Guy Daniels, TelecomTV (00:27):
Straight into our next session then, which is our first market perspectives debate, and we're going to start with the automotive sector and a session on supporting the market growth of connected vehicles, some context here before we start V2X technologies currently driving this market for connected and autonomous vehicles. We'll hear more about that I'm sure. And the next step will likely be the introduction of the so-called software defined vehicle. Great opportunity there as well. In theory, massive value to be had for all stakeholders, but there's also a degree of uncertainty around the roles of the various players here. So we're going to try and identify the differences and the commonalities between auto and telco and we'll see where we all go from here. But first we're going to meet our panelists and I'm going to ask 'em to introduce themselves briefly starting on my left with Maxime.

Maxime Flament, 5GAA (01:25):
Hello from my side. Thank you very much to telecom TV for the invitation. Really appreciate. Now I'm representing 5G 5G Automotive Association. It's an association putting together the telcos on one side in the automotive industry and it's all about understanding one and the other. So I think the panel is pretty right to the point and it's also understanding opportunities needs but also pain points and problems and put them on the table and discuss about them and solve them together

Guy Daniels, TelecomTV (01:59):
Exactly what we need. Thanks Maxime, Warren.

Warren Bayek, Wind River (02:01):
Hi guy. Thanks. Happy to be back here at the DSP Leaders World Forum. Again, I'm Warren Baek, VP in the technology office at Wind River, focusing on Intelligent Edge. Thanks Ron. Colin,

Prof. Colin Lee, JLR (02:15):
Thank you for having me. Colin, professor Colin Lee of Jaguar Land Rover. I'm also an honorary professor of practice for Queen University Belfast. So that's Ron.

Guy Daniels, TelecomTV (02:26):
Great, thanks Scott. Thanks everyone. Colin is our co-host for this session. He's going to help us frame our discussion with a special DSP leaders address. So Colin, I'm going to ask you if you'd like to take the lectern or the side of the lectern or behind whichever's comfortable for you.

Prof. Colin Lee, JLR (02:46):
Thank you very much. Well welcome everybody. I'm just going to get a piece of paper. It's not really much saying on it. It's about five pages long. I'm kidding. I wanted to talk about automotive versus the telecom industry. I know this is going to sound weird at the beginning, but it will pan out. So what are the common paths that has been done by the telco industry that the OEMs are now going through or the automotive industries now going through and just at different times? So let's go back a few years. When we first started looking at phones, even going back, we literally congest at talking to people and that was it. So if you look at, obviously you guys know all this, I'm just literally using this as an example. If you took a phone in the old days it used to be just a phone.

(03:38):
And if you look at the business plan at that point you are probably looking at it trying to figure out, okay, how can I make more money and how can I make more signals go over the air at that time? Then all of a sudden they added a camera, which I'm pretty sure wasn't on the original business plan. And then all of a sudden we went into a world of digital and apps and things like that, which you could nowhere near. You would never have known back at the beginning that that world was ever going to happen. And yet there you are. And if you look at your business now compared to perhaps you were back then, I'm pretty sure that you would be going, yeah, I'm looking back now. I wish we knew what we knew now because we could have done this earlier or something like that.

(04:23):
Well, the automotive industry is at that beginning point. Even now it's looking at the, we are used to making very fast cars and making you feel great as you're going around country roads and taking the family out and enjoying themselves. But it's gone beyond cars now. It is now, well we can actually do software over the air and do fixes to your vehicle. We can diagnose issues, we can do all these nice things that we could never have done without being connected. Then in 2019 they decided to mandate, I think it's 2019, I could be wrong, am I right? I never remember. But basically we were mandated to put for equal, which is basically means that if the car is an emergency situation, you're able to get some emergency support. But we recognize and we are still looking now looking forward like you were back in the phone, we just knew cars, right?

(05:26):
What could we possibly do that will be more than just cars, more than just phones? Well, once we start thinking about that, SDV comes to mind because we could really just take that external network and create our own network in a car. Oh, that's quite interesting. That can we use APIs? We can really do things that make sense for the customer on top of just the car, but hang on, whoa. Now we're connected to the outside world, which means that we're going to be in trouble if we get someone decides to hack the car in some form or fashion. So if someone hacked your phone, okay, the phone doesn't work or perhaps they've got some data on you which you can change the password and perhaps deal with it later. But if that happens in a car, it could be potentially much worse.

(06:18):
So the adas side of cars tend to be very skeptical of connectivity in terms of if you've got your adas here and you've got your connectivity, there is no connection at all between those two. So you can say slow down, slow down, slow down. But these guys will sit there and go, I'm not sure about that. I would put it an example to you. All these years you've been alive, you already know that you shouldn't put your head in the oven, right? So if I asked you to put your head in the oven, you won't do it. And there's a reason for that. So the car has a very similar mindset. It says well slow down. Well why slow down? I've got one person telling me to slow down. Oh you need to slow down now because there's 200 people telling you, okay, I might consider it, but eventually the ADAS system will make a decision.

(07:06):
The point I'm trying to make from a car perspective, the value of connectivity is to improve the capability of the vehicle. So if we can add value to the ADAS system, you'll have a much better experience of the car. So up until now though, we've been focusing on in adding value things like adding Apple CarPlay to give you a better feeling from A to B. The second stage would be to enhance on that and then the final stage would be to do the ADAS at the end. But up until that very end point there is a problem with value for the OEMs because, so it's a leap of faith. That's what I'm trying to say. And I think when we get to the panel, I think this will start building up, but what is clear, we all go in there and we all want to do it. We're just trying to figure out the road ahead. So if you go back to your phone, you were looking forward struggling, now you're looking back going, wow, we are there looking forward struggling, but we may need your help to look back. Got that capability, we can't. We might need to work much closer than we've ever done before. That's me done, I think. Thank you very

Guy Daniels, TelecomTV (08:24):
Much indeed. Applause for Colin. Thanks Colin. Fantastic. That's got us all thinking. Hopefully we're going to pick up on this one. We should really start by looking at what are the commonalities, what are the differences between telcos and auto? What exactly is the market opportunity? Because a room full of telcos might see the market opportunity a little bit differently to a room full of, well I have to use OEMs, don't I? Which let's Maxim, do you want to explain to our audience what OEMs are? I'm not being patronizing, but in our business an OEM is probably something different to your business.

Maxime Flament, 5GAA (09:07):
Well we would say automotive OEMs and so that's original equipment manufacturer basically putting pieces together and the supply chain comes along with it. So you have tier ones, tiers two suppliers, and at the end of the day the vehicle manufacturers with all due respect is just putting this together so that it works and it's nice product.

Guy Daniels, TelecomTV (09:28):
So we telcos will look at the connectivity potential here. Perhaps I'm being a bit simplistic here, but selling the connectivity, selling the bandwidth, selling that capability. We're hearing from Colin and we heard from you last year as well at 5G AA event it, it's not quite that simple, is it? Can we try and look at what the commonalities are, what the differences are? You said we hope we can do this. You say we really want to do this. What's the big stumbling block, Colin?

Prof. Colin Lee, JLR (10:05):
Well, I mean we're very similar to phones and sense. We've got batteries, we've got communications, but we need to always be connected if we can have an ADAS environment. So you were talking about similarities? Yeah, so the similarities are that we talked, you spoke earlier about trying to save money. We've got to look at also saving money. We've got so much per megabyte per car and you are trying to expand your network. We're trying to do it the same. We're trying to go worldwide, we're trying to work the same in one country as we are with another. We want to give the customer the same feeling wherever they are in the world, whether you have the experiences on the phone or whether you are driving down the road, you don't want a situation on the phone where you're talking and all of a sudden they're cut out and come back in again.

(10:59):
I don't want a situation where the car starts spluttering or slowing down and starting up again because it got the wrong information. Very similar experiences because what will happen is they'll blame the person that might be the phone or the actual m and o that's running the phone, but if it's a car experience they'll be blaming the OE em or the people like us. So that's a very similar thing. At the end of the day, we're giving a service to the customer with which they are very much in a digital domain and it's getting closer and closer and closer as every day goes past

Guy Daniels, TelecomTV (11:32):
Maxim, the five GA brings together OEMs, it brings together the telco community. Is there an equal participation here? Do you see an equal desire from both sides? Have you still got issues that you need to sort out?

Maxime Flament, 5GAA (11:51):
So perhaps come back on the topic of similarities. I think that there are a lot of similarities, but I can depict perhaps the viewpoint of the vehicle manufacturers or the automotive where we are. So vehicle manufacturers are relatively slow, slow to adopt technology and even less so how can I say that? To adopt connectivity, it took about eight years to adopt 4G from standardization to full adoption of the automotive in 2016. Around that these eight years, if we look at the 5G adoption, we're still not there. So we still have a little bit of time. Now what the vehicle manufacturers like to do is to wait until they have to do or have to actually implement. And very often they are responding to regulatory soft or hard regulatory frameworks to make decisions. And currently where the regulatory frameworks are pushing them is in terms of efficiency, in terms of electrification, in terms of automation, not too much in terms of connectivity. Connectivity is just a means to the end to answer the connectivity. Some of the regulatory aspect that I just talked about. Now where we are struggling is in terms of innovation, in terms of adopting 5G with all the promises when these promises have not been delivered yet. And so that's I think where in the next couple of years in terms of coverage, in terms of API availability in terms of other edge computing availability, we are going to see progress on the telco side where then the vehicle manufacturer can embrace these opportunities.

Guy Daniels, TelecomTV (13:57):
Absolutely fascinating. And Warren, your company applies to help us as a market and it applies to the automotive OEMs. As a market, what are you seeing in terms of the different approaches in the two?

Warren Bayek, Wind River (14:08):
Well, and as further, as you probably know, our parent company, Optiv is actually an OEM supplier. So we really have a unique view on this. I actually want to talk really quickly though about some similarities. One of the similarities, and Maxine touched on this and Collin stole most of my talking points, heavily regulated industries and regulation does tend to drive a lot of what happens in the automotive industry as you just alluded to. So the move to connectivity is not about necessarily something they want to do now, it's something they're kind of having to do to reach some of the targets that the regulatory bodies are bringing to them. Like telcos, the automotive industry and automotive Williams in particular are trying, there's two pieces and we've talked about this all through the morning, the revenue side and the cost side. Driving costs down is still a very important part of this industry.

(15:02):
And the move to, we'll get to software-defined vehicles in a bit, but that's a big piece of the connectivity puzzle. The move to the software-defined vehicle is largely going to drive costs down in a myriad of ways. Much like cloud should drive down costs in telco, right? You're taking a lot of what used to be tens, many tens, 50, 60, 70 processing units in a single car and you're bringing it down to four or five. So you've eliminated a ton of manufacturing, you've eliminated a ton of cabling, you've eliminated a ton of things that can go wrong during the manufacturing process, lowering the costs. So that's kind of one of the similarities is they're looking at interesting ways to drive down costs. Going back to the question of how are OEMs looking at this, I think the question, and this is probably I'll lead you into your next question, it's the monetization question is the big question before us. And that's something I hear in all my conversations with automotive customers, with telco customers looking at this industry. And that's a part where we need to play together in the collaboration of these markets and the market players is going to drive vehicle connectivity. And as con alluded to, the vehicle isn't a vehicle anymore, it's a rolling data center. So I'll leave you with that and we'll get more into that in the next question.

Guy Daniels, TelecomTV (16:21):
We'll take you up on that. We will go straight into monetization. That's a great way for another talking point we ought to discuss. And if Maxim, I can come back to you first because I've been following what you've been doing with V two X off the back of the three GPP standards of V two X. But as you were saying just a few minutes ago, there's a problem in that the industry is maybe not believing the 5G claims or wants to see more evidence or more progress. How does this lead to monetization? What do we need to do?

Maxime Flament, 5GAA (16:57):
No. Okay. Connectivity haven't seen as a cost center for the vehicle manufacturer. So it's not a given that is going to bring money. And an example of this is just safety. So safety is not going to pay the bill, it's just going to be an answer to a brand, an answer to a branding, to something that you're defending. But now where connectivity really is breaking is when you are talking about over the air updates. And again coming back to the regulatory aspect, it's not necessarily to bring new features all the time, like the Teslas of this world would do is to the heavy cost of recalls. That's one of the things. And so Tesla has perhaps done that first, but then where the mindset needs to change is when you start using this connectivity, you have started to have reduced your cost with recourse, it motivates the connectivity and after that you really have a data strategy, have a product strategy based on your connectivity and always learn from your vehicles when they are in the field, understand better the drivers, et cetera. There is loads of new opportunities to bring when you have this connectivity, but it's first and foremost motivated by this first thing, the recourse and the cost saving that you can incur to that.

Guy Daniels, TelecomTV (18:42):
Sure understand. Thanks Colin.

Prof. Colin Lee, JLR (18:45):
Well connectivity is amazingly the way I look at connectivity over the future. You've got the customer convenience side linking to them and their world, whether it be at home or whether it be at work or whether it's making it much easier, see them panic and stuff like that. But for the vehicle industry world, it allows us to really look at the car in a totally different light than we've ever done. We can start looking at things that are not quite working and we can monitor that over time. We can check your vehicle to see if it's just you. So if say you went around a corner too fast, you always do that for some reason and I don't the car would know that and when you get in the car we could adjust the car for your driving so that we have less of a problem on the steering the tire where for example they're just ideas of what we can do inside the car.

(19:38):
So there's a lot of things that we really want to be connected for. Other things like the V two X where you can't see beyond the horizon, so you can't see in front of the lorry in a way, but being able to know what's going, like a heat map around you of what's around you allows you to make a decision. Like you, when you go in somewhere, you've got your ears. Ears are like the V two X or the connectivity. You can hear what's going on and based on the danger you will act accordingly. So if someone shouts, you immediately turn your head. So if there's a heat map that's really bad, we quickly look over there and decide does it affect us? If it's not anything serious, you just get on your daily life, you carry on with the journey, you don't even tell the customer, but if it's serious you will automatically walk away from it.

(20:23):
I guarantee whatever serious it is. But the car itself may go, okay, what do I do here? Is it something dangerous for the car? Is it something dangerous for the customer? Is it going to impose on their comfort by getting to the airport on time and so on. So we cover a multitude of areas where we have to rethink well what's it for? And what I like about it, if you look at the car, we can improve things and make the car much better, which is why we always try to be connected. But as soon as you open up the connectivity to everything that the customer sees, they have a completely different world, totally different feeling of what's going on in the vehicle. And I think that is really where all OEMs are heading is to make your life so much easier so you don't worry.

Warren Bayek, Wind River (21:09):
Yeah, I'm glad you brought that up. We actually demoed at MWC two years ago and again this year a more advanced solution, that exact problem statement that you showed, which is it's sort of crowdsourcing if you will. And again, you made a great point. The car will never take action based on one thing, it gets right, it's going to build a picture and frankly the really hardcore safety things that happen in a car are never going to be crowdsourced, right? The braking system, some of the things that are really critical, but building a more complete picture of what's happening all around you is a very important part of connectivity. How do we monetize that? That's an interesting question and one of the answers we're getting from OEMs and that we're looking at across the industry is the largest industry in the world is the insurance industry of which the automotive part is a pretty big slice. Insurance companies are looking at any way to save money on claims. One obvious way to save money on claims is to create less accidents and less costly accidents. If they can get better drivers and better car driving through connectivity, they're willing to pay for that because it saves them so much on the backend and through partnerships, everybody who plays a part in that connectivity story from the car to the telco to the consumer can take a piece of that pie.

Guy Daniels, TelecomTV (22:35):
That's fascinating because I was going to ask Colin, this is all well and good and we're talking about data as a service network, APIs, there's a lot of things telcos can offer, but do the OEMs see the value of this? And we were talking previously about say, subscriptions for features that may make the drive or passengers life slightly nicer, but someone's got to pay for them and it's not going to be me as an individual who particularly wants to pay for them. So I guess I'm expecting the car company to pay for them, but the car company's thinking, well that's another line of my balance sheet and no thank you. So how on earth do we do it? So there's some interesting ideas here.

Prof. Colin Lee, JLR (23:11):
It's a good point. I mean OEMC value differently as I said before, but for us to be connected, somebody has to pay for it. And this is one of the biggest things we have a problem with to get over the next hurdle. And I guess you guys are the way forward on that because how many people here have a subscription on their vehicle?

Guy Daniels, TelecomTV (23:34):
I don't see any hands going back. Hands definitely don't. Two or three,

Prof. Colin Lee, JLR (23:38):
There's one.

(23:39):
So OEMs tend to put a period of allowing the customer to have all the access to the service. When you first buy a car, the idea is you get the feeling of it, you get to see what's going on, but along sudden it turns off because clearly we are saying, would you like to read and subscribe? Of course they do subscribe, but it's not as much as we would like, right? And that's right across the O em. It's not just one OE em. So how do we go from that moment where we could make the whole lot free forever and put it on the beginning but we're still not going to, that model will not last. It's certainly not scalable. So we do struggle in the sense of saying, okay, and then of course you're looking at what do you offer? So we offer maybe this today might offer another two services, then perhaps another 15 services later when we add V two X.

(24:28):
But if the customer doesn't understand what they're going for and the customers may not want some of these things or understand why that's important, then turn it off and realize their mate's got something that they don't have and they don't understand why they don't have it. So there's a whole host of things behind here we're struggling with at the front end and there's no clear picture to resolve that we are doing little and often, just as what you said there Maxim is we take a long time, there's a reason for that. We're trying to see does it make sense? We talk to our customers, we get their feedback. Did you know that happened? No. Oh, okay. Is that because we didn't tell you at the beginning or is it because you've not learned since or something like this? So those are the things that we are still learning as we go through. We know what to do on cars, we do that all the time. But the connectivity side of it is it's a different mindset and what we could do is just amazing. I look after strategy in the vehicle, connected data platform and all I can tell you is that I've got a plan that's going to make the future look great, but without connectivity I've got, let's talk about Colin's plan.

Guy Daniels, TelecomTV (25:31):
I like your plan. I know unfortunately we probably have to wait a few years to hear your plan from you at the right time, but we will and we will look forward to hearing that.

Prof. Colin Lee, JLR (25:37):
But we are looking forward to connectivity in a big way. Why not and why wouldn't we be? Absolutely

Maxime Flament, 5GAA (25:44):
The whole point, we'll come to SDVI believe, but the whole point subscription based kind of economies is to end up with a vehicle that is so personalized mean. So when you purchase today or perhaps 10 years ago, you spend a lot of time just picking your options and then saying, oh, is it worth it? Et cetera. Now you will have one year, two years perhaps to even choose which options you want perhaps one year after you say, oh, but I would like this special feature that they're offering. And then you press on the button and then two months later you just say, oh, I don't want it anymore. And then you take it off. But it's all about giving the opportunity after the purchase to get these options on board.

Warren Bayek, Wind River (26:40):
Right. Can I mention something about this? So the subscription story's interesting. No one raised their hand. If you asked that question 10 years ago or 15 or something, whatever, how many, I didn't have my phone with me, but how many subscriptions are you paying for that run on your phone? I bet

Maxime Flament, 5GAA (26:55):
Exactly the same.

Warren Bayek, Wind River (26:56):
Two people would've raised their hands today. I bet everyone in this room would. And I think the same thing is going to happen in cars. It's culturally strange for us to pay for something in our car today, but it's because we still consider cars just transportation. For the most part though, we're starting to change that perception even amongst the population. But in five years, none of us are going to consider cars just a way to get from A to B and that's going to drive the subscription

Guy Daniels, TelecomTV (27:26):
Off and that's going to move us nicely to our next talking point because I want to go back to your big picture at the lectern there looking ahead and to me ahead is the software defined vehicle at some point and we hear a lot about it and what it like you can do and be Warren, I'm going to hopefully start with come straight back to you for this one because I know this is a particular adjust to you. So can you explain what the software defined vehicle is and what it may mean for both our industries?

Warren Bayek, Wind River (27:53):
Yeah, well so the software defined vehicle is exactly what it sounds like, right? It's putting everything about a car into software. So it's basically turning a car from a bunch of hardware components that make the car work into software. So software that runs the braking system, just everything about a car. How that's going to change the world is you want to go from the 10% growth bandwidth usage back to those a hundred percent a year. The software defined vehicle is what's going to do it. Frankly. I just saw a study by IBM that said if just 20% of cars become software defined, meaning overall including high levels autonomous driving, so bear with me, they would generate hundreds of zetabytes of data and for those of you that don't grab your phone for Google, that's 21 zeros. So that is a staggering amount of information. Now, not all of that will leave the car granted, but a large percentage will. So the software defined vehicle in the end is going to drive a massive uptick in bandwidth requirements and a lot of the other things people have talked about which is you're going to need slicing and guaranteed timeframes to be able to get to bandwidth and connectivity everywhere will be critical. So all of those kinds of things are going to be a requirement to make the software defined vehicle everything that it could be.

Prof. Colin Lee, JLR (29:18):
I'd like to add on that because very true of course, but what I think we like about software defined vehicle is you would have every ECU or every part of the car will have an API associated with it. So you can quite easily sit on your bench with a whole car in front of you and you can do millions of tests that you could do in the market. You could say, well what happens if I make it minus 3 million degrees and all of a sudden you get an answer. So we can do an awful lot of pre-testing, lots of pre what happens if we did this? What happens if we did that? How do we adjust the suspension in this situation? We can do an all kinds of things within a small area and then when we then take all this into the real car, which basically is a car, but you've now put it in a car, you should have the same feeling. It's a bit like an aircraft kind of where you don't fly an aircraft, it's simulator, right? So I think that gives us a really, really big advantage of what we're doing. There was something else but I've now completely forgotten it'll come back to it.

Guy Daniels, TelecomTV (30:24):
I'll come to Maxim because is this the direction 5G AA is moving to try and facilitate and accelerate the adoption of this?

Maxime Flament, 5GAA (30:36):
Well as DV dvs certainly a part of our work program now where I was talking about over the air firmware update or updates or map updates, et cetera. But all these can be opportunistic kind of updates. So when you get to a wireless line, you just make your update or something like that. Where things are changing is when the feature or the product is actually relying on the connectivity to actually perform its function or perhaps do it better let's say. And there we are entering a difficult point with the mobile network operators. Why would a feature rely on quality on demand network API when only you have only this kind of network feature network, API in Germany and Spain let's say, or perhaps a little bit in the uk. So we need a comprehensive knowledge of what kind of APIs are going to be available at a region level. Otherwise the vacuum manufacturers will just not pay attention to these opportunities or perhaps miss the opportunities and then not build up on this relationship, this great relationship between vacant manufacturers and MNOs.

Guy Daniels, TelecomTV (32:02):
I was supposed to go

Prof. Colin Lee, JLR (32:03):
On, I just remember what I was going to say. So having done all this, we then put digital twin up in the ethos, which is basically as your car leaves the factory, it's exactly what it is and we can now monitor, I mean all OEMs are looking at this. I know this because we talk openly about how we can improve things in life, but basically with five GA under control conditions. But the idea then is that as the life of the car goes through using the connectivity, you can start looking at the digital twin versus what's here, see the differences. And if that car itself is starting to go wrong, you can deal with that car. But if it turns out to be a collection of model, a model of a car across the whole world that's got a particular issue, you can start going back to suppliers and start improving it that way. So there is a whole host of things we're really interested in doing and this is just open, we've been talking openly in 5G about pushing this capability, how each individual OEM deals with that is up to them, right?

Guy Daniels, TelecomTV (33:05):
I'd imagine the telcos here would like all that connectivity to be cellular provided, but there's going to be a mix surely. And we have to admit there are not spots. There's plenty of not spots around. I live in a not spot, there's lots of 'em around and there's tunnels and there's places where there's no coverage. So where does this bring in? I mean satellites going to play. NTN will play a role somewhere tunnels, what's that in building type coverage where you haven't got access? What do you do there? There's a lot of scenarios I think where the connectivity then drops off. How do we smooth that out? Do we have any idea yet how it's going to happen?

Maxime Flament, 5GAA (33:41):
There are a couple of demands. Let's say first on predictive quality on service. So just not only about the hotspot or the lack of coverage, but just the drop of quality of service. That's an important point. So when you approach an area where your quality of service drops, I dunno from a hundred megabits to 10 megabits per second, whatever you need to know it. And then the functions in the car can just cope with it because they perhaps ask you to go from level three automation to a level two automation and then you have to put back your hands on the wheel and look ahead. But all this to say that they are mechanism. Now you mentioned non-terrestrial network and that's a wave that is coming up, which is at the moment using very narrow band kind of services and perhaps for answering the needs specifically for emergency calls or resilience, but not yet. And certainly not replacing the role of the Ian network,

Guy Daniels, TelecomTV (34:55):
Not but augmenting,

Maxime Flament, 5GAA (34:57):
Yeah,

Guy Daniels, TelecomTV (34:57):
Augmenting the role.

Maxime Flament, 5GAA (34:58):
Yeah,

Warren Bayek, Wind River (34:59):
I'll build on that a little bit. That's another thing that cars will do and are currently actually doing already is as we talked about, you're never going to rely completely on the connectivity to do pretty much anything. Cars now can understand and with all the intelligence and networks now when they're going to hit a tunnel, when they're going to hit a dead spot, when they're going to hit a 3G spot and ahead of time the car makes the decision to stop using the things that require the connectivity I currently have. So that's another thing that will get only better and better going forward in the future.

Prof. Colin Lee, JLR (35:34):
A good example of actually is if you are in say a kind of platoon situation where the advantage of platooning, of course you can be much closer. I've actually done some experiments myself where I've been no more than 10 meters, bear in mind now cars of 4.55 meters long and we're only 10 meters doing 70 mile an hour, 10 meters on our test track by the way,

Guy Daniels, TelecomTV (35:57):
Thanks for paying.

Prof. Colin Lee, JLR (35:59):
We're doing 70 mile an hour and then the guy in front slams his brakes on and you can imagine how I felt. Well of course because it's V two X, as soon as that break, it saw the break go within 50 milliseconds we breaking. So that's a classic example where it backs up what you just said there. That experience does happen.

Guy Daniels, TelecomTV (36:19):
Let's face it, this is automotive. We're not just talking cars are we? We're also talking lorries. And platooning I guess would be a great application for

Prof. Colin Lee, JLR (36:26):
What the trouble is of is once you go into the tunnel, and especially if you're going through a third party, and this is why we not think in that way at the moment, it's okay if it's direct because it still works in the tunnel. But if you went through, I don't know, say even an edge technology where you're going through say the local or thereabouts cellular, the tower, the trouble is there. Of course, soon as you're going to the tunnel you lose connection. Now if you are 10 meters apart and he puts a brake on, you're doomed. There's nothing you can do about it. So we have to make those decisions based on the quality of service, which is really important to us. Sorry, I wanted to close that one off.

Maxime Flament, 5GAA (37:02):
Yeah, that's why predictive quality of service really comes into play. Now there is a challenge here you are in the scenario where you are only connected to one mobile network operator. Usually you have two, three or four. And there is the connectivity of the edge of the mobile. Edge computing needs to be connected between different MNOs, the pro, well it is standardized, we have all the tools to do it, it's just not implemented yet. And so again, another hurdle in terms of adoption, okay, why would I need to use the edge compute when they are not interconnected between the MNOs and eventually not possible to actually detect a heartbreaking in your header like describe.

Prof. Colin Lee, JLR (37:55):
So based on that, we'll definitely step back on that type of thing. So we would typically not connect that type of use case through a third party because for several reasons. One, we dunno where it's gone, will it turn up on the time we expect it? There'd be one reason. And the second is if you do end up finding, you've lost it. You all watch Del Boy leaning against the bar and he stands up and he goes back to lean on it. If falls through, well if you're relying on something, you're okay. But if you suddenly find it's gone, boom, you fall over. So I think a lot of these guys here would recognize what I'm saying there. OEMs will, the automotive world will start relying on connectivity a lot more as time goes by, but the trouble is we have to make sure that every time we get more reliant on it, we don't get the devil boy syndrome. I think that's really important to know.

Guy Daniels, TelecomTV (38:50):
Well, Colin's Jaguar has done an emergency stop at 70 miles an hour on his test track and we've got to do an emergency stop now for this session. Nicely done. I thought we have to wrap up this panel now. Thanks very much for the discussion. It's been absolutely fascinating. I'd just like to say for our online audience, stay tuned, don't go away because this discussion will continue on extra shot with Tony Pulos. That's going to start in about 10 minutes time. Tony and his guests over there ready to take your questions, so please send one in to Tony. And don't forget, take part in our poll, our online poll. If you haven't yet done so, you'll find it on the website for our in-room audience, charity pinball. Don't forget coffee. Absolutely essential refreshments out there. We'll be back on the main stage in 30 minutes time. So let's have a round of applause for our guests. Thank you very much.