Autoplay: ON Autoplay: OFF
Detroit Auto Show 2017
V2X Breakdown: What is my future connected car talking to, and why?Roadshow Editor in Chief Tim Stevens moderates a panel explaining connected-car technology and discusses advancements in the road ahead.
[MUSIC] Hello and welcome to the Roadshow via CNET's live stage in the North American International Auto Show here in Detroit. I'm Tim Stevens, Editor-in-Chief of Roadshow. And between here in Detroit and last week in Las Vegas, we've been hearing a lot about connected cars. But not a lot about win and how we're going to get them. Joining me now to discuss that are Paul Mahoter, Director of Connected vehicles at Audi. Patrick Morgan, VP and GM of Ades product line in XP. Roger Lonto, Associate Director of Strategy Analytics. And Matt Smith, who is Intelligent Transportation Systems Manager of the Michigan DOT. Before we start talking, we put together a short video to help level the playing field. We live in a world where everything is connected. There are smart crockpots and smart thermostats and coffee makers, alarm clocks, even smart frying pans And yet our cars still don't talk to each other. Sure there are plenty of cars that you can start from your smart phone or even your smart watch. But still in 2017 we don't have a reliable way for one car to radio back down the road to another car and warn it about traffic or road conditions. The technology is called v to x or vehicle to whatever communications. In the eyes of some future cars will be able to talk to each other and to smart roads and smart traffic lights. Always knowing exactly what's happening around them. [MUSIC] But yet, over a decade after this tech was proposed, it still hasn't happened. And so, what is the state of v to x? Are we any closer to this ideal being a reality? Or have we gotten to the point where the industry has evolved, and doesn't even need this anymore? It turns out when I said a decade, I was actually a little bit cutting it short. Pom, how long have we been talking about V to X anymore, is it V to E, V to I, V to what? It's interesting, the first time that the ITS organization had proposed that spectrum be set aside For active safety related applications for V2V communication was in 1997. So 20 years now. It's been 20 years. Yeah. That's quite a while to be talking about a technology revolution for transportation. It has been a long time and I think that at the time there was also a lot of understanding that this In order to achieve the kinds of communications that we were looking for, then you would need to have a standard that was based on something like an [UNKNOWN] 2011 type of technology. And so, the the DSRC center was actually proposed, because at the time, that was the fastest available. And if you look at cellular at the time, cellular was Analog. And so what has happened over the years if you've started seeing more and more of the data application now happen to senator and I think the dialog now is more around what still needs to happen to the src and what can be achieved [INAUDIBLE]. Maybe we're looking at more of a harmonious combination of these two technologies working together. At Audi we actually look at connected vehicles as a pie. And each sector of a pie is a domain that the connected vehicle actually touches. So we've been walking our journey from vehicle to destination, to vehicle to passenger, to so on. Vehicle to infrastructure we actually touched last year with the launch of a service called Traffic Light Information in Las Vegas. We will be rolling it out nation wide but what this does is it uses LTE technology in the car and the fact that some cities have already managed to centralize their traffic management systems, and use the data from these centers, run them through smart algorithms in the cloud that use machine learning. To create a forecast, to advice the driver of when the light will actually turn green so this has been a quietly, quite an interesting way to see how you can use two different ecosystems that evolve independently. One on the city side, one on the color side and bring them together in an intelligent way. That makes sense to the driver as well. And we're gonna be talking a lot about ecosystems in different platforms. You touched on basically mobile wireless networks as well, which we'll get into a little bit because that's, I know, a little bit of a hot topic. And I think we've had some differing opinions on that. Patrick, what is some of the big advantage of this technology? We talked a little bit about traffic lights there and other communications. But what sort of things can we expect from this sort of communication? What is the grant idea of having all this This connectivity in our cars. Yeah, it's a great question, I mean what I think the question that we also ought to think about as drivers is what is it that the benefit that's provided V2V technology compared to other technologies that are in the car? And let's also go back to what is the function of an autonomous driving vehicle, I mean at the heart of it, The function of autonomous vehicle is to safely transport us from point A to point B. So what does V to V provide that the other sensors don't? There's two aspects. The first is it provides an ability to see around corners or communicate with traffic lights or other cars or pedestrians. And secondly, it provides the ability to do this in near real time. So, therefore what are the benefits? Let me maybe walk through an example, because it can provide near real-time and ability to see around corners. You can even see around objects or other cars. Imagine a situation where you got a line of cars in a highway driving at high speed. And imagine a situation where perhaps one of the cars on the road needs to emergency brake due to some unforeseen event. Well using V2V, the message can be sent from that car, received by another car that would be equipped with V2V, and then be able to take evasive action or perhaps emergency brake And in that line of cars, the intention there obviously is to save lives and to take a face of action. Where ultimately, doing this safely and securely, and saving lives. It seems like an ideal situation, something that we would all want in our cars, and it's something we've been talking about 20 years. Roger, why is this taking so long? [LAUGH] It's a very broad question, but are there some reasons why this has taken so long to get to the point where [CROSSTALK] could talk about implementation? Because the intention was to put it in a large volume passenger vehicle segment, which means the standards are going to be much higher, the requirements in order to put it in there. That hasn't done anything to bring down the cost, because the cost is still going to be around $300. It is a little bit Dumbfounding when you look at the mandate itself is solely for a here I am message being broadcast ten times a second, okay? The beauty of it is its simplicity, so So a car is announcing where they are. But of course, you have to have that built into every car. So some of the applications, Patrick could've referenced for example. So he was talking about automatic emergency braking, maybe a car or two ahead you can get notification of that car emergency braking without seeing that car. Although we are assuming Reception of the message which actually isnt mandated and the applications aren't mandated but they are very interesting applications so for example if I'm going down a two lane highway opposing traffic and I want to pass a car. I could get a notification telling me if there's a car coming, you don't want to pass right now. Or if I'm trying to make a left turn, don't try and turn in front of this car you can't make it. Or the traffic light connection is an application that can be done with cellular and DSRC or together, actually, based on the, you know, how close you are to that traffic light. So, it's a compelling application, but the devil's in the details, so after 12 years of 20 years spectrum was set aside. For 12 years, I'm familiar with the work that's been going on, on setting the standard na writing the code to support this. And there's still quite a bit of work to be done while we are within site cuz you're asking at the beginning, are we any close? Of course we're closer but there's a big obstacles to overcome Aside from the various protocols of the National Highway Traffic Safety Administration security. Security wasn't such a obsession, I'd say 12, 20 years ago. Now, the government is in the position of proposing to mandate an attack surface in the car. And depending on how we deliver the Credential revocation list, whether we're going to do that over cellular or whether we're going to use Sirius XM satellite to let vehicles know that there's a bad actor and to block that signal. It's a significant challenge to overcome, how we're gonna communicate that information. Still to be determined. Also, infrastructure was not part of the mandate Either. And some people look at it as mandating the creation of an entirely new network which as you might imagine, probably is gonna cost a little bit of money. Absolutely, can you talk a little bit about the infrastructure or we We talked about cars talking to each other, but there is something of an expectation that the municipal governments will have a hand in this as well, with being able to transmit information to vehicles about lane status and that kind of thing, and that requires some investment on that end, too, right? It does require some investment but it's still exciting about the potential that the technology has When you look at, obviously the vehicle to vehicle communications, and when you're driving down crash rates is always huge. But there's a couple of key applications from the infrastructure side that speaks to the same goals that most transportations agencies have, which is getting people, I think I already mentioned here, safely from point A to point B. And there's a couple of applications out there. One in particular that we see is the SRC base, others that can use emerging and different communication technologies that require some level of investment from the infrastructure site, but I think the benefits we can receive from it, rival, putting money and building a [UNKNOWN] intersection You're proving the geometric from the intersection. If you look at something like the red light violation warning. In this application you have the traffic signal broadcast over D.S.R. C. The traffic signal phasing timing from the signal. So at 10 times per second. So equipped vehicles like Roger mentioned under the NIST mandate equipments are equipped. To transmit but, conceivably if we're broadcasting the information out, vehicles can also be equipped to under the same communication protocol, receive the information back. So, it can receive the signal pacing and timing, receive Then in the vehicle, the application onboard the vehicle can determine based on the vehicles position and heading, hey, I'm going to be running this red light and start broadcasting warnings to the driver. That, to us, is huge, because one of our biggest safety problems are red-light crashes. And if we can start driving down those crashes with a minimal investment and infrastructure, that's big. Is some kind of investment in infrastructure. There is three, $400 per intersection minimum, depending on how modern your signal control is. It could be upwards of five, six, $7000 to equip a signalized intersection, to be able to broadcast out this information. But still, when you look at what the alternatives are and what you're trying to do, it's an easy case to make from a benefits cost ratio But there's big education component to it to the public agency standpoint. We are, it's not like there's gonna be a big pot of money to deploy this kind of infrastructure and application support. It's going to be competing for money, the same money that's used to fill pot holes and improve bridges. There's an education component to it where we need to Educator agencies about the benefits of the consume, we can see from this technology most likely. I wanna make one clarification with regard to the Here I Am message. It's not a priv, there's no privacy violation implied in there. First of all, there's no data saved and everything's anonymized anyway. So extensive measures have been taking to ensure privacy as preserved. It's not a data collection platform, so you were talking about data, etc. It's not really implied here. I think that you asked the question why it's taking so long. I think part of the reason is not just maturation of the technology, but the business module as you mentioned If making this effective requires every car or most cars to have the technology and to have every intersection or every part of the road to also have the technology there's a major investment that needs to be made in putting that technology in. Now I think that on the OEM side. We've seen a lot of efforts because of the growth of the connected vehicle space. Where we were able to build a commercial model for putting elements of that technology in the car already. To the point that if you needed to upgrade to the next level, it can be done very easily. What we haven't seen yet is the rollout on the infrastructure side, on the same technology. And so what has happened instead is The cities have been working on connecting existing traffic infrastructure, for instance traffic lights. And when we did our plans for the launch of the TLI service, we actually found that there's like three tiers of cities. There's cities that have had the benefit of taking a lot of the decisions over the years, where they've centralized Their traffic management, they have a lot of the intersections already equipped with lights that are connected and centrally controlled. And they've kept the software all up to date. So for cities like that to switch over to a service such as the one we have in the Audi now is actually a very minimal cost proposition. I think when we were having this discussion in Las Vegas at our launch, Tina Quigley who's the head of the Regional Transportation Center in southern Nevada mentioned that it's $6 million to build a mile of asphalt. Versus, they probably got this done with about 1,100 connected lights for probably less than the cost of an Audi. So if you compare that The return benefit to the city for doing that is, during the development phase, they've been receiving information from us, from our engineers where they might be getting a strange signal that could be a result of a pedestrian button that was stuck, it could be a camera that is off. And these type if information would normally require investment and a lot of infrastructure on site in order to get that. And then over time, What you're able to do is provide quarterly information back to the city on traffic patterns which can help them understand how to manage this. So a tier one city, for them a decision is very easy. A tier two city is where they do have connected traffic infrastructure but they may not have kept all the softwares up to date so there is a little bit of investment that needs to go in there. But then there are some cities that don't have that hardware in place at all but what do you do about those cities and I think that those tier three cities I think it needs to come down to are we going to follow the. Vehicle to cellular play and leverage that technology or should we just invest in VSCRC right now and use that as a platform? And I think that's an unanswered question at this point. Well the federal government obviously has its thumb on the scale as to Yeah. how they, but it's a pretty light Thumb. It's not a very thumb. There's not enough money to really push it through. But I'm surprised you didn't mention what the value proposition is to a city, for the traffic light information service that you're offering. Which is congestion reduction, greater fuel efficiency and Potentially even life saving elements. Absolutely, thank you, yes. I thought my monologue was going on too long. I needed to stop it. [INAUDIBLE] Emission reduction, I'm sorry, emission reduction. Yes. All of that from the congestion emission safety standpoint. And then, we've actually made a point of mentioning that if you actually observe people that are in these cars, that are talking to the traffic lights It's a different driving behavior. You're approaching a light and the system's telling you you're not going to make the light so you might as well start slowing down. You put your foot off the accelerator. You know, you kind of relax, versus trying to make the intersection. I think that has some additional societal benefits overall. You were saying, Matt? Good point that you brought up about. What cities I mean this [UNKNOWN] can do to prepare? We hear that a lot. We work a lot with US COT and [UNKNOWN] and [UNKNOWN] on some BSOC infrastructure, and BSOC based applications, some side ware based application. When we hear back from our peer who say, hey what can we do to pair for this [UNKNOWN] world? First impulse is like well I don't know. You have to wait and see before you start committing a whole lot of money to any one technology. But really what the answer is that we really get down to is you can prepare for any kind of communications technologies by making sure you have your infrastructures communicating on a backend. You have your back hall communication whether that's fibre, cellular, read the file type scenario. You have the back end systems in place, whether it's your own or you're contracting out or you have a third party. And be sure that you have your staff You'd be getting your staff up to speed.>> Yeah.>> On knowing how to operate in this new environment. You could go a long way and it doesn't matter if you are DRSC based communication, or cyber based communications. If you've got all the parts in place to start transmitting your information. However you need.>> Yeah.>> So that's.>> That's a great point. In fact one of the things we realized is. When that because the traffic agencies have never needed to send data real time to an external party. When they started doing it. It took a little while to get all the kinks removed. And so there were instances where the data wasn't always clean. Or there were gaps in the data. And so that was all the way from Systems not being up to snuff to the fact that in office hours, the system is sharing it with people doing normal work, and so you know those types if issues come up but you don't necessarily envision, but it does become part of the learning curve that needs to be crossed. Yeah, I mean, the real-time aspect of it is one thing that really differentiates some of the early standards that are build around, yeah, for vehicle to vehicle communications. That's a really good point to make it, this emergency breaking application is one, traffic light information another. Maybe that's also part of what makes the technology so exciting, is that there's so many different applications, vehicle to vehicle. Vehicles to obviously the infrastructure, obviously to even pedestrians and things, as we start to think into the future. Obviously there's going to be a disruption of business models. Even things like autonomous trucking, like we saw in the demo earlier, could also be a very interesting application. Yeah. What other commercial implications [INAUDIBLE] obviously safety considerations are very much the same, if not even more important as we're talking a large Tray carrying a lot of cargo. Are there other implications for commercial applications than there are for the consumer applications? Definitely. So let's think about that. So first of all, on the screen on the video we saw the truck platooning demo with the daft trucks. This was something that we participated in. We were quite excited by it. The way that the truck platooning demonstration worked, is basically the idea is that Trucks follow each other in a line, very short distance in between them. And the idea is that as the lead truck makes driving decisions, that signal gets sent over to the trucks in the platoon. And they're able to act as essentially one vehicle, moving seemlessly. We're able to show that that happened even with a latency, a guaranteed latency between the trucks, that actually was faster than the fastest human is even able to react. And so this is also a really really interesting type of application. Not only do you have benefits on things like efficiency, you know less drag and so forth in the convoy. But you also have Benefits obviously of things like safety and some other types of applications. So the interesting part about what you're describing is you were able to do that without any infrastructure involved, right. So this is cars talking to cars. And I remember way back in I think the 90s, there were demonstrations being done with magnets on the roads and ares platooning based on that. And that was supposed to be the way of the future and obviously We've now come to the point where technology has evolved to the point where you can use other ways for completely bypassing that investment. I think a big turning point of that was back in the early part of the development of the DSRC. Standards where the initial thought was hey, we could get this to work if we have one of those DSRC radios at every intersection and every critical part across the country. Well, that worked out to be about 250, 300,000 DSRC locations at $20,000 a piece. It quickly became apparent, hey, how are public agencies going to To pay for that, so let's start seeing, are there other V to V technology or something else that makes a lot more sense in the short term to pursue, then the V to I benefits can start water falling from those V- Those are the figures that I've heard. $300,000, $20,000 per. But in Europe, they've taken a different approach with V to V, they've pushed it more toward commercial vehicles and those kind of enterprise applications So, you have [UNKNOWN] in the Port of Porto in Portugal using day to day technology in trucks for logistics purposes. Certainly for collision avoidance, as well, and safety, and platooning throughout Europe, they're building corridors for trucks, but Enterprise applications makes sense immediately, and we'd be having a very different conversation here today. If all those guys who got together and women, twenty years ago or twelve years ago, said hey why don't we put this on emergency and service vehicles and commercial vehicles first. That would have been the quick path to market. That's where Nitza had a mandate authority and could have fast-tracked things. If you're trying to put things into you know passenger vehicles The protocols take much longer through in this. They don't have the power to just snap their fingers and get this implemented. Going with the cost of course nobody wants to add that cost unto the product that they're selling to consumers. Are there standards in place for anything we're talking about concerning the things like that? Are there any standards in place for that sort of thing? We haven't [UNKNOWN] interesting standards for one of those communications from one vehicle to another. But is there anything as far as protocols for transmitting information about vehicle state that would enable something like pontooning? So, I mean there's... I'll start. I think Roger you probably have a lot more detail on this as well. But the SE has a couple of standards that define the message set as well as some of the protocol level details. I EEE has some Standards for how the actual network technology level operates on the 802 level layer. So all of these come together and I think from our standpoint it's not so much about coming up with new standards but intelligently using the standards that are already there. And again I'll use an example from how we implemented the TLI system The communication that's coming from our service provider to Audi in the vehicle over the LTE channels is actually talking the same language that was defined by the SEGA 2735 Standard. It's essentially the same methods. It's using the same [INAUDIBLE] and map data sets. But we've trimmed it down to make it a little bit more efficient over cellular technology, and also because of these little messages were not really needed at that point. So the car is prepared to receive this information, and then the user interfaces, the intelligence that's being built in the car in terms of when do you show information to the customer, when do you not do it, what's the best way to do it All of that can be done independently. And then whether or not the transport protocol is over cellular, or it's coming over DSRC in the future, doesn't really matter. Because the car is going to receive it in the same language, it knows what to do with it. The other thing that I think that we do need to understand is the reason that we can actually see much further acceleration now. With these technologies is we've sort of divided the world into the OEM world and the infrastructure world, and if they could only talk. I think what we're forgetting is you can connect this world, and you can connect this world, but you need an intelligence layer in the middle. And whether that's in the cloud or in a data center somewhere It's machine learning. It's algorithms that actually take the data that's coming in in real time and have the super computing capacity to actually be able to build forecasting models. Build the information that actually needs to be channeled to the vehicle in the right way at the right time in order for it to be responding. And that is now available. So, for instance, our provider is Traffic Technology Services, and this is what they do for us in the cloud. But I believe for any DSRC loaded application as well, you will need that level of intelligence in the middle as well. There's another partner. So you mentioned infrastructure companies, and I've spoken to many of them, and they're completely frustrated. As you say, they don't know who to talk to at the car companies, and I'm not sure the car companies are ready to talk to them. But what about the cities? So I think the cities are looking at a scenario where they have to look at their transportation resources as kind of like real estates. So if you have a city you plan out the real estate, that's gonna be commercial, that's gonna be mix used, that's gonna be residential, that's gonna be I don't know nuclear waste dump, whatever it is. They have to start looking at transportation in the same way and suddenly they have a plan and different assets they can play with, public transportation, car sharing, ride [UNKNOWN] And then controlling access to the city, like London, [UNKNOWN]. So, and they're being forced into these kinds of decisions because we have an intractable traffic problem. Detroit is a good example, virtually no public transportation coming in to Detroit. So, the city is unprepared quite yet to take on that task, but they desperately realized they have a problem DSRC is a potential solution, but I think what they're really interested in, like Tina Quigley, who I know you know very well in Las Vegas. I couldnt think of a more enthusiastic entertainment transportation executive in the country. Is interested in speed to market. So you gave her a quick solution to deliver into the car over the a cellular modem in the car. That's what people are interested in today, solving problems right away. Exactly. [INAUDIBLE] And the fact that the DSRC is here today actually speaks exactly to that speed argument. And the other thing that's really super- [INAUDIBLE] Super important is to make sure that we definitely have a A strong partnership between companies. Different types of companies right. No question this whole idea around business models disruption, power companies and cities gonna work together to really make money here. Threre's a lot of different things. Ride sharing is one. Driver assistance is another, traffic light, traffic efficiency, truck platooning. There's a variety of ways that definitely as we all work together [INAUDIBLE]. Since we're talking about V2X here, I think we somehow make this connection between V2X and immediately jump to VSRC. And I think what the panel is telling you is that's not the case. That's one solution for a Large set of technology than services that need to come together. One interesting area that we're involved is in the city of Somerville in Massachusetts, we're actually working with the city there to develop in autonomous parking filer. And now this is now taking self driving technology in the car. And connecting that to the city's parking infrastructure so that you can have these cars that go off and park in allocated parking spaces that may be in parts of the city that are less utilized but now can be utilized for parking. And because they're all autonomous You probably only need about 40 percent of the space of a normal parking lot, because you don't need pedestrian walking spaces, you don't need stairs in order to get that all congested. That helps in removing the congestion from the city, as well as improving city planning in a big way. So I think that you have to look at this world as. Not just we came up with a standard at 20 years ago, and why hasn't it been implemented yet? It's like, the world has been doing a lot of things along the way. And what's gonna happen is gonna be a combination of all of these things coming together in the most efficient way in which you can make the offering available to the customer. You hit the nail from a head. Say from an end user, that public transportation [UNKNOWN] Standpoint, we are interested in the end result. We're interested in the application. We're interested in preventing crashes, more efficient traffic flow, whether it's urban or freeway, getting road weather information out there. Really the means of getting that, it really doesn't matter to us. Give us the most efficient way to do it. Maybe it's [UNKNOWN], maybe it's cellular, maybe it's a combination, maybe it's satellite. Maybe it's something else that we don't even know of yet. That really doesn't matter as a public agency. We want the benefits of the application. So, you're right. The beat the access, really we're focused on the end result, not necessarily the communication. But I think for federal regulators, they're looking at a hundred or more being killed everyday And the toll rising based on statistics for the first half. Right. Which drove, I think, the decision right at the very end of the Obama administration. We need to do something. This is the only arrow we have in our quiver. We're gonna string it up and fire this arrow. And the information that I have is that The incoming Secretary Elaine Chao is 100% supportive of the SRC. Unfortunately, the process nevertheless is still going to take time to come to market. But meanwhile, the US military is demanding this technology because they have equipment, they need to move all over the country. Mm-hm. And they want to start using platooning immediately. immediately. So, they have and immediate application. Again, commercial vehicle segment. Makes more sense. I speak through the (inaudible), through the VX (SP) applications, that have already been demonstrated, that are fairly mature. It's just getting out there into marketplace and into vehicle penetration. I think it's there. I don't think there's that much left in the development part of things. I'm not an application evolver just you'll see radio guy so I can't really say that 100% level of confidence but the majority is their and it's just a matter of getting out and making sure you've got the unusual. Users, the infrastructure people, the end users in city states, and the application developers and the vehicle manufacturers all going on the same path and we will hopefully that's why we're interested. We want to get that traffic fatality curved down to less than a 1000 people are dying on the roads every year in Michigan. We've got to get that down. And it's common, it's there, it's encouraging to hear that there's still, at least the next four years there's still support for that going forward and I think [UNKNOWN] Benefit of living outside Washington DC [LAUGH] [UNKNOWN] but are you going to ask us about cellular [UNKNOWN]. So unless if you're gonna take a step back, we've been talking around this a lot, do we have an Wanna actually delve into these actual discussion of cellular slash 5G slash LTE versus this DSRC we've been talking about. So let's take a minute, let's re-define these two things, and then let's compare and contrast them a little bit. Because I think we should do that. So let's talk DSRC first. That is what we've been talking about. That's the standard that's been coming around for 20 years now. Wi-fi based. Patrick, do you wanna iterate through some of the advantages of this direct point-to-point DSRC standard? Sure so if you take a step back and you look at the original intent of the standard, I mean 20 years ago it was definitely recognize that there needed to be some kind of standards developed in applied to cars that would provide dedicated short-range low-latency wireless communications. And what the standards group decided to do was to branch, and this is called a 802.11p And what the [UNKNOWN] Group decided to do was to use 802.11 as a basis, because 802.11 is a great standard that provides a really good framework for this. But the downside of standard 802.11 is it doesn't provide guarantees for latency or for bandwidth. So if you wanna have Real time communication, short range between cars you had to do something. And so therefore a new standard was needed. And that was the genesis of the AO2.11P And now so the other alternative would be to piggyback on existing cellular networks effectively that are built up around the country already, largely commercially run. Roger, what's the advantage of going to a cellular based system, a solution to connect one to another, rather than a direct Point-to-point to be. [CROSSTALK] So, the point of view of using cellular it's already going into a lot of cars today, LTE isn't just one protocol. It's advancing and developing, so, it's expected that release 14, LTE which is more or less 4.5G I will have a direct device device to communication which is not very well understood so I let a lot of people I talk to an industry that's with the exception of These Fine gentlemen there's a new interface that allows for direct communication that doesn't have to go to the Tower and so or less the same millisecond latency But it will take some time for that to percolate into the market. And it obviously takes longer to get into cars. So I would project that cellular V to V capability, technology, would probably arrive under an ideal scenario. Giving DSRC the benefit of the doubt, around. 2021. So an essence, that capability will arise simultaneously in the market. [UNKNOWN] has eluded to the fact actually that there's an opportunity for cooperation and collaboration between the two technologies. Cellular is very likely is going to be needed for credential [UNKNOWN] list And also, there are some applications like the traffic signal timing that could take advantage of both DSRC and cellular signals. But the point is, again, to Tom's earlier point, there's a business model for cellular, okay. We're finding ways to create direct, like selling WiFi data Or indirect business models or around that vehicle connection, which would involve data collection, okay? And of course security and privacy concerns, etc, etc. But that is the other revolution that's happening, that data collection, to take better care of the car, better care of the customer, mitigate destruction, prevent crashes and collisions. And create a more enduring relationship with the vehicle, a full product life cycle engagement with the car. So the business model is there for cellular. We're still kind of proving it out. Some car makers are still in that I don't want to know phase. But that's falling down. They realize they can't be in denial. With the 100 million [UNKNOWN] recalls. And the amount of software code that's going into cars. You need to be connected to cars, to do the software updates, to find those customers. So the business model is very robust for cellular. Right as you mentioned there are commercial incentives. You can tell services like the ability to start your car remotely and other things like that that are enabled by having a modem in the car and so we're seeing more and more manufacturers [INAUDIBLE] being on the cutting edge of having. I think in-car communications or vehicle-to-vehicle communications like this, you don't really have a choice. Because ultimately the SOC isn't really deployed anywhere else. But those modems are already in the car providing data for things like Google Maps and other things. Yeah, so it's an interesting time we're in right now. Because it may seem like the conversation needs to shift immediately towards Need to see because, you know, Roger was eluding to the 5G Standard which is suppose to come in the 2021 time frame and everything that we hear indicates that it will be very comparable, if not, even better in many cases to what the DSR super performance can be. Of course, yet to be proven widely and yet to be confirmed But if that's the case then you've got one check mark on the list of things that you need. I think there's a couple of other things. There's from a quality of service of the network standpoint, having a cellular network provider that is incentivized to keep that up and running is obviously a good thing. So if there's a commercial model that already exists to keep that alive Then that is a second thing that sort of works in favor of cellular. The part that is still in my mind still a bit of a question mark, and probably will be clarified over time is, what about the fact that you have active safety, and non safety [UNKNOWN] applications working on the same spectrum Does that cause any kind of a hazard in terms of what you need, because VSRC was kind of designed as a dedicated channel to only allow communication between vehicle to vehicle. Do you now if you allow other types of communication to happen in that same spectrum, enter that kind of an issue? And I think we hear both sides and we hear that the industry is Quickly understanding how we might be able to resolve some of those things. The bigger one, that I still don't see good a solution for is, the salary industry has a tendency to migrate technologies every 10-15 years. And so, if every 10-15 years, a whole bunch of cars becomes incapable of communicating on these technologies, or need to be upgraded, That's a cost factor that needs to be considered ahead of time. Or you need to find a way to obsolete those cars. Or if we really think in a more progressive way. I think what we need is some baseline standard that needs to be maintained. Even if the technology migrates from one to other. So that even older cars will have the same active safety benefits. I think As we look at these 4 questions, two of them clearly point towards cellular, one of them are little bit, is moving towards but, is still on the DSRC cam. The last one right now is where DSRC is strongest which is it is the one confirmed technology that will stay stagnant or at least, that's the conception. That's for 21 years from your chance. The one that goes completely against DSRC is that the DSRC standard is 20 years old In 20 years how standards for network technologies are written has evolved quite a bit. Yeah. Efficiencies, all of that has changed. And so it is an older protocol. It's like talking to your grandma over a cup of tea versus IMing or Snapchatting. And so that difference also needs to be kept in mind. So quality service is a big issue. You can't have your Netflix stream interrupting The communication from an outcoming car, security is always a big issue too, and playing up so much is definitely [INAUDIBLE] From a DOT standpoint, do you have any skin in the game here? Is there a difference or an opinion as far as which is a preferred implementation or just giving it to market as quick as possible. Yeah, absolutely, it's probably the biggest when it comes to the Sawer's standpoint. Great point for both before and for And don't know yet about cellular. [LAUGH] The big thing from us on the public standpoint is, we have to make sure that cellular applications or what have you aren't subscription based. We don't wanna have safety applications that we're behind it again. That might be my ignorance [UNKNOWN] to what the standards are going to be. Sure. But we have to make sure that what we're doing isn't only available to those that pay for a service or on those ends. And again we don't have enough familiarity with dealing with those standards yet or those applications yet to see how that's gonna work. That's a big thing. The point you brought up about the spectrum sharing. Everything we heard about VSRC and spectrum sharing. The same hesitation about going down that route. And relying on a spectrum that could be shared for safety and non-safety. That holds true for the cellular standpoint, too. If it's something that's going to be shared for data streams or safety related, non safety related. That start causes concern for us, too if there's going to be an intermingling, but those are probably the two big ones right there. Yeah, exactly. I think the key point that you are hearing from the panel here is that it really is driven by the application and also by the availability and the maturity of the different solutions. Our view is that absolutely there's complimentary applications. There's complimentary adoption capabilities. And from my perspective as a driver or a passenger in these things, I mean bring it on. Obviously I would like more technology in the care to keep us safe and keep us secure as drivers. As a technology provider, as one of the technology leaders in the industry, yeah absolutely let's use what we have [UNKNOWN] Get the wheels turning in terms of business models, adoptions, smart cities and so forth.>> It's funny what Matt said about not having to pay a subscription. I was on a panel a month or two ago in Tokyo where there was a Telenor Connection executive there a Wurth European wireless carrier. And they were talking about a subscription model for automated driving and I said, Wait a minute, I can't even always count on getting a connection to get a phone call. How am I going to have dependence for automated driving on a wireless connection to the cellular network? But on a more serious note. I was surprised you didn't bring up a challenge for the DSRC implementation which is, we haven't figured out what to do about the secure credential management system, who's going to provide that and how is that exactly gonna work. That's a big question. Yeah, I mean you're right. I mean we haven't quite figured out business models either, like Matt brought up the investment is not really clear where it's gonna come from And maybe what we're talking about is that there are certain areas where DSRC will be viable as a technology that can be financed. And then there's other areas where we have to rely on cellular, and the combination of the two is really where we need to evolve, too. Or maybe there's a smarter way of doing things. And Cars connecting to cars. Eliminated the need of putting [UNKNOWN] infrastructure. How far can we go with that idea? Especially if you're now the cars are also autonomous. So that they're not just dependent on the communication between the cars but the center of [UNKNOWN] is becoming a lot smaller, there's artificial intelligence being built in the cars that is learning the behavior of [UNKNOWN] And overtime can learn how to react to a certain situation. I think all of these factors are gonna influence the safety records that we have in this country to help that come down. And I think that the best thing we can do as leaders within our respective field here is to watch for What can be done today, and how can we make it easier to implement? And I'll be a little selfish here, our traffic light information service we think has a lot of benefits, and we'd like to roll it out a lot faster than we're seeing we're able to. And that's primarily because of the state of evolution of different cities. Some cities are up there, ready to go. There's other cities that need a little bit more time. And I think that you can do something like that with a comfort and convenience feature, like traffic light information, but if it's a safety-related feature, you don't wanna have a part of the population that's able to access that, but not a different part. How do you actually make these things move a lot faster, and I think what you're going to have to do is let the comfort and convenience features allow the spread of the infrastructure development to happen and then we'll switch on the safety benefits. Might be the way to get it out there. No one's going to buy a widget because it's got the safety capabilities to it, so. But the convenience factor, technology grow And building a safety. I think that's- Consumers want safety, in all of our surveys all over the world. That's good. That's good to hear. But I do wanna add an interesting vision, which was shared with me by an Ericsson executive, also a few months ago, which was the idea of virtual guard rails. Imagine if we can figure this thing out, you don't even need guard rails on the highway anymore. I mean, I know that's a little scary thought but If we can do it right, if we can get the coverage that we need, and maybe it requires both the SRC I don't know, but it involves the signs. Maybe you don't even need street lights anymore. [LAUGH] [CROSSTALK] More narrow lanes. Yeah. [INAUDIBLE] Footprint roadways, so I think what we can do with our budget if we get rid of signs and traffic signals- [LAUGH] That'd be huge. No one's gonna hit a pothole in Michigan [UNKNOWN] if we get to that point, so. We're talking around autonomy, and obviously, I think a lot of people connect to [UNKNOWN] autonomy, and almost bring them together in their minds, but We're seeing a lot of development of course in the world of autonomy, despite the fact that none of these [UNKNOWN] that stuff, it's really come to creation yet. So, the question is, do we actually mean [CROSSTALK] We've got one in Las Vegas that's working. True, that's true. There are some out there. Do we need [UNKNOWN] before we can get to autonomy? Or can the autonomous cars move on and Steam ahead forward even if we don't bring this to markets. Yes. I want to go back and touch on that point around safety. Absolutely we have consumers, passengers, or drivers absolutely expect that when we get into an autonomous vehicle it keeps us safe, it keeps us secure. Conversely, any type of possible risks. Is, has the risk to absolutely delay or disrupt the adoption, so therefore, it's incumbent upon us as providers and leaders to ensure that of these types of technology are needed for autonomous driving are actually in placed. Example I gave before is being able to see around corners. Even with other obstructions maybe around other vehicles and so forth. That's provided real-time with V2V. That's an important, essential element of cars today. Now as we move forward, are there gonna be more technologies, and more sophisticated aspects and more business models and things developing? Absolutely But I think the safest security aspect for autonomous driving is a really important aspect of it.>> I would say that as long as we're going to have mixed use in the roadway, which for the foreseeable future probably within most of our lifetimes, that's what we're going to see, autonomous vehicles with. With a dumb vehicle, so to speak, I think V to X is going to be critical. You look at what Audi is doing with the traffic signal systems. Yes, it's not DSRC but that's really still V to I, you're still getting infrastructure information broadcasted into vehicles, you're still getting signal information as another input into an AV So, I think as long as you have the mixed queues. That ability to communicate with other vehicles that aren't quite as smart or have the centric capabilities and ability to have the infrastructure assist in making those not quite a smart vehicle smart and behave better. I think there is still a market for that. I would imagine that 30 years minimum [INAUDIBLE] [CROSSTALK] That's going to happen for sure. Yeah. And then, I think the connectivity part adds to it. I think we were out of time [INAUDIBLE] and I think we could probably talk about this all afternoon. But unfortunately, we do have to close things down. Thank you, Paul and Patrick, Roger, and Matt for taking the time out here in Detroit to talk about V2X communications, and V2I and V2E and V2 everything