So, hello everyone. My name's Neil Jackson. I'm from ARM. I'm from the IoT business unit there inside ARM. And it's my pleasure to talk to you today about ARM Embed. So I'm actually not from Norway, but I'm actually from the Nordics.

I'm working from the offices in Olwen, Finland. So I joined ARM via an acquisition about three years ago when ARM acquired a small company called Sensinode, which specialized in IP-based IT solutions. And that's how our office came to be.

We started off like 10 guys and now we're about 70 today. So it's a very fast-growing business unit inside ARM. The IoT business unit has been around for about three years now. It's based on a research team which has been around inside ARM for several years.

So we just decided to sort of spin that off into a separate group inside ARM to focus on developing IoT solutions and sort of take it outside the actual business itself to sort of give it freedom to do what it needs to do.

We're just out of interest. So anybody in the audience today who's heard about Embed? OK. So as I can see, it's not very popular in Norway. I don't know. We have about 170,000 Embed developers, but I'm not sure how many are in Norway. Maybe probably not very many based on this.

So moving on. Before we go into the details of the Embed story, I just wanted to sort of give some context about ARM. Some of you most probably know about ARM.

Just give you some insight on how ARM has come to be involved in the IT industry. So ARM's basic business is about design of processes, microprocesses, using devices from small sensors up to large servers. So the business model is built around licensing IP to companies that build the hardware, the chips, the sensors, and so on.

So as a result, ARM has quite a good knowledge of IoT in general. And as you can see here, this is a range of devices that are based on ARM's Cortex-M. So you see a lot of things that are out there today based on Cortex-M.

And this has sort of been happening because ARM has, the way it works, has a very large ecosystem of partners, hundreds of partners that work together with ARM to build this system. So in terms of volume, just over 17 billion of these Cortex-M devices shipped so far.

This is from 2015. Just 6.4 billion of those were last year alone. So you can see a lot of devices being shipped using Cortex-M. And so you can imagine that a lot of IoT devices, I think, out there that are based on ARM.

So that's a bit of background. So let's look about Embed next. So let's look at the problems that we're trying to solve with Embed. Embed itself was launched about 10 years ago as a web-based microcontroller development environment.

So something to give to developers that they can base, that they can develop around Cortex-M processes. In 2014, we set out to give that a new direction. It was folks previously at prototyping who wanted to take that one step forward to move towards making Embed a platform commercial deployment of IoT solutions.

Because we see that IoT is a very complex problem to solve today, to be successful in IoT, you need to be able to create a lot of, to solve a lot of problems around productivity, security, connectivity, management, and efficiency.

And these are things we've seen by working together with a lot of our partners over the years. So it's not just about creating the devices. You need to be able to connect them using different radios for different applications, connect them back to the cloud, and be able to do something based on that data. And it's our perception that despite being a huge prediction of IoT devices,

unless we can make that easy to do, it's not going to happen that quickly. So it has to be an easy way to actually get there. So these are the things we focus on. So the first one is management of IoT devices. So that's on the connectivity. So how do you easily and efficiently connect wide-range devices to services?

So what kind of range you need, all those kind of things. Management in general for devices. Once they're connected, how do you manage them at scale? How do you provision them? How do you update them? Things like that. Productivity. So once they're connected, how do you make sure you have the right,

you get the data efficiently from the device, that's the backend, and how you control those devices, so these kind of things. And the next area is around the development of IoT devices. How do you make that efficient? How do you have the right tools for the job? Is it secure? There's been a lot of discussion today about security.

It has to be a solution for an end-to-end security solution. And then productivity, so having the right tools for the job so you can do things quickly and efficiently. So that's what we aim to sort of solve with embed and try and build a solution that can be sort of reused in several IoT cases.

So you have the basic plumbing, so to speak, done, and then the developers, like yourselves, can then concentrate on your applications, what you're most interested in, not in sort of getting the basic connectivity going in the first place. So that's what we're trying to solve. So this is the Arm embed system.

It's a fully IP-based system, end-to-end. So what we have is a complete solution. So it's based on Cortex-M architecture. We have an OS for that, embed OS for the devices. And then we have what we call an embed device connector service,

which is something the devices connect to. And that offers REST APIs then for cloud service to connect into. And that's where we work together with our cloud partners so they can connect to the devices. And we also work with our embed silicon partners

to make sure that they're getting the hardware support for our embed OS system. And also we have an embed-enable program to give some kind of confidence that the devices that are embed-enabled have some kind of promise there. They actually work together.

So today I'd just like to give a brief overview of where we are today. So just to show how the Arm embed-based solutions are leading in key growth areas, I've got some examples around that. Look at the embed OS technologies that we have today.

Look at things around scalable security solutions that we're working on. And then to look at our embed developer and partnership growth and what we're doing there. Okay, so look at some solutions first. The first one here is an example in the construction area.

It's a company called Conferge from the UK. They have been working on an automatic concrete monitoring system based on wireless mesh network. So what this does is enables automatic wireless monitoring of concrete as it actually dries.

And apparently this is a very laborious, intensive process and I've spent a lot of time doing this. So having an automated system is very valuable. So it's an example of what you can do today. It's based on what we have already today as a smart city reference design. They've taken that and then used that to solve a problem that they have specifically.

Next one is in the area of healthcare. It's not a device, but it's something created from a part of our school, Zebra. And they've taken our platform to build some kind of a time tracking solution for the heart patients.

This solution has already gone through pilot review and they're moving on to a commercial phase in the near future. These are real things that are being developed today using embed. Two examples here from a recent conference, Mobile World Congress.

Two examples based on long range wide area network technology. In this case, LoRa. So that's a technology used today for wireless battery operated things. So it'll be in regional, national, global networks.

So it's one of the many networks that we use today to sort of deploy. In addition to some cellular, you can use LoRa or you can use something based on 6-Lopan. Something that's not regulated. So there's two demos here. One from Multitech and one from Stemtech.

The one at the top there is more of an interesting demo for a show. They have a dummy there with a hard hat on and somebody goes and hits the dummy on the head with a hard object and that sort of triggers a information that something happened to that guy and that information is then passed over the wireless network back to the back end

and then that triggers an alarm somewhere saying this guy has probably most likely just fallen over and hit his head somewhere. So that's always a nice eye-catcher at the events. And the second one was a smart city taxi application. Also using LoRa, this long range radio, to track the taxis.

So they're two very recent examples. Next one is the BBC Microbit. I don't know if you've heard about this but it's something that's been launched a while back now and it's actually shipping, started shipping last month. So it's an arm-based embedded system designed by the BBC.

So it's for use in computer education in the UK. So the idea behind this is to encourage children to get actually involved in writing software for computers and building new things. So it's been actually made available for every 12-year-old in the UK.

So the hardware is like a 4 by 5 centimetre device running an Arm Cortex-M0. It's got an accelerometer, magnetometer, sensors, Bluetooth, USB. It's got a display made up out of 25 LEDs and two sort of programmer buttons on there.

So it's either powered by USB or an external battery pack. So it's a very, very simple device. Something that's very simple, sort of cost-effective to produce and give to children. But then again it's something that they can actually go out there and sort of program and make it do things that they want it to do. So it's a very interesting project that's just starting off now.

So I'm interested to see how that all rolls out. So moving on from the examples now, just recapping on the IT opportunity, that's probably been talked about today in several talks, but it is a huge opportunity out there. So the volumes are huge but also the growth growth rates

are sort of astoundingly large. There's a lot of growth there, a lot of new businesses. And because of the size of that, there's going to be a lot of greenfield opportunities for new technologies, new platforms, new ecosystems and so on. So it's a huge chance for new companies to get involved.

And we've been working with partners now for quite a while. And during those talks we've sort of learnt a lot about what they're looking for. And so what you see here is three things that are sort of highlighted from those discussions.

Things that they want to see resolved. The first one of those is sort of fast time to market. This is something difficult today. How do you get out there very quickly with a device? The ecosystem's not mature. A lot of things, a lot of complex things you need to do. You need to source the components, the right technologies, the software,

security models, a lot of things you need to take care of. And this is something we're trying to address, to pull things together to make a solution available. The next area is around cloud services. So once you have the devices, then how do you actually connect them to the cloud? What are the REST APIs you need to have to allow that information to be accessed and controlled?

And that's something that we get asked about a lot. And then the last one is how to deploy this at scale. I mean it's quite simple to make a prototype, a small scale solution, but actually developing that into something that's very, very large scale, in the millions, and can be easily and effectively rolled out.

That's a huge challenge, and that's something that we see popping up in the discussions. So on these topics then, these are the things that we've recently announced to address these things. So the embed reference designs, embed OS, and the embed device connector.

And I'll talk about those now. Next. So let's look first at embed enabled and the reference designs. So if I kick off with embed enabled, so we want to enable scale.

So we launch the embed enable program, so we can give confidence to people that are using the embed, that their devices will be interoperable with each other. Some kind of confidence that they'll work together. So for example, if you're developing for embed your software on one device, then you'd like to know that that same software will also run on another device,

and that's the kind of thing we want to enable. Because at the end of the day, if you're going to invest in an IT solution, you want to have an ecosystem that sort of provides you choice. You don't want to be locked down. Sorry.

It's for all hardware that supports ARM's architecture here. So all the ones you mentioned here, they all support that.

So ARM Cortex-M is a hugely popular architecture and used by all the main silicon vendors. We're starting this program with developers, so we just want to make them confident that when they do develop for embed, that it's not just a one-off thing, but it can be reused for other solutions.

And then we'll expand that to cover our embed classic, where we have two versions, embed classic, embed OS, and our cloud services. Then moving on to the reference designs. The first one here is a smart city reference design.

So the concrete company example was the first one, sort of using this as a basis for their design. So this is for the smart city segment, and one of the things that pops up in smart city solutions is how do you build a sort of good network.

You might not want it to be sort of sailor-based because that may be expensive, difficult to employ. So there are alternatives. So, for example, using these long-range radios and also using the low-range wireless 802.15.4. Sub-Ghz is a good option for that.

It gives you long radio sort of reach with good power levels. So this is an example of how you can build such a system, and it gives you the hold. It's based on a six-low pad network, so that's a mesh network. So, for example, if it's based on a street lighting system,

then each of the lights then connect to each other, and you can have hops in that system, like up to 15 or more hops in a system. So each node connects to the next, which connects to the next, and it performs like a series of devices which sort of build their own network. So this is something that we've built as an example that anyone can go away and take this

and sort of base their solution on this. And it's based on IP, IPv6, right down to the end node, and today getting an IPv6 connection can be a bit of a problem, so we've also put like a tunneling solution in there,

an example of how you could do that, because getting a good IPv6 connection might not be an easy thing to do in rural areas. So just an example of what kind of things we're building now to sort of accelerate the take-up of the technology. Another one we have here, just recently launched,

is an Embed Wearable reference design. This is something that was built at Arm. It's been a project that's going on for several years, and sort of goes back to the research team that's been in Arm now for a long time. They wanted to build a watch just to see what it actually means to build a device like this,

what problems you need to solve. And so this project sort of came to an end just recently and was launched. It's available today on embed.com. If you go there, you can see all the resources. So you can get all the hardware design. You can get the electronics design.

You can get the software for that. And this is a showcase for EmbedOS, so our operating system for the Internet of Things. And actually, I don't have a device myself, but this guy here from Silicon Labs is a very proud wearer of one of these devices,

so it's actually real. So if you want to go and look at this later, just go and grab him and see what it is. But it's an interesting concept. So anyone can go away, take this blueprint now, and then sort of design their own device based on this. So it's just one where we want to sort of accelerate

the use of our technology and our solutions. And sort of an interesting sort of side effect of this project was the battery life is extremely long, so we've seen up to 11 weeks of battery life with these devices, so as you can see, we've been designing this for low energy,

and that was sort of something that came out. It sort of lasts for several weeks, so that was good to see using this device. So next we'll look at some of the Embed products. But to start with sort of a recent statement now.

There's been a lot of discussion about security today, and this is something that's taken from the GSM-A IoT security guidelines. So the fact that even the easiest to develop type of endpoint device must behave in a reliable, high quality, secure manner, because it's expected to participate in a network that could eventually span up to millions of devices in size.

So the basic message there is that security can't be ignored for any of the devices in the IoT space, because at some point, they're most likely gonna connect to the network, and that means that they're gonna be then vulnerable to security attacks. So we believe this is a problem

that has to be solved to make IoT reality. So I'll jump next to the, look at EmbedOS in a bit more detail. So this is our operating system for IoT devices. It's sort of a modular, efficient, secure,

open source OS for IoT. So look at the coloring scheme here. The bits that are sort of bluish, they're the open source parts, and the ones that are sort of purple boxes there, they are actually binaries. We don't make all the software available

as open source to the community. Things that are in sort of development phase still, things like Thread, some BLE drivers and so on. But if you're an Embed partner, you get access to all of this, so even the purple bits on the chart here.

But I'll talk about the partnership later on. Okay, so just a quick look at this now. So what it is essentially is a complete operating system for IoT devices. Starting at the bottom, you have the interface of the hardware, which is designed specifically for Cortex-M. So that's the 16 point,

well, 17.1 billion devices that have been shipped so far. Based on that then is what we call the Microvisor. And someone was talking about this morning in one of the talks that one of the things you need to do is to compartmentize your software. So the critical parts need to be put in a compartment, and then all the things that you can't trust then

are in another compartment. So we built a solution that supports this. And also the newer versions of Cortex-M, the version 8 that's coming out in the near future, has been announced recently that actually supports this in the hardware. So starting right from the very basic elements,

we are implementing good sort of security policy in the OS. So on top of the Microvisors, we have the Drivers there. Then we have the IP Stack on top of that.

And then we have APIs for each of the network interfaces. And then we have what we call Embed-TLS on top of that. So it's a secure library. That's actually based on a stack, an open source stack that ARM acquired just under a year ago now from a company called Ofspark.

So that's a sort of very, very proven and widely used TLS stack that can be used in embedded devices and also in servers. So that's the sort of critical part of our system. And then on top of that, then we have what we call an Embed-Client, which manages the connection then to the backend.

So it doesn't matter what kind of device you have, the Embed-Client will make sure that you're being connected to the backend in an efficient way using efficient protocols and in a secure manner. And then on top we have the, of course, we have the high-level APIs, and then the application will run on top of that

with some libraries. On the left then we have the schedulers there. And that's about it really. But the idea is to have a complete operating system. So you wouldn't need to go out and source different pieces from different places, but it's all there to choose from.

So you can choose the bits you want to use and leave out the ones you don't. So there's a good set of components to choose from. Okay, but it's more than just an OS. It's a full software solution with a development toolkit.

So it's targeted for production-grade devices, which is different from, we started off like 10 years ago what we call now Embed-Classic, which you can see on the left-hand side here. That was designed initially to make microcontroller development easily

for rapid prototyping. And that achieved its goal. We now have around 170,000 developers working on Embed. It's based on sort of free online tools. And what we're doing now is taking that basis and we've taken that and built upon that

to create what we call Embed OS. So you can see we have two different tracks for our operating system here, and Embed OS is the most recent one. So we took the good things from Embed, our Embed SDK or Embed Classic, and we added security, more connectivity, made it some maturity for production use

and trying to improve the overall developer experience and also to bring in our connectivity to the cloud. So moving on to our Embed device connector now. This provides Embed developers with a simple, secure,

and energy-efficient way to connect any Embed-enabled device to the cloud. It provides connectivity that's secure and built for constrained IT devices and provides a good set of REST APIs for application developers. So as you can see here,

this sort of gives a diagram of how you would use that, but what you need to do is you go to our Embed.com site. You would log in, get your Embed account there, choose your hardware, design your device, and when you do that, it will be automatically configured

to work with our device connector, which is our cloud service. So when you've done that, you power your device, it should automatically then make a connection back to our network, and it should register as a device on our cloud service.

So you can go in there, and you can see all your devices popping up in the portal there. And once it's in there, there's a REST API available that you can then build your application on top of. So if you're building a street lighting service,

then you have your street lighting application running on top of that, accessing the data from the devices using that REST API, and also controlling those devices with the same REST API. So it's a very simple way to get up and running with your complete end-to-end solution. Do the device, connect to the cloud, build your application, and then you can very quickly

start playing around with the application, which I think for most IoT companies, that is the actual thing you want to be concentrating on, how to make your own application and business. So we've introduced this, like, as a free service for our embed developers.

You can connect currently up to 100 devices, have 10,000 events per hour. So that's enough devices sort of to get up and running, connect a few devices, play about with it, see how it works. It's a strong security story, so it's based on our embed OS and our embed TLS and so on.

So it has end-to-end security, supports TLS and DTS, for example. They are open standards based on Co-App, which is an efficient protocol for transferring data between the device and the backend, and then it has the REST APIs that you can use to integrate with other systems.

And if you need to go beyond that 100 devices, then you just need to contact us, and we'll see how we can take you further. Most likely, it'll be one of our cloud partners that can then take you beyond that and towards a commercial service.

So next, pushing along now, just to look at some of the technologies. I'm going to look at Red, 6.0 Pan. I don't have much on BLE, but I've got a couple of slides now on Thread and 6.0 Pan. So for those of you here this morning

listening to the keynote, Nest was mentioned there. Nest is seen as being a very interesting and good example of something in IoT, and Nest are actually working. They are part of the Thread group, which have created the specifications for Thread and are developing that.

And Arm is a founding member of that group, and we've been developing our own Thread support for EmbedOS. So it's a secure wireless mesh network topology phone. So there have been a lot of technologies out there. This is sort of a new way of looking at this.

We've taken existing standards, and just put them together into a package that can be a good solution for solving the network solution for connected homes. It doesn't solve the application layer. It's sort of application layer agnostic, but it does provide a very good networking solution for home networks,

including sort of good support for commissioning, which is an integral part of that. So it's built on proven existing Internet technologies, IP, UDP, DTLS. Six Lopan pulls these together into a solution that sort of works nicely end-to-end.

So we're just seeing devices coming out using this now, and we'll have support for this in EmbedOS. Six Lopan mentioned this briefly in the smart city reference design. So it can be utilized in large commercial networks where you can connect, for example,

several hundred nodes or endpoints to one sort of gateway. So you can have, if it's a street lighting example, you could have hundreds of street lamps connected to mesh network, and they'd report back to a one gateway, which, for example, is connected to a cellular connection. So you can cover quite a huge area geographically with this technology.

And these radios, the 802.15.4, a lot of those have been shipped, so it's a good sort of proven technology. So moving on to security in a little more detail now.

So the challenge is, I think, to make all devices secure, relevant if they are what size they are. So we know there's going to be a sort of huge diversity of endpoints, like from sneakers up to servers. It's a common problem that needs to be solved.

So, and it's not just the endpoint itself. We have to solve the solution from end to end. So there are lots of places in that chain, starting right from manufacture up to the deployment of the service, use of the service, and sort of taking the device out of service.

A lot of places in that chain that have potential security problems. This is a very simplified view. We just want to sort of emphasize that it's quite a complicated solution to solve. And just to sort of emphasize this,

so we know this needs to be a low-cost solution. So for example, taking the sneakers, they're going to be very, very low-cost devices going into those wearables. And that means that security needs to be easy. If it's a very low-cost device,

they don't have the luxury to focus a lot on security. So having tools and software solutions that are available and just work out of the box is a sort of key to making this all happen. Otherwise, we'll have the problem where devices that are sort of lower-cost, they can't afford to do security,

then we'll see a lot of unpleasant things happening there. But it needs to happen through all the points of the chain. And just to sort of see how we're approaching this in Embedded Wear is a quick example now. So to have a secure device, we have solutions here.

It starts with hardware, so you need to be able to store information securely on the devices, be able to use a random number generator built into devices. A lot of the security stuff we have now, we just made a recent acquisition of a company called Sansa. They brought a lot of this technology with them into the company,

so we have a lot of good solutions on there, how to make a device secure. Moving up the stack then, we have the drivers, they need to be secure, so we have the microvisor for that. Also, the lifecycle security information, so all the things that need to be secure during the lifecycle of that device,

all the keys that need to be secure when it's taken to use, when the user's changing, all that information needs to be very secure. Moving up the stack now, then on the top, we have the communication layer, so need to make sure that the connectivity to the back end is secured, and that's where our EmbedTLS solution comes in.

So we're working to make a complete solution. It's not just about the device, but the whole end-to-end story of securing that device in the system. So, running out of time now. So, just wrap up now, talk about our partnership. We've seen good momentum since we announced the IoT device platform in 2014.

We have a good take-up of Embed developers. We've seen the number of developers double during last year, so that's very nice to see. The numbers aren't huge, but if you think about the size of the embedded community today, that's quite a nice portion that we have already,

so we're quite proud of that. And also, there's a good sign of momentum. And also, nice to see the developer sessions also going up, so people are actually using it and developing stuff with our solutions. And supporting developers is very important for us, so something we take very seriously there.

And also, in our partner network, it's good to see that growing. We have over 250 engineers from our partners contributing already to Embed. We have Silicon partners, ecosystem partners, cloud partners, all working together with us.

We have over 55 partners to date, so that's growing nicely. And what's interesting is that these partners have been contributing 20 core technologies towards the ecosystem, which is nice to see. It's not just things that we are doing, but our partners are bringing in solutions and software also to broaden the ecosystem, which is great.

So, just to wrap up now, this is the latest picture of our ecosystem, so this is our partner network. IoT is a very complex system, and we know that we can't do this on our own,

so that's why we have this ecosystem approach, that we can build a community that has a common goal of making a solution that works for a wider audience. And we believe this sort of ecosystem approach is the way to go, and we work with a variety of partners.

The bottom here, you can see our Silicon partners, so they are porting their devices to support EmbedOS, so when their devices are shipped, and if somebody wants to use EmbedOS, then they know they can find the drivers on the EmbedOS site, and actually go ahead and develop on top of those.

On the top, we have our cloud partners, so these are the partners that are plugging into the REST APIs that are provided from our device connector, and they provide all kind of analytics, all kind of applications for different kind of services, so there's some big names there. They see the value of us bringing in the devices by Embed

to their sort of their own systems and solutions, and they can use that data generated from the device, and use that for new services and solutions. On the right-hand side, we have what we call our ecosystem partners. These are people that are working together to build solutions.

They are bringing tools along, or they are developing end-to-end services or different kind of things, so they're contributing also to the ecosystem as a whole, supporting Embed and helping it to move forward.

So, that's about all I had today, so we're looking forward to sort of moving Embed forward, and to summarize, so if you haven't already, please go and take a look at Embed, and if you're interested, then why not talk about joining a partnership,

what that means, joining our ecosystem. What we see this year, we hope to see continued success of our solutions. I'd like to see more OEMs joining and sort of taking Embed into use and making more solutions. And as a final note, I'd just like you to go away with the information

that Embed today already offers, the key technologies to enable secure, robust, and scalable IoT for services. I won't be around this evening, but a colleague of mine, John Riley, is also here, so if you want to hear more about Embed, then please grab him and have a chat.

But that's all from me today, so thank you very much.