The next killer 5G app
(with Simon Buckingham)
Podcast episode 40
The next “killer app” will run on a 5G network.
Simon Buckingham of development house, Nonvoice Agency, says that developers need support — and possibly deep pockets — if they’re going to leverage the high speed, low latency, and power of the Fourth Industrial Revolution.
Below is a transcript of this conversation. Some parts have been edited for clarity.
Michael Hainsworth: 5G stands on the edge of a brave new world, one that will be built by app developers. Simon Buckingham represents more than 40 coders champing at the bit to build that world for us. As the CEO of Nonvoice Agency, it's his job to shepherd young programmers through the process of creating the next killer app. But he says that just like the evolution from 3G to 4G, we need to understand the evolving needs of the developer community under 5G. And before we can build that 5G future, we have to understand the past.
Simon Buckingham: In the beginning of the app store, a lot of the apps were utilities. You had a lot of note taking apps like Evernote, you had your first financial apps like Mint, you had a bunch of quite useful utility. Shazam could recognize your song. You had your first games that were coming out, like Angry Birds and things like that, so you had a lot of pretty good innovation in the beginning of the apps world in 3G.
And then as we moved to 4G, we really started to see video. Video streaming, video apps like Netflix. You started to see Zoom, and Teams, and those types of video conferencing apps were launched around the LTE era because you had a lot more bandwidth and capability. Dating apps like Tinder started to come to the floor and things like that. So, there was again another wave of innovation as we moved to 4G, and we first started to see more richer and deeper games experiences. I'm thinking about Roblox, and Pokemon Go. I'm thinking about those types of experiences that took a simple game mechanic and moved it to a whole other level.
What we're seeing with 5G is a whole other iteration again. We're starting to see ultra high definition video, ray-rendered videos, we're starting to see augmented reality effects becoming a lot more prevalent. We're starting to see multiple video chats concurrently during a game, or multiple video chats while a grandparent reads a child a book, and you can see the expression and you can read together.
And we're also starting, I think, to see a lot more innovation around augmented reality, so I think the killer application for 5G is augmented reality. I think the two technologies go together hand in glove, like mac and cheese, like bangers and mash, peanut butter and jelly - whatever your analogy is wherever you come from in the world. But we're finding that augmented reality really shows off the 5G capability because what you're doing is you're overlaying digital experiences on top of physical experiences.
MH: That's fascinating, you mention the 3G to 4G evolution, and the innovation from that leap was largely because of the bandwidth increase, the speed increase, the throughput increase we got gave us that video capability in a way we never had before.
Now, 4G to 5G, it's not just a massive speed increase, it's also ultra low latency, and it's also ultra low power devices not having to rely on an always-on connection. But it strikes me that you're suggesting the killer app in 5G has less to do with the super speed or the low power, it has everything to do with this ultra low latency where we're going from hundreds of milliseconds round trip for a bit, to one millisecond.
SB: Yes, I think you make a very good point. Speed always helps. These 10x transitions from 4G to 5G undoubtedly is going to enable a lot richer experiences. But I totally agree with you, Michael, I think that the low latency, what we call Mobile Edge Compute, ultimately you got to understand that the world we're living in - You're gonna have a pair of glasses which are gonna be the ultimate thin client. They're really gonna do nothing but sense your retinas and that kind of thing. All of the computation will move into the network. Everything will essentially come from the cloud as we evolve 5G and as augmented reality, and indeed virtual reality, become more prevalent. So I agree, those types of applications and innovation can only work in an ultra low latency environment.
The more usage you get, the more success you get, the higher your cloud costs could end up being.
MH: So then, what role does the evolution of cloud technology, from that vast, distant data center, to edge cloud, to near edge cloud, to far edge cloud... What does that ultimately mean for an app developer?
SB: The app developers that we work with at Nonvoice, we represent about 40 app developers in our agency, 40 of the best ones, and they love the fact that they can do more. They have more computational power at their fingertips, so they can do a lot more processing of experiences, and deliver richer, more interactive experiences. But, they also ask me all the time, "Well, what about the costs?"
At the moment, I'm using the device to do a lot of the work for me, and I'm a little bit worried if everything happens in the cloud, that actually has a lot of cost implications for an application developer because a lot of the apps are free [to download]. So actually, the more usage you get, the more success you get, the higher your cloud costs could end up being.
So one of the things I'm watching very carefully is how the mobile edge compute costs come down over time because that will actually have a big impact on the viability of a lot of the innovation that I'm talking about. So, that's one thing that we need to watch because if you're a young app developer with a very innovative app and a great idea, obviously you need to be able to fund that, scale that, offer that on a wide scale. Millimeter wave and ultra mid-band capacity, the fact that US telcos just spent $80 billion plus on mid-band 5G spectrum... The economics can't be underplayed either because obviously that money has to be paid back as well in some way, shape, or form. We're constantly balancing economics and innovation but as always, I think innovation will win.
MH: That's fascinating that as we move into 5G, the consumer is concerned that that 10x jump in throughput and bandwidth capability means that suddenly they're going to blow through their data caps. It sounds like the developer community has the exact same concern!
SB: Absolutely, yeah! Because on all of these experiences, we're moving to a world in which spatial computing, augmented reality, virtual reality, high speed video conferencing, avatars, all of these things, if you have a digital twin, what's going to happen to your data plan? Is it going to double because you send off your electronic helpers to run errands for you? We don't really know the cost implications yet because obviously the amount of bandwidth you're going to use will increase. Already, studies from around the world have shown that 5G device users are using significantly more, double digits more, data already just naturally.
Over time, as more of these experiences are released, I think you're gonna see a significant amount of incremental data usage, which is a good thing but obviously also has a cost implication for the telcos. You think about education, health, you think about working, you think about all the things now that we do from home, that are done remotely that we used to probably go to the office, or we might be going to the doctors office, or we might be attending school in person, all of these things have transformed themselves into remote and digital experiences. So, the wonderful thing is that 5G came at just the right time to enable a lot of those experiences.
MH: Do CSPs understand the revolving role in all of this from dumb pipe under 3G, and certainly lesser but still to a degree under 4G, into a full on app partner in 5G?
SB: The telcos, the communication service providers we talk to, they understand. They’re saying, “Okay, look, we're going to have to provide networks for licensing in order to give more capability. We're actually gonna have to provide cloud rendering at the edge of our network. We're going to have to evolve our infrastructure to support these applications.”
So I do believe the CSPs have some visibility of it, but inevitably they'll be surprised, I think, by how innovative the app developers are, and the requirements that the app developer needs. Because app developers need to know what capabilities are available at any given moment. As you're moving around, if you're moving from a 5G network ultra wide-band and you're moving into 5G mid-band, then maybe you're moving out of 5G in its entirety and you're on an LTE 4G network, all of those changes have implications for the app developer experience that they're trying to deliver. One of the questions our app developers constantly ask me is, "Can you provide a stasis flag for the telco capabilities and telco availability?" So, those kinds of traffic lights, that type of traffic signal, is it green for go, red for stop, or orange? These types of capabilities need to be worked out because they're becoming more and more important as we move forward and we encourage this innovation.
MH: So much of the innovation that is taking place takes place behind the scenes between those various levels of the core, all the way out to the radio access network. So that's the kind of software technology that telecommunication equipment providers have been working on in conjunction with the CSPs who have bought their gear. Who, though, should be responsible for building the API ecosystem so app developers can drag and drop the necessary components to build their 5G apps that leverage all of the new components that make up 5G?
SB: Yeah, I think right now to be honest, Michael, it's a little too early to know. Nonvoice, we have actually started to provide some APIs to our app developers because they've asked us for things like attribution. They wanna know where downloads are happening from, they wanna know where revenues are getting generated from, they wanna be able to track things a lot more carefully.
I think that there needs to be a lot more work done because it needs to be uniform. What you want to do is, you want to have a single access point to everything, and right now that doesn't exist. So I think we need a little bit more ecosystem collaboration from the GSMA and MEF, the Mobile Ecosystem Forum, some of the other trade and industry groups, 3GPP, etc. need to come together. Because at the moment, there are certain CSPs that are working with AWS, some are partnering with Microsoft Azure, some of them are obviously working with Google Cloud, others are working with proprietary mobile edge compute platforms like Mobile Edge X. And then simultaneously with that you have a whole bunch of innovation in the O-RAN, the Open Ran, and obviously Nokia, there's innovation from there.
So, basically what we need to do as an industry... I don't think people have really understood yet just how much innovation there's gonna be from 5G and how much more richness and capabilities are needed for app developers to deliver innovative new experiences. So I think there's still work to be done.
MH: If the killer app from 3G to 4G was video that took advantage of the throughput, and your suggestion that the killer app for 5G involves ultra low latency, things like augmented reality, what are the challenges here? Let's talk about that, and start by explaining the percentage of the overall app revenue pie that you think is going to come from, say, Industry 4.0-based apps versus the kind of apps that are more consumer driven.
SB: I think we'll probably be somewhere 50/50. I think there's a tremendous number of great enterprise apps. We work with some people who are doing a lot of blockchain-related innovation, in digital couponing, you see a lot of volumetric video applications, which can apply both to enterprise and consumer, where you have a lot of cameras and you're able to try on clothes, try on shoes, and do things like that in a retail environment. You see a lot of innovation around training and education because you can't easily jump on a plane and repair something anymore, so by being able to look at that and the overlayer schematic on top of it, to help somebody repair something is a tremendously valuable capability.
That being said, I think there's a tremendous amount of consumer innovation as well, a lot of it driven by the pandemic, because as I said, learning a lot of 3D and augmented reality books... My two year old son talks to my mom in England via Caribou, which is a video calling app with two video windows that open, and books. My mom reads to my son and they're doing it thousands of miles apart, things like that. Great little acts like that. On the consumer side they require quite a lot of bandwidth as well. So, I'm probably equally enthusiastic about enterprise and consumer. I don't know, it probably won't be 50/50, Michael, but in simplistic terms I think they're both going to be big buckets of opportunity.
The biggest issue with IoT has always been the chicken and egg scenario.
MH: So what are the challenges then, in developing apps that leverage URLLC, that ultra-reliable low-latency communication?
SB: Two things, one: availability, there's availability on a nationwide basis. So, obviously you have to think as a developer, "Okay, I'm gonna build something in a lab environment, or I'm gonna build something in an incubator type environment, but I wanna scale that out nationwide. At what point is there going to be enough millimeter wave or enough mobile edge compute instances where I've always got remote GPUs available to me?”
So, availability is one and, as I think I mentioned earlier, cost will be the other one. Who's paying for that capability, and how much is it gonna cost because there is a cost to all of that infrastructure. And at the end of the day, somebody's gonna pay for it. Whether it's Google and Apple who pay for it, or the CSPs, or the app developers, or a combination thereof, I think remains to be figured out.
MH: You mentioned that URLLC gives us the ability to leverage volumetric video, the idea of understanding video in, maybe not 360 degrees but- three dimensions. A lot of that kind of technology is already sitting in the palms of many of our hands thanks to the addition of Lidar based technologies in our smartphones and things like that. What role do the device manufacturers play in the innovation that comes out of 5G?
SB: The device manufacturers are pivotal, really. Like it or not, whenever Apple does anything, all of a sudden everything becomes a lot more real and a lot more serious. So, as soon as Apple started to put Lidar sensors in their high end iPads and iPhones, all of a sudden the augmented reality industry, which has been around to be honest, for years and years and years, suddenly reached a bit of a tipping point, and started to take off. All of a sudden when you have a Lidar scanner, the customization that you can give to the experience is extraordinary. Because you can put things on the surface of a table, you can put things on a wall, on a ceiling. You can really really create a much richer experience in AR when done in combination with Lidar. And I really think that the handset manufacturers need to continue to build those types of capabilities in.
We believe that over time, the Lidar sensor, which is quite basic at the moment, will get better and better and better, and therefore those experiences will get richer and richer and richer. And when you think about the glasses, which are just basically a screen with a couple of sensors, and all the computing is done via the ultra low latency cloud, it's gonna be a totally different environment to the way applications are processed, delivered, consumed now. And it will only work if all those enablers across the ecosystem are in place. That's how we deliver those truly innovative, brand new experiences for enterprises and consumers, I think.
MH: We're going to be able to take those Lidar sensors, those laser radar sensors, put them in a tiny little box, stick them in the corner of a room and be able to accurately indicate the number of people in a room, whatever that occupancy is. All sorts of other types of sensors as well that won't leverage ultra low latency, as much as they'll leverage the ultra low power requirements that goes with 5G as well. So, what do you see as app development in IoT looking like?
SB: 5G will enable, because of its bandwidth, its availability, I think IoT will continue to spread. I think we'll see more and more smart applications. Everything delivered into the home will be done through a wireless sensor. There won't be any meter reading now. All of these home appliances, smart home appliances, will be turning everything on and off automatically, whether it's thermostats, or heaters, or coolers, or whatever. So I just think that over time things will get smarter, and more intelligent, and more automated. And again, that experience will be something that consumers will really enjoy, I think. It will be a significant upgrade from what we currently experience, what we've experienced for the last 50 years. We’ve probably done things in the same way for a large amount of time. And I think 5G is actually going to enable that jump. If we get the standards right, if we get interoperability between devices right, if we get cost right, because you're right about the little Lidar sensors, but they need to be pennies. They need to be significantly less expensive so that they ultimately just get embedded into every room and every environment.
As you know, the biggest issue with IoT has always been the chicken and egg scenario. The hardware needs to get cheap enough for the applications to be worth doing. We're getting to that point now with Raspberry Pi. I see a large number of my application developers working with Raspberry Pi devices, whether it's to synchronize live music in remote environments, or whether it's to put it in a remote control vehicle and be able to patrol that vehicle by the phone. So we need more device proliferation and lower cost devices, and we need 5G to be built into devices like that. And then all of a sudden, we'll see that prevalence in conjunction with the telcos closing down their 3G networks, which then forces any legacy devices, like fire alarm systems, to be upgraded into the high speed networks. So we need to look at all of these things, Michael, for the ecosystem to evolve, I think.
MH: I am of an age where I do not know life before the invention of the microwave oven. A kitchen, in my world, has never existed without one. And that was sort of a world changing advance in culinary, positive or negative depending on your opinion. But, in a similar vein, social media made a huge wholesale change in the way the world operates. It strikes me that what you're saying is 5G is going to make those same sort of wholesale changes to society that we couldn't have predicted when Mark Zuckerberg opened up Facebook to the world in 2007, or the microwave was invented in the ‘60s.
SB: I think you're absolutely right. Having immersed myself in 5G over the last year or so, I've become more and more enthused by it, more and more intrigued. You have to understand, these are 20-something year old developers that are using a lot of tools like Unity and building things that, when I was a kid, on my Acorn computer and 64K, I was never able to do. But the kids today, in their 20s, they're totally immersed, and they're building experiences and using tools and using capabilities that we can't even imagine.
So as a country, what we really need to do is figure out how we put that infrastructure in place to allow that innovation to flourish? Because believe you me, the app developers will find a way of delivering it. They'll innovate, they'll create. I really do believe that, having spoken to a lot of the young people that we represent, that we work with that I speak to. But, we need to make sure that there is a little bit of coordination to make sure that that infrastructure is available universally, at scale, at a reasonable cost.
MH: So then what do app developers need from CSPs in this relationship?
SB: Well, to be honest app developers don't need anything. They just need a network, fairly ubiquitous, fairly reliable, fairly good. That's a bit of a flippant answer, but in essence it's the truth. But, to the extent that CSPs can provide remote cloud rendering capabilities and network slicing and those other things, particularly cloud rendering, I think they would be very valuable additions to the ecosystem.
MH: What then, do app developers need from the telecom equipment providers, like Nokia?
SB: They need to understand what that equipment can give them. They need millimeter wave and they need network slicing. All of those things need to be supported by that equipment. They need interoperability, they need privacy, they wanna make sure they're transmitting data over a secure network, you know, cyber security. They also need, to some degree, authentication. I see a huge need over time, if I send my avatar out into the world, my question is always, "How do you authenticate that that is in fact a digital representative of Simon Buckingham out there?" So, in my opinion, there's a huge amount of value that could be added by CSPs and network infrastructure providers in authentication of a user's identity, and protection of a user's privacy.
MH: Simon, this has been fascinating. Thank you for your time, if you are in fact Simon Buckingham.
SB: We will never know, we'll never know.