Navigating the Future with Nokia: Trending Tech Podcast Featuring Azfar Aslam. In this captivating conversation, we dive deep into the intersection of technology, security, and industrial solutions. Azfar, a notable figure in the tech industry, lends his expertise on the significance of the Internet of Things (IoT), the rise of the metaverse, and the importance of security in technology deployment.
Listen to all episodes
[00:00:00] Jim: Hi, and welcome to episode 38 of the Trending Tech Podcast, brought to you by the Team at iot-now.com, vanillaplus.com and the Evolving Enterprise, or TheEE.Ai. For this episode of the podcast, I’m joined by Azfar Aslan, Vice President & Chief Technology Officer, Europe at Nokia. But first, a warm welcome to all of you around the world who are listening in.
My name is Jim Morrish, co-founder of Transforma Insights, a firm of industry analysts focused on all things related to enterprise digital transformation. This conversation with Azfar is very timely since Nokia has recently published a couple of very interesting studies focusing on the metaverse and security respectively.
The first report in conjunction with EY found that 80% of experienced survey respondents expect the industrial metaverse to have a significant or transformative impact on their business. Whilst the second found that the number of IoT devices or bots engaged in denial of service attacks rose from around 200,000 a year ago, approximately 1 million today.
Maybe this is the beginnings of the long awaited rise of the machines. I’m sure that we’ll unpick both of these topics in more detail in our discussion. So, Azfar, welcome to the podcast. It’s great to have you as a guest.
[00:01:10] Azfar: Thank you very much, Jim, and pleasure to be here.
[00:01:12] Jim: Great. So, let’s start with the basics. How does Nokia define the Metaverse? It’s not all about legless avatars, is it?
[00:01:20] Azfar: It’s not just that. And, and by the way, for us, the journey in the metaverse started almost eight or 10 years ago. We didn’t know it was called metaverse at that time. We referred to it as human augmentation and the industrial augmentation capabilities that we were working on and researching.
And in doing that, we were looking at effectively how we could augment humans, not just with more electronics devices, but also what we call the biological devices, exoskeletons to make them more productive. So, productivity and safety was the mantra there. And then we started to do the same thing in the industrial environment and we started to build on that, creating sensors, data analysis.
And eventually what we ended up with was digital twins. So going through that journey at some point we called it Mirror Worlds where everything would be done virtually in terms of simulating the real environment and vice versa. And this now is referred to as Metaverse.
And by the way, for us, there is no one metaverse. There are many. And so we use a plural there, just call it metaverses and I think the last count as of last week was roughly 50 metaverses are available around the world.
[00:02:27] Jim: 50 Metaverses. I guess all we need is teleports from one to the other. But what you described there is, is something that really isn’t a consumer only thing and I’m sure there are real enterprise benefits to be had and in fact that first piece of recent research that I mentioned with EY found that companies that have already deployed industrial metaverse use cases are seeing more benefits than those still in the planning phase expect.
So, these seem to be projects that generally exceed expectations. Can you give some examples to illustrate some of the expected and I guess, unexpected benefits?
[00:03:02] Azfar: Right. So, first of all, I think we should, we have to differentiate between the segments within the metaverses these days. So, 50 out there. But they’re roughly in three segments. One is the consumer metaverse, enterprise metaverse and then the industrial metaverse. I think the, the last two is where we are predominantly focused on right now, because that’s where we see a real problem being solved around the world.
If you look at the diffusion of these new technologies over the last 40 years or so, there’s a trend that shows us if something will work out or not. If you go back into the 90’s we started with the computers in the enterprises, economies of scale happened and then computers found their way into homes.
So, the 2000’s was all about broadband, connecting these computers for entertainment. So, there was a use case there, enterprise, then a consumer led use case. In the 2010’s, whatever we were doing at home, we just did it in mobile. That was the next evolution. And the 2020s to 2030 is the next phase of that evolution of the digital evolution, both in the consumer world as well as the enterprise and the industrial world.
So, as we do that transition what we realised is that pre-pandemic, so during the last decade, the thesis was that Industry 4.0 was there to save the industry in terms of, driving more productivity, efficiency and safety in the physical industries. And we were focused on delivering capabilities, developing capabilities for those physical industries.
What we realised after the pandemic was that actually there’s a lot more work that needs to happen in the digital side, in the virtual side of things as well, so that we can collaborate better, we can view the systems better. And that led to the development of component-based digital twins to the system based digital twins.
So, to give you an example, we used to have a digital twin of a turbine and you could analyse how the machine was going to perform, behave, and when it required maintenance. But now we can expand that to a whole industrial plan. And when we are able to do that with the compute and cloud capability that we have, we are able to visualise things more holistically, so how the whole system will operate, when do we need to really fix something?
And those kind of capabilities have brought OPEX advantages to the companies because in the good old days, we used to have scheduled maintenance for the plants, manufacturing plants, industrial plants. You could have an outage, a planned outage, for roughly one or two days stops the production.
That’s a cost to the business. But now we are getting to a point where using these capabilities, all the things that we did in the last decade in Industry 4.0, now combine that with the virtualisation simulation capability of the entire plant. Starting with a digital twin, now moving into the Metaverse virtual environment, what we are now able to do is avoid those costs to the businesses, but the real outcome that we are seeing in terms of a productivity gain is that we don’t have to do a lot of the scheduled downtimes anymore.
We can deal with the problems as they’re predicted by the analysis and we solve those issues on a need basis, meaning we are more productive in those industries. And in the research that you referred to we looked at four particular industry segments. One of those was automotive where we looked at, okay, what are the predominant use cases that those companies are evaluating already or implementing already.
And then we looked at the industrial goods and manufacturing segment. We looked at the power and utilities, and then we looked at the transportation and the supply chain. By looking at all those four industry segments, one thing that came out really clearly of those companies out there, and by the way, we interviewed or worked with roughly 840 plus companies out there in the world across all the region in the world, what, what came out very strongly was that more than a half of these companies who are talking about implementing the metaverse use cases developing on the digital twin, half of them have already implemented at least one use case, or they have trialled out one use case in the form of proof of concept and the learnings then coming out, which is what you’re referring to, that the benefits are better than expected are, are exactly that.
When we look at the prediction of the benefit of a particular use case in automation, and we come back to those use cases later on in automotive industry, then we saw a much better or higher gain. We, that environment we saw same scenario, play out in the utilities, same in the other industries.
So as an aggregate, what we found was, and if you just narrow things down to the financials, the improvement in CAPEX that was expected to be delivered by these use cases was beaten already. So, we were looking at something like 15% improvement in the CAPEX deployment which will be improved by the implementation of these use cases.
Similarly, on the OPEX improvement, use cases that we saw across these four industries, we saw across the board a higher gain of roughly 6%. So, if you narrow everything down into financial, this is the net result that we are seeing, a significant improvement better than expected by each of these industries.
And that’s a good starting point for this industry because normally what happens is you predict something big and you fail to materialise it.
[00:08:23] Jim: It’s extremely interesting. So, you’re positioning metaverse really as an interface to a very sophisticated digital twin environment and system. And highlighting the difference between, you know, the business perspective and the consumer’s perspective.
So, in the business perspective, if the business case makes sense, it makes sense and you move forward and you do it. Whereas the consumer side is much more reliant on the network effects and comes in with more of a lag. But also you’re highlighting, that the kind of a reinforcement effect when people have learnt and have experienced from that first deployment, they get more out of the next deployment.
So, I guess the lesson is that if the business cases are there, then adopt quickly.
[00:09:00] Azfar: Absolutely, I’m going to pick up on what Apple stated in their launch of the new device. What they said was it, it’s a new computing platform. It’s a new computing paradigm, and that sits really well back to the situation back in the 90’s when we launched the first computing platforms there.
But it was the enterprises that adopted them first, and once there was enough economies of scale there, the cost came down enough for the consumers to adopt it. And I think we’re going to see a similar trend play out in the metaverse adoption, solve the bigger business and enterprise challenges, and then eventually the cost will come down enough for the consumers to be able to enjoy these capabilities as well.
[00:09:39] Jim: Yes. I think you’re right with that dynamic. A change in tack slightly. It seems to me that, you know, high quality networks, high quality connectivity, high performance networks, edge processing and so on, will be critical for many of these metaverse use cases. And that makes them, I guess, very suitable to private networks, for example?
[00:09:57] Azfar: Absolutely. I think there are really two underlying infrastructural components. One is the cloud technology itself, and the second is a network. And, and the two have to work hand in hand, and which is why in Nokia we, we talk about cloud optimised networking now. the reason that the two have to work hand in hand is that if you’re running a Metaverse platform, which is connecting multiple parties together, it could be different parts of the same company together, or in the consumer world, it could be multiple companies coming together to create a new platform.
In each of those instances, what you require is a very heavy compute capability with very low latency, and there’s a lot of data that is flowing between the end devices, between the machine and machine devices. And so on. And all of that data is going into the cloud over a network. So, what we are finding is that the performance of the cloud needs to increase quite significantly, and the capability delivered by the networks needs to increase quite significantly.
And in simplest terms, we are talking about very high throughput in a very short period of time, meaning the latency of service between the device to the cloud, device to a device has to come down quite significantly. So, the role that both the cloud and the networks will play in this era of the metaverse is quite significant.
It’s not just about the platform device anymore. It’s about delivering the right service, because if the service isn’t usable, if it causes challenges to your brain then adoption of those services will be very difficult in a wider scale. So, for the service to be fully utilised both in the, in the industrial environment, but also in the consumer environment, you need very strong network and a very strong cloud platform.
[00:11:39] Jim: Yes, that’s something I wanted to pick up on, in fact, because if we stick with the enterprise side for a moment in a private context, that’s a much more controllable context and you can control the throughput on the networks or ensure the grades of service and the requisite access to cloud and so on.
But I’m sure that there’ll be additional challenges when companies want to access those same use cases in the field, for instance, to support field engineers, et cetera. What, what are the key differences between deploying metaverse centric use cases in a closed controlled, private environment compared to a public wide area environment?
[00:12:14] Azfar: That’s a great question, Jim, because we are seeing that challenge play out already in the industrial manufacturing environment where the R&D teams, planning and design teams, those people who are responsible for evolving these manufacturing platforms from the previous generation to robotics and so on and so forth.
What they’re finding as a challenge is that there’s a lot of data that is dispersed across the world, in edge cloud locations that are within industrial plants, within an R&D organisation and so on. But when you try to bring that data, what you find is that either the throughput isn’t always available on the circuits that you’re working on, or the latency is too high.
So, when you look at that kind of environment and you say, okay, the role of the network will, will evolve quite significantly. We will have to find a way of connecting a very high performing campus network with another high performing campus network to the wide area network. And this is where capabilities like network slicing that we are introducing into our networks are going to help our customers achieve that higher level of throughput so that the cloud can stay in sync across multiple locations and these engineers or product designers or facility managers can work together.
So overall, when we go forward a few years into the future, the technologies or the capabilities like slicing, which are quite nascent today, will become quite pervasive. The performance that each campus requires from their service provider, again, in turn from their equipment and services that come from Nokia, each of those capabilities will be added over time. And the way we are adding those capabilities into the network infrastructure, it actually seeps into the cloud infrastructure as well, is under the banner of Future Ready Performance. And what we mean by that is that in some parts of the world, that challenge will happen earlier than other parts of the world.
And so when that happens, for example, in the United States and in the United Kingdom, which are actually leading in this research that we talked about earlier in terms of adoption of the metaverse use cases, we will be ready to provide those additional capabilities in time for our customers to connect their very high performing campuses to the wider area networks, which are generally owned and run by CSPs around the world.
[00:14:28] Jim: That’s interesting. So, there’s a key concept there around, you know, network slicing and effectively a dedicated grade of service slice across a wide area network to connect those local area network pools where it’s a more controllable environment. And as you say, it’s incoming technology, but to be rolled out soon I’m sure. Security… security must be a considerable challenge. Both in terms of application security but also the potential for IoT devices, for instance, to be co-opted into botnets. What are the critical considerations here?
[00:14:59] Azfar: Yeah, you referred to the security research that we published quite recently on the evolution of that threat from the botnets in particular where the IoT devices can be hijacked and used for purposes not intended, that they were intended for. I think the first thing that you’ve got to look at is there are two types of well, two segments of these solutions that we’re dealing with. One is the very high end premium industrial IoT devices that include the robotic connections, AGVs and so on, which come in with an integrated security framework that we can actually work, manipulate and customise based on the situation and the environment we’re in.
And then you have the lower end devices where a lot of these risks reside today. When we look at the overall threat vector towards the industrial environment we are finding that for the cases that we are looking at either in automotive and utilities and manufacturing, these are very well contained use cases and mostly running in a private environment.
So, the control over the security and access to the system is quite limited in those environments. When you start to connect these capabilities into wide area network to a public cloud environment, again, that’s a very secure connection that we are able to provide from the campus to the cloud environment.
But nonetheless, what happens is that from the lower end where the security levels with the frameworks that are implement are actually weak, that threat is continuing to proliferate as shown by the research. And that’s an area where, as an industry, we need to work with the equipment manufacturers, the IoT device manufacturers, the security frameworks that we are implementing, the standardisation of that.
So, all these actors need to collaborate together to solve that particular problem, which needs to be solved. But there’s a piece which is a little bit more secure and more available and highly reliable. And look, what happens in the real world is that you do not use inherently insecure systems in a very mission critical, business critical environment.
So, the cases that we are talking about are mainly in that category of. “hey, if something fails, my entire business process is going to fail down”, so I cannot take a security risk by bringing in weaker devices, weaker solution, weaker software. So, I think the way we are dealing with this situation is by being very selective, which cases we implement with what capabilities and what security frameworks are already available.
Of course, there’s another piece where we are looking at how these IoT devices could be hijacked, how they become the threat to the wider internet. The piece that we are doing in areas like deep field, inside Nokia, where we are tracking these devices, where we are tracking these botnets, we are tracking the behaviour, the kind of viruses, the threats that are coming.
We are sharing that back with the broader industry so that producers of those devices, the managers of those devices, are aware of what’s going on and start to take the corrective actions.
[00:18:02] Jim: Okay. Thank you. It seems that two distinct types of device, there’s a premium device with those built in security frameworks and those lower end devices, which are perhaps not quite so intrinsically secure. How do we scale these kind of industrial solutions?
Because potentially one day people might want to bring their own devices, a headset for instance, to access corporate applications. How do you start extending those frameworks to accommodate some of these less secure devices or devices that you haven’t prescribed to be part of the applications you’re deploying?
[00:18:33] Azfar: I think that bring your own devices to already a quite a popular item in all around the world these days. So that, that phenomena has been around for the last few years, I think. And there again, security frameworks at the IT level can deal with that particular challenge. But when we talk about the OT, which is the operational technologies in the industry and the enterprise environment, the entire business process runs on those operational technologies.
I think the way we are dealing with that right now is through these private networks, securing them in the right way. And there’s a reason why we are building these private networks with our customers all around the world. Because one of the first requirement that comes from these enterprises or the manufacturing customers or the automotive customers or the power and utility companies, is that it has to be super secure.
They cannot take any risks. So, security becomes one of the fundamental reasons for implementing a private wireless or a private fibre network for these customers. And that notion will continue, as I say, for the next few years in terms of the applications that we are talking about in the metaverse, where we’re simulating the whole plant, which, for which the data now needs to be accessed by let’s say an R&D engineer, which is sitting in another part of the world. So again, the secure cloud connection from centralised or the edge cloud needs to be provided to that engineer. And again, there are frameworks available to ensure that security requirement is met.
So, what I’m saying is that security was a fundamental reason why we went into the private wireless implementation on top of the productivity requirements. So, the two went hand in hand. The real challenge that will emerge over the next few years is when we start to connect multiple of these campus networks together, which may not necessarily be controlled or connected via a single service provider, meaning they might not have control over all the policies for the transport networks.
That will be an interesting arena. But this is where, again, the network slicing will come into play, which it can create across multiple service providers. So, a campus in the US could connect with a campus in the UK in a more secure manner because we have more confidence that the slices that these service providers are providing across the pond are actually secure, they meet the requirements, not just from the reliability throughput and latency point of view, but also from the security point of view. So, there are mechanisms which will keep evolving, but when we look at the biggest challenges that the industry is trying to solve, we have a way forward with those.
[00:20:54] Jim: Fantastic. It sounds like a very exciting future it’s been an extremely interesting discussion and thank you for joining us.
[00:21:01] Azfar: Thank you very much for me here.
[00:21:03] Jim: And, and with that, I think that we can draw this podcast to a close. And just a reminder that you can subscribe to the Trending Tech podcast wherever you found us today.
And indeed, thank you for joining us, and we’re delighted to have you listening in as part of our growing audience around the world. We’ll be back with another edition of the Trending Tech podcast soon, focusing on another aspect of digital transformation. In the meantime, please keep checking iot-now.com, vanillaplus.com, and theEE.ai, where you’ll find more tech news plus videos, top level interviews and event reviews, plus a whole lot more. You might even find something that I have written there.
Thanks again for joining, and bye for now.