam, you could talk a little bit about what that looks like in terms of what the edge means today. We have different capabilities on it than maybe just the processing units themselves, right?

 

Sam George:

It's a great question. Early IoT, you could think of the topology as I have a cloud solution and that cloud solution is directly connected network connection by network connection to a set of IoT devices. Again, those IoT devices are monitoring the things I care about. What was fascinating that we saw as we were about two years into our IoT journey is that a typical example of the value that IoT brings is to predict the maintenance of assets that are important to you. So, the way that typically works is I'm collecting data about... I'll pick on an example of a water pump. I'm collecting data of a water pump that's maybe temperature and vibration. Over time, what I do is I build a corpus of data, store it in the cloud and then correlate failure points. When did the pump fail and what was the data [inaudible 00:23:40] time series leading up to that that will help predict that?

 

Sam George:

I use technologies like Azure machine learning in order to build the models that will predict those failure points. So then, now I've got predictive maintenance working, and I can tell with a very high percentage, let's say 98% or so accuracy, what the remaining useful life is of a water pump, and when it's going to need servicing or a replacement. That's great. The way that solution works is I have a device that's always sending data up to the cloud, and then as data enters the cloud, it gets run through a service like Azure stream analytics after it goes through IoT hub, and stream analytics has support for machine learning models, right in line as it's processing those data streams coming through. If it spots something that looks like your remaining useful life is going to be too short, then it sends an alert and then you do something about it.

 

Sam George:

What's interesting though is as more and more companies got comfortable with and learned to depend on the ability to predict maintenance, they started to ask questions like, well now what happens if someone in my area digs up a fiber network cable with a backhoe, and that device now is unprotected and it can't send an alert to a local technician that it needs servicing? We started seeing some customers that would then duplicate the machine learning model that they'd already had running in the cloud, and they duplicated it out on the device itself. We looked at that and we said, "That really looks like a problem. If I've got this great cloud solution model or this cloud model for developing a machine learning model, why can't I just take that same machine learning model and push it right out to the edge on that device?" Then, instead of the device sending all the data to the cloud, it's just scoring it against that machine learning model on the edge, and then the only thing that it sends to the cloud is an alert that says, "My remaining useful life change, and here's what it is now."

 

Sam George:

Then, there's also this really interesting loop where it happens then where data is then periodically bulk uploaded from that device to the cloud, and then because one of the things about machine learning models is you have to keep retraining them because as data changes, the machine learning models need to as well. So, edge computing was born out of this need to have workload portability between things that I'd already developed in the cloud and things out on the edge. We came up with Azure IoT edge several years ago now, and it's a container based system. For those that are familiar with containers, you know that containers are great for portability. What we've done is we've gone around to a set of Azure services like Azure machine learning and Azure stream analytics and Azure cognitive services for easy AI, and a bunch of others, and they are now able to containerize themselves. Then, you can deploy them right out to IoT devices running this IoT edge runtime.

 

David Starr:

That's fantastic, and brings this solution much closer to the problem. Man, AI on the edge, that's just cool. Jeff, I have a question for you with regard to that, and that is does any specific edge solution that might be available in the Microsoft commercial marketplace come to mind for you when you hear this type of solution?

 

Jeff Cornwell:

Certainly. This is one of those fast growing areas with not only a wide range of solutions from Microsoft, but also from our partner ecosystem. What's exciting is we talk about the ability to use marketplace to deploy more effectively. We have the concept of what we call an IoT edge module in our marketplace where this container resides and can ultimately be deployed. We have a ton of great solutions out there, a couple from Microsoft. We have the OPC publisher, which allows you to connect to existing OPC UA servers, some great solutions from partners like NVDIA that allow you to perform that AI on the edge that we're talking about across multiple video streams simultaneously. A lot of really great solutions not only from Microsoft from a first party perspective, but also from our partner ecosystem.

 

David Starr:

Now, let's take a moment out to listen to this very important message.

 

Speaker 5:

Did you know the Microsoft commercial marketplace allows you to find and purchase leading Microsoft certified solutions from Microsoft partners? The Microsoft commercial marketplace includes the Microsoft app source and Azure marketplace. Each storefront serves unique customer requirements and different target audiences, so publishers can ensure solutions are available to the right customers. For applications that integrate with Microsoft 365 products, visit appsource.microsoft.com. Get solutions tailored to your industry that work with the products you already use. For B2B Azure based solutions, visit azuremarketplace.microsoft.com. Here, you can discover, try, and deploy the cloud software solutions you want.

 

David Starr:

Let's talk a little bit about... We've looked at and we've listened to Sam and to Jeff talking about the technology, the partner opportunity, and super, super exciting. I'm sure the listeners out there... Sam, I look to you to start off and then I'll hand over to Jeff. Let's double click a little bit and think about from an industry point of view, let's think about industry scenarios that are point of mind for you that you're seeing that really we're driving innovation with IoT. The goal here is to try and make it real for the listeners. Let's think about business [inaudible 00:29:14] in industry and how really the technology is helping. Let's start with you, Sam, and then Jeff, you can chime in as well afterwards with your perspective. Sam, over to you.

 

Sam George:

In industry, this is where the rubber really meets road, where something that you... an IoT pattern starts to turn into a business solution. A business solution, problem that you're solving for example in manufacturing looks different than a problem that you're solving in agriculture or in healthcare. The underlying technology is the same and the patterns are the same, but the solutions are different. I've got to talk about... Let me talk quickly about a few different industries and the typical solutions that we see there, and then I also want to highlight our partner network as part of those. I'll take manufacturing as an example. In manufacturing, the things that you're concerned with are improving the quality of the things that you're manufacturing, predicting the maintenance of all of your manufacturing equipment to ensure as small a downtime as possible and coordinated downtime windows, improving the overall equipment efficiency of your factory assets. These are the biggest concerns of manufacturers, and ensuring safety.

 

Sam George:

If you look at something like agriculture, what you're doing is you're trying to reduce the amount of water that you're applying to be able to grow foods, because water is a precious and limited resource. You're trying to reduce the amount of energy that you're using, you're trying to reduce the amount of pesticides. You can use IoT and censors and even edge computing in agriculture to tell you exactly what parts of the fields need water, what parts have pest infestation, which parts need pesticides and which don't, and dramatically reduce how much water and energy and natural resources you're consuming.

 

Sam George:

In manufacturing, you're improving that equipment efficiency, in healthcare, we see a lot of scenarios around remote patient monitoring, especially in the time of COVID-19. If someone's got an injury to a leg or you see people that are monitoring for injuries and they might have diabetes, and they want to make sure that they're not getting an infection in their legs for example, there's monitoring cuffs that go around your feet that are connected back to your doctor in a way that fully respects privacy and all of the HIPAA compliance rules. But, instead of having to go into the doctor every time, the device that you're wearing is actually interacting back with your doctor's office, and then they can send you alerts when it's time to meet you.

 

Sam George:

There's so many of these types of scenarios for different industries, and the great news about Microsoft is that not only do we have deep expertise in all of these industries, we also are working with the Fortune 5000 companies that are in all of these industries, and most importantly we have this really deep partner network. As an example, I'll pick on manufacturing. We partner with some of the largest and most influential ISVs in manufacturing, Siemens and Rockwell and Schneider and PGC and there's a long list of companies that we partner with in manufacturing. As you start looking at other areas like agriculture or smart energy or healthcare, there's similar ISVs that we're working with. Industry is a really important thing because that's really when IoT becomes real.

 

David Starr:

Thanks, Sam, that's super helpful. Jeff, maybe I'll look to you. We mentioned about obviously the Microsoft commercial marketplace. Maybe you could share some thoughts on some of the partners that you and our teams have been working on, and bringing those solutions to the marketplace. Could you share some thought here?

 

Jeff Cornwell:

I would love to, and Sam, I could just listen to you all day. I think that's one of the most exciting parts about this business, is that IoT has so many use cases and is making such a dramatic impact in so many different industries and scenarios. We could have a two hour conversation on this, but just to highlight a few areas, I mean, I think I mentioned it earlier but there's some pretty impressive horizontal IoT solutions that are solving problems across multiple verticals. I think of things like connectivity solutions from [inaudible 00:33:35], their quick edge connect solution, I think of analytic solutions like [inaudible 00:33:39] doing some really impressive things. But, from a vertical perspective in our outsource catalog, we've a lot of very impressive solutions across the board that are really dedicated to solving specific problems.

 

Jeff Cornwell:

I think if I look at some of the broader categories of activity where a lot of energy is right now, we see a lot of energy and intelligent supply chain, companies like [inaudible 00:34:00] with their intelligent supply chain solution doing some pretty impressive things. I also think about companies like [Mariner 00:34:06], who are doing spyglass visual inspection, a visual inspection tool that's cloud based, leveraging AI and IoT to improve the quality of products across multiple industries. So, just a ton of examples from asset management to connected products to infrastructure to even vehicles and mobility. A ton of great solutions that are really exciting, and quite frankly like I said, this is why I get excited. There's so many more ideas and solutions that are being addressed that we don't even know about yet, so great time to be in IoT.

 

David Starr:

You know, I have a question about security because that's one of the things I hear talked about most in the world of IoT. In fact, you can talk with any given chief security officer, and they might fall over when we start talking about IoT solutions. What would alleviate some of those concerns for people?

 

Sam George:

Yeah, I'm glad you ask. You can't talk about IoT without talking about security. It is a very important topic. The good news for our partners and our customers is it's one where we have brought our full expertise to bear. Microsoft has this long history of being very security focused, starting back in the day with Windows XP Service Pack 2, and it's in our culture now. It's in our engineering culture, from when engineers first join us to the most senior levels of leadership. In IoT, we took a very firm stance when we first set out on IoT, and one of the things that we said from the very beginning is, number 1, from our cloud gateways, connecting two devices, we're going to only support highly secure connections. We had customers at the time that were saying, "Hey, I'd like something that's not using TLS security for example. I'd like to do unencrypted," and we frankly just said, "No." Because we said, "That would be a really bad idea," and I'm glad we did.

 

Sam George:

We've had this posture of secure by default and to push every single security technique that we can bring to bear. I'll give a couple examples in addition to only supporting TLS connections down to the device. For example, sporting X.509 certificates for device identity. We brought Azure security center to market so that Azure security center, which works now with our IoT services, with Azure security center for IoT, actually does active monitoring now of your IoT and IoT edge devices. It monitors them both at the device level and at the cloud level. As an example, when you deploy an IoT device and you turn on Azure security center for IoT, it will tell you, it will warn you if your device is starting to behave in an anomalous way. For example, if that device all of a sudden opens up a network connection to a known bot net that we know about in the cloud, we will alert you. That all comes from all of our security operation teams that operate worldwide monitoring everything that we do worldwide, and all of the different offerings that we provide.

 

Sam George:

The same time, if let's say for example, you have an administrator account that's managing your cloud services and it's backed by active directory, and we notice that there's been a spurious logon, for example you had a simultaneous logon in two different parts of the world, we can flag it immediately and say, "Looks like you have a compromised account." So, it really is this, we've gotten to the point where it's not just best practices, but we have proactive security monitoring. Then, we just acquired a company, we announced this recently. We just acquired a company called CyberX, who is a cyber physical security expert, and they actually do really sophisticated... We're integrating that in Azure security center for IoT. They do very sophisticated network inspection, so they can tell number one, what are all the IoT devices on my network because quite frankly if you're in enterprise, you probably don't know all of the IoT devices that are on your enterprise, that people have just brought with them.

 

Sam George:

So, it'll tell you all of that and better yet it has a catalog of all of the different behaviors that are normal for those types of devices and it'll warn you when it's behaving abnormally. We're really getting to the phase where we're super proactive. With Azure sphere, we've gotten to the point where we have, from the very ground up, from silicon to operating system, [inaudible 00:38:38] to a security monitoring service. We have an end to end turn key package. Microsoft stands behind the security of a device. I honestly can't think of another company that's doing more in security in IoT than what we are.

 

David Starr:

Speaking about standing behind the security of a device, even throughout, I believe it was $100,000 challenge to hack Azure sphere, that was a fun program just to illustrate the multiple levels of security on that device, of which there were many including hardware level.

 

Sam George:

Yes. Yes, very much.

 

Paul Maher:

Excellent, and hopefully no one won the challenge just to be clear. But, so security of course is key and integral to us all. Sam, I'm going to put myself in the footsteps of some of our more experienced users out there that perhaps have familiarity with IoT and some of challenges. I'm thinking about, could you share a little bit from your perspective on the work we've been doing? On how we're solving some of the gnarly problems, so I'm thinking about we have a proliferation of devices out there that don't necessarily talk to each other. There's problem on protocols and standardization and being able to manage the device's state. These I think are some of the bread and butter, some of the key issues, that are hard and gnarly, and I'm sure some of our listeners are hoping that we've cracked the code on, or at least making lives a little bit simpler for them. Could you share a little bit of what we've done there and how we're making lives easier if you will?

 

Sam George:

Very much. Let me give a bit of an analogy. One of the biggest problems that we see in, I'll call it, the early IoT was when I build a solution, let's say again, I'll go back to that water pump example. When I have a device that's monitoring that water pump, I built software that's on that device that is hard coded to monitor those particular sensors, temperature and humidity. I'm sending it to the cloud in a certain format and then the cloud is expecting that format. It's going over a secure connection and everything, but there's this hard coding between the software that's running on my device and the solution that's listening to that device or sending it commands in the cloud. The communication between that is already fairly standard, right? It's typically in QTT over a TLS connection, and the world's starting to standardize on [inaudible 00:41:04] QTT for IoT. But, the real problem that we saw was the hard coded nature between the software on the device and the software on the cloud.

 

Sam George:

We saw a really similar pattern years ago. For those listeners that are old enough to remember, when Windows first came out, this is before Windows 95, when you wanted to run a peripheral, let's say you had a microphone for example and you wanted to use that microphone with your Windows PC, you had to install software first on Windows. That software knew exactly what to expect from that microphone. Of course, that was a fine first step, but eventually and years later after it all started working super smoothly, we had this thing called plug and play. What plug and play at it's simplest was was a capability model where a device could self describe, and when the device plugged into Windows, it would say, here's who I am, here's what I do, and Windows knew how to interact with it. That's where the analogy stops.

 

Sam George:

The thing that was interesting about it, was that the device had a capability model that it sent along with the connection to Windows. It was just one of these big aha moments that we had a couple years ago that IoT was fundamentally missing that and that that was a problem. So, we came up with this approach, we call it IoT plug and play because everyone knows what it does just from listening to it. The way IoT plug and play works is it enables you to create a schema of the interaction pattern for that device, specifically the telemetry that the device sends, any properties that it reports, any state that it synchronizes or any commands that it receives. You can model that in a JSON-LD schema, and then once you've modeled that, then we can do a whole bunch of really interesting things with what's effective metadata that describes an interaction pattern.

 

Sam George:

The first thing we can do is we can take visual studio code and generate software on the device that adheres to that contract, that interaction pattern. Then, all you have to do is write a little bit of code to listen to the sensors themselves. The other thing we can do is as that device connects to the cloud, in particular to solutions that support plug and play like IoT central or like partner solutions, what it means is that as the device connects to that solution like IoT central, even if IoT central has never seen that device before, it can automatically populate a dashboard for it, it can automatically start collecting data from it, and it can automatically enable you to send commands and synchronize state with that device, all without writing any code.

 

Sam George:

Why that's interesting is it dramatically reduces the amount of time that it takes to build software for the device and software for the solution, but it also means that our partner ecosystem can now, which they've started to do, produce plug and play devices. With a plug and play device, what it means is that I can use the same device across multiple different solutions. For example, that device that was monitoring temperature and vibration on a water pump, now all of a sudden, I can start to use that exact same device to perhaps do predictive maintenance for a washing machine or anything else. What that means is that device providers can now start providing standard devices that are mass produced. They can be used very quickly and easily by any solution.

 

Sam George:

That's one of the biggest step forward things that we've done as it relates to the connection between devices and the cloud. The other thing that we've done is we've lined that up with our new digital twins for environments initiative, and in fact, that schema that describes what a device interaction model is is the exact same scheme that we use to describe a physical environment. You can use that same schema to describe a factory and the people in it and the processes that are happening. What that means is that as a device connects to Azure digital twins, it just shows up as another digital twin.

 

David Starr:

Sam and Jeff, I want to thank both of you so much for being on the show. I've learned a lot today. I don't know about you Paul, but there are some things in this podcast that I certainly wasn't aware of, and I'm going to have to go digging and learn more about.

 

Paul Maher:

Oh, absolutely, learnt a ton. It's been, as always, a pleasure listening to Sam, and also, it's been great having Jeff onboard too, learnt a ton.

 

David Starr:

Well, we'll have links to all of the different technologies and some of the projects that were discussed in today's episode, along with social links so that you can follow Jeff and Sam at their respective social handles, along with Sam's blog. With that, I want to thank you, Sam, so very much for being on today, it was a pleasure to have you.

 

Sam George:

Thanks for having me. It was great talking to everybody, and it's never been a better time to start taking advantage of IoT.

 

David Starr:

And Jeff, thanks so much for representing the marketplace. Good job.

 

Jeff Cornwell:

My pleasure. Appreciate you having me.

 

David Starr:

Thank you for joining us for this episode of the Azure for Industry podcast, the show that explores how industry experts are transforming our world with Azure. For show topics recommendations or other feedback, reach out to us at industrypodcast@microsoft.com.