test test test test test test

test test

captioning performed by vitac it is becoming almost a standard. why are we connecting cars to each other? just as a wireless connection in the car or regular cellular connection isn't an obvious value proposition, why does my car have to talk to other cars along the road? why is that happening? i think the analogy to an intervehicle communication would be waze, most would look at this vehicle to vehicle communication

conversation and say, i've got waze, i'm already communicating with the other drivers and going to the server and coming back and data is being aggregated and interpreted and it's a great help for my navigation. what we're talking about today is collision avoidance. that is a direct kmuks with cars to help them avoid running into one another. cars notifying all surrounding vehicles of their position in real time, i think it's eight times a second. what kind of progress have we made in 20 years of development? not a whole lot. gm a couple of years ago starting putting -- the single lowest volume vehicle and the volume has continued to go down since they've made that decision, but it was a way for gm to say we want the spectrum don't take it away from us. we're putting it in our cars. come to find out that i think gm was a little surprised that

nobody followed their lead in this approach. voekz wagen meanwhile has been the next car maker to step up. toyota was before everybody in deploying in a large number of cars millions of cars in japan but on a different spectrum and with a different value proposition and completely different marketplace. vokz wagen has said beginning in 2019 the two cars will get the connections, only for europe though. it's important to understand the context, europe does not have a mandate and the mandate has been fought off and beaten back by the industry essentially both the automotive industry and the wireless industry. interesting volkswagen's biggest market is china, they will use cb to x which you'll hear about today. the energizer bunny, dsrc just won't go away. no one here is going to tell you

it's dead, even though i've been quoted as saying that. it's not going away yet, not debt yet. the news in this space is the voices coming from dj move in europe and the u.s. is technology, neutral. we're not going to put our thumb on the scale anymore. the significance of that, when we had the small city challenge in the u.s. for example, all of the proposals required dsrc, if you were trying to get a smart city grant you had to include dsrc in the proposal. it looks like that's going to go away. u.s. dot will stop pushing dsrc. some states have said -- state level dots will continue to push p but not all 50, maybe about 18. it's not a unified front on dsrc at this time and so the regulators in government is taking a step back and saying,

maybe the quicker path to market is for us to get out of the way and let the market forces take over. this is a little technical, i'm not going to get in too much of the details but i did want to highlight a couple of things about cb to x, which you might be hearing about the same time. cv2x is using almost identical if not superior capable. so low lat entcy and high band witne width and can operate without network assistance. i can't tell you how many senior industry engineer executivewise far more advanced degrees than i'll see in my life telling me -- carriers will never allow direct communications without access to the network that you're going to have to communicate with the network and there's going to be too much late tency for this to be a safety relevant application. it's not true and don't know why the confusion persists but

wanted to emphasize with this audience to make sure we're all on the same page, direct vehicle to vehicle communication without the network. and finally, again, they are using the same spectrum and there's other relevant issues here and a little bit ever qualcomm snark, so this comparison was sourced from qualcomm. what are we seeing? how does this unfold in the market and what does it mean for the average consumer? hopefully it means we're going to save lives because vehicles will be able to communicate with each in the early days. the problem as you probably sur miced by now, if a cadillac cts and no other car has it. the only car the cts can avoid colliding with is another model year 2018 cadillac cts. this is a problem. but volkswagen will have the

same problem, model year 2018 or 19 only avoid colliding with another volkswagen year 2019. the value proposition really only works once all cars or some would say a certain percentage of cars are so equipped. there is a day one capability that would be relevant though and that is vehicle to infrastructure. if you have traffic lights that are enabled with dsrc technology, they would be able to communicate with the cadillac cts but that application would have to be developed and likelihood is that it would be a very small number of traffic lights because deploying cv2x instead of using cellular technology is way more expensive. you're not going to go with a more expensive route. but that would be a day one application. ford has said 2019 they will do it because they were one of the early and hard core developers

for them to say they are doing cv2x is a very big deal. others have come out in particular and said we're doing this. i was just at the mobile world congress and talking with sam sung and qualcomm who say avoids have already taking place, car companies adopting 5g, it doesn't even exist yet. how are they doing that? the significance of that information is that typically car companies have been three or four years behind each evolution of the wireless network. so when we were getting 3x, cars were getting 2g, and until they discovered like gm when we went from analog to digital, that when analog got turned off, all of the cars got turned off and they had a class action lawsuit on their hands. the automotive industry is now working hand in glove with the wireless industry on this

implementation of this next generation and they want to have cars with the latest technology at day one. because that device will be in there for 15 years, you don't want to save moneys on an old network apology that won't be around in 15 years. so, the evolution nar path though, this is the qualcomm snark, they are saying, cv2x will be forward compatible with 5g and dsrc only kpatable with dsrc. haven't heard this a lot but there's a little confuse about interoperability, there's none between the two technology so the industry has to make a choice. they have to decide one or the other. they won't communicate with each other but cv2x will work on a 5 also g network in the future. and you'll have capable with

this kind of technology for edge commuting and vehicle connectivity solutions around traffic and other value propositions taking advantage of the network as well as working where there is no network. so the key is -- i want to double down on this message, the pc5 interface that allows direct communication between vehicles and infrastructure, which is not present in our current wireless network but will be available in a network that's been evolved to cv2 x and eventually 5g. so this is not a comprehensive list to the outdate, functional equivalent, wider spuk trum and business model, cost and consumer acceptance. these are the issues we'll be discussing today and that's my table setting contribution.

>> so we'll we'll invite the panel to come up and tara will take over. >> thank you all panelists for joining the conversation, i'm going to briefly introduce each person and let them get their own introduction of what they are working on and where they are on this topic. we've got michael calabrese of the wireless feet futures product and mark with the competitive enterprise institute, we've got mary brown with cisco, senior director of government affairs, and we've got daniel paneras with the

int internet and television association. and finally we've got roger who has just generously given us a great background on the subject. michael, i'll start with you. >> thanks for doing this. and you know as roger said, it's looking increasingly unlikely that there will be a mandate for dsrc as the specific technology or really any mandate for a vehicle to vehicle safety signaling. particularly if there's no mandate, the fcc should use the opportunity to make take an immediate fresh look at the high use of this fairly large and now incritical conditi

increasingly value spectrum which lays empty. this should be fairly obvious, quite a bit has changed since 1999 when it was first allocated. and i'll mention a few of those first, there is a revolutionary new auto safety technology to avoid most serious accidents, it's just not radio communications technology. as automakers develop ought mated vehicles they are already incorporating and rapidly improving sophisticated crash avoidance technologies that include radar, automatic breaking, ultrasonic sensing and on board sensors such as drousyness detection. they said v2 signaling won't be proven or effective for 15 to 20 years or longer, even if it's mandated. because it takes 15 years for the entire vehicle fleet in the

united states to turn over. by then we'll live in a very different transportation world. second, cars will be connected just not the dsr -- not by dsrc. even if the dsrc was mandated for vehicle to vehicle signaling, soon all new cars will be connected to the general purpose mobile networks for all kinds of other purposes. that should make v2x safety more cost effective if it's simply part of the general purpose network rather than being a stand alone proprietary dscc network. but it means the band can be reorganized to maximize the public interest since this would be starting a fresh. third, whichever technology is used there remains a critical distension between real time

safety of life applications, which must be -- which are narrow band -- must be uniform and interoperable or commercial operations such as getting an advertising -- ad flashed on the windshield when you stop at the red light near mcdonald's or downloading maps or swapping videos with other cars. we can use other spectrum for that purpose or share the commercial part of the band with wi-fi. when commissioners spoke on this topic two years ago, believe it or not and nothing has really changed, since then, commissioner o'reilly in particular emphasized that the noncritical safety use of this band should be shared with wi-fi. fourth, cellular v2 x is starting from scratch.

they are going to again valid dating testing this spring in southern california. now is the time to decide how it will use the public air waves. there's likely spectrum with better prop pull gas station and closer to core mobile carrier bands to be used. that minimum there's no reason that safety channels using 5g radios cannot operate at the top of the band and allow wi-fi, which is adjacent at the bottom of the band to share the rest. finally, the public interest must factor in the enormous and growing value of wi-fi. wi-fi bands are congested in busy places and in a 5g world, consumers will need much wider channels of shared spectrum that appear available only in the 5 and 6 gigahertsbands, the car band is a vacant desert island

smack in the middle of what will be the car band needed to make wi-fi more available and fast and approvable for consumers. it treemly really is time for look. >> thanks for hosting us here today. i'm going to cover -- i guess you could say two categories here. first i'll discuss some specific problems in the national highway safety administration's v2v notice of proposed rule making that was published by a week before president trump took office and then since been moth balled as a long term action as michael and roger have discussed. first, kndsrc approach would ha required they estimated nearly

50,000 roadside equipment units to be built out along the national highway system. but as nitsa quickly noted. they don't have funds to build out a network and don't have authority to regulate and manage the network. so when they were talking about mandating this specific vehicle technology that would have required a brand-new nationwide network, they basically said we'll figure this out later. and this -- you'll see this is sort of a theme that nitsa thought it could figure this out later rather than actually presenting an actual proposal to the public to comment on. big one here and this is likely would have resulted in litigation if nit sa decided to submit forward but they left a large glaring bracketed hole where the regulatory text discussing the secure systems. obviously we were going to encrypt the basic safety messages transmitted between the

cars and also you would have to have a way to replenish security certificates over time. they said we'll figure this out sometime between now and the final rule but we're not going to give you anything to evaluate. so lawyers can't look at this and most importantly engineers couldn't look at this to see what may be going wrong with their approach towards cyber security. so to avoid litigation and this would have derailed their proposal rollout, they likely would have needed to issue a supplemental notice of proposed rule making and opened up another comment period so commoners could evaluate that cyber security text that was omitted from the proposal. another problem, nitsa also noted that it likely doesn't have the authority to mandate

that users accept critical updates or replenishment of a security certificate. so what that would mean, if a user refuses critical updates, the v2v device becomes inoperative. it's no longer receiving or broadcasting. they said they could have tried to mitigate this kind of thing with not including an off switch but also installing a tell tale light or chime that would prod users, annoy users into accepting these wireless updates. but another problem since they can't mandate that users accept these updates, the privacy conscious or even just the apathetic, i like to call this the apathy rate not considered in the nprm, but given that we have about 10% of cars on the road today that display a check engine tell tale, now, so that's

mostly people who just don't care and taken -- they've taken the meaning of that tell tale to not be worth very much in terms of their safety or the operation of their vehicle, but if you have privacy conscious people who may be hostile actively hostile to this forced connectivity, all they would have to do is refuse updates and then the device on their vehicle becomes inop tif. you just don't do anything. finally, on these sort of deficiencies and really sort of as what michael was saying, the trump administration you know, has staked out a deregulatory approach. executive order 13771 which was the two out one in order requires cost neutrality and given that the regulatory impact analysis that was published along with this proposed rule

estimated a $5 billion annual cost and a $108 billion total cost by 2060, that would give the trump dot a lot less maneuverability in pursuing other goals if they were going to add this gigantic new cost, would have been by far the most costly auto safety regulation in a number of years. so i think part of it is practical because if they were to continue pursuing this, we would have -- they would have an issue in pursuing things that are their priorities as well. so it would limit their flexibility. then quickly, and i hope to touch on this more, but ought mated vehicle developers and many of the leading ones such as waymo, the self-driving car project and tesla and bmw weighed in and strongly negative of the v2v proposal. i think from the per sfektive of an ought mated vehicle, when they were looking at this and

remember, this was only hazard warnings, this is going to alert the driver of an imminent hazard, not going to actively intervene to stop a crash like automatic emergency breaking would. the best scenario for self-driving car developers would be they have to install this useless device that may display some annoying tell tales that riders and their vehicles can't respond to. so you might -- if you're in a full self-driving vehicle and you were to see this alert, you might realize that a car is about to hit you but you can't do anything about it. you'll be terrified going into a crash. but the worst case scenario that they -- that they -- and they repeatedly weigh in on this in the comment period, was well, if they didn't resolve the cyber security issues, what happens if there's some sort of link between the v2v device and the technology, the ought mated

technology that's directing the core vehicle functions, what if we somehow have a malicious attack on that. their point was, without any additional information from the federal government, all they were doing was increasing the number of attack vectors and there's no point of increasing the attack surface if you're trying to promote safety when you're then also cutting against another i think far more promising safety technology. and then finally, i would say that a top down dsrc could convert resources that automakers are currently spending in the upstart developers are currently spending on automation technology which can save far more lives and best case scenario under a hazards warning, shifting into this. i like to think of it envisioned by nitsa as the many tale of

connected vehicle technologies. it sounded great -- they may have had had a case back in 2005 to mandate this but i think the time has long passed and we're going to see with 5g far more promising connected vehicle technologies. so with that i'll let mary go. before i hear mary's top line thoughts, i would love to return to the idea of consumers having to update the system and i just imagine some government entity forcing people to update i phones, the country would collapse, right? nobody does those on time. i'd like to return -- and the check engine light in people's cars, how often do we all ignore those, probably too often. so love to return to that. mary, let's hear from you. >> thanks, tara and thanks to new america for having me today.

why is cisco sitting in front of you? because everyone involved in the debate is a customer or partner and we are trying to figure out what is going to happen. this is one of the oddest issues i have worked on in my 35-year career in public policy. two decades ago the department of transportation had the bright idea that you could take radio technologies and introduce them into the transportation sector to generate efficiencies, make us more safe and probably render some environmental benefits to boot. today what the transportation department has done is by any measure somewhat complete but not complete. with every passing year, the views about what should happen to this idea of using radio to make us safer has become increasingly splintered. and i now count five mutually

exclusive views, some of which you've already heard today. the first, dsrc is and should remain the intelligent transportation technology of choice. some auto manufacturers, the department of transportation itself, and the national highway transportation safety administration, appear to agree. more recently state highway departments added their voice to the chorus, that said, nitsa has not mandated the use of vehicle to vehicle technology. has not mandated dsrc. if you look on their website, it remains the significant rule making and it's a key part of the u.s. department of transportation its strategic plan and more infrastructure deployments are happening at the state level funded by tax dollars. new equipment is being introduced by vendors and just last week the ieee body stood up a stud die grogroup to refresh

updat the standard. cellular vehicle to everything should be the designated technology of choice, this late to the party technology challenger uses exactly the same spectrum as dsrc but not int interoperable technology. it was created by the global cellular industry in response to china's decision to move to a cellular based system to do this. we're still in the united states in very early days. it has not been tested anywhere near the level of dsrc and not been tested by government. it may well be a great technology. it may well be a better tech knowing than dsrc but that's a proof point that needs to be developed along with some explanation of what the business model might look like because that may have very different

implications for both auto manufacturerers and state highway departments and most importantly consumers. third position, its has too much spectrum devoted to it and that amount should be reduced in favor of repurposing the spectrum for more wi-fi. it is certainly true that wi-fi needs more spectrum and that spectrum has not been used and is just sitting there today. some parties are urging nitsa not to mandate any radio technology for safety because they say autonomous technologies are good enough. my own view on that based on the engineering studies i've seen is that whatever radio technology you talk about, dsrc or cellular based system, both of them see beyond autonomous technologies and would give the car more information to help keep the passengers in that vehicle safe. and i think the auto manufacturers would contemplate whatever system that is that's

ultimately comes into use that would -- that information would be integrated into the autonomous system. it would become another data input. and lastly, others argue, leave the radio solution to the market. but what's less clear is how would the market resolve the very real network effects problem. namely, we all have to be on the same or interoperable technology to receive benefits. splintering is never good for public policy decision-making, you want to bring parties together. i see this debate going in the opposite direction and i hope we can discuss more about the implications of all of this as we move through the discussion on q and a, thanks. >> thanks, mary. danielle, let's hear from you. >> in keeping with michael's theme of change, i wanted to talk about changes in the

marketplace and some things that haven't changed either since 1999 or more recently about five years ago when the fcc first opened the proceeding, we've heard about changes in the ought moat tif marketplace and i would like to talk about the spectrum environment and highlight the need for my wi-fi spectrum has not diminished in the last five years. in fact that need is becoming more acute, we're seeing fast paced growth and consumer demand for wi-fi and we'll also need additional wi-fi spectrum to enable speeds. there have been a couple of important studies using different methodologies that consume consumers we will need a gigahe gigahertz in the next few years to support wi-fi. we may need more than that when it comes to enabling unlicensed technologies. from the perspective, 5.9 remains the best near term option for additional mid band

unlicensed spectrum. why is that? it is immediately adjacent to the uni 3 band, most important in the world. by opening this band of wi-fi we thinks manufacturers will experience important economies of scale and providers can bring more broadband to the market more quickly. despite this rhetoric the deployment is just around the corner after 20 years as others have highlighted. these incumbent services remain in the pilot phase and may never see commercial deployment. there is no other mid band spectrum today that's so under utilized and so few incumbent operations. today broadband providers are delivering gig asbit broadband but it could become the bottleneck. they'll lack the wide 160 channels they need for the next generation broadband to consumers, which is how many of us experience the internet. so while we think 5.9 remains a critical band for wi-fi, there's been alet of change in the last

few years. there's a sift to cellular v2 , this is going to more cellular technologies that means arguments about some costs or kufrnt channelization might need to be maintained are really now suspect. we've heard from a couple of folks that the cv2x spek suggests the band could be used for applications but this is a new technology that may not need to be the case, we've also heard cv2x has a path forward to 5g and it may be appropriate to think more in the 5-g bucket and think about spectrum applications accordingly. i think the time is right for the fcc to take a fresh look and step back and think about what are the spectrum needs for this shifting landscape in automotive communications technologies. and what is the right spectrum home for these different

options. in fact we've heard some in the community suggest that exclusive license spectrum might be a better choice for safety applications than the dsc2 model. i want to talk about changes in the spectrum environment since the fcc first conceived of the 5.9 band 20 years ago. in that time, the lower adjacent uni 3 band widely used for wi-fi has seen widespread deployment. millions of devices here in the united states. what hasn't been talked about too much yet here today that upper adjacent band to 5.9. right now immediately adjacent are high powered fixed links and satellite communication systems and fcc has said -- opened a proceeding and notice of inquiry to look at authorizing unlicensed use about adjacent 6 gigahert spectrum. it may no longer be appropriate to have an island of auto

safety. might not be a good engineering choice to have an automotive safety band there. it may be time for the fcc to take a step back and take a fresh look. they propose the fcc issue a further nt further notice of proposed rule making for unlicensed use to help meet the growing demand for wi-fi and seeking comment on what the spectrum needs are for the applications that have changed in nature since 1999 and even in the last five years and to examine whether there might be other suitable spectrum for the communications technologies. >> roger, i'll go to you next. i summed it up but did want to comment on the statements made here and some of the things i didn't mention. i did want to highlight what michael was saying about the fact that there are a lot of technologies that entered into the market since the conversation about dsrc started. cameras in particular, more or less did not exist on cars as a

safety tool and radar was in very early stages and too expensive in most cases. we certainly weren't hearing about lydar, i can tell you in the industry we've gone from one provider to 50 providers and i don't have to tell you what that means for scaleability, cost reduction and the delivery of that technology to vehicles to enhance driving safety. dare i say without any wireless connection on the car. what we haven't talked a lot about on the panel here is what we're working toward in the industry, autonomous driving, enabling an autonomous experience with no driver and 5g is very likely to play a major role in that with what some of you may have seen today or in recent weeks and months, the need for remote control of the vehicle, okay? we've seen hackers do that but now it may become -- not only an

sae standard being defined for yet another level of ought ton me, remote control but a requirement. it may become a requirement that the vehicle -- something goes wrong, you want to have a fall back of some kind of remote control. i can tell you the wireless connection 5g provides precisely the band width you would need to do that although companies are doing it today with lte. believe me, i'm as skeptical as anybody about the scaleability of remote control of large numbers of vehicles but it is not only being contemplated but very may well be a required standard. in fact, finland autonomous driving law provides for the driver that there must be a driver but the driver doesn't have to be in the car. so draw your own conclusions where this is leading. but so i would be the first person to say most of the ought ton mouse vehicles do not have a wireless connection and they are

perfectly safe. but in the future, we're going to want to have all of the technology at our disposal that we possibly can have. a belt and suspenders kind of approach to autonomous driving and wireless technology will play that critical role and it very likely will be cellular. i will say i was at the satellite 2018 conference here in d.c. yesterday and certainly satellite wants to play a role in this game as well. but i think the points were covered very nicely and with very little overlap by the other speakers so back to you. >> roger, thanks for that. i do think it's important to talk about the role of autonomous vehicles today in the policy discussions. i for one wonder if we'll keep using the phrase behind the wheel to refer to somebody who is in control of something after we're no longer behind the wheel literally of our cars. but moving on, it's clear that this debate in some ways is

quite old in the sense that some idea of intelligent transportation and spectrum has been in the works for decades. but it's also very new with things like self-driving cars. so i want to shift gears and pun partly intended and ask michael, should there be a mandate at all for vehicle to vehicle communication? >> okay, we haven't focused on that -- on the -- we're not claiming to be safety experts exactly and mark should weigh in on that. i think really the -- some of the considerations i mentioned at the beginning are very important, which is given the -- given the cost and given how long it would take to become

effective, i think the administration reelg really needs to look at how much of a difference will it make. i think it was just mentioned the trajectory of driver assist technologies, the radars and lasers and cameras and all of that is only improving and it makes each car safer and makes the whole driving environment incrementally safer for each new car. the problem -- a big problem with dsrc that nitsa was fairly frank about in its notice of rule making is that it will take at least -- first of all, it will be a few years -- even if they adopted it tomorrow, it will be a few years before they start really doing it and then it will take at least 15 years for -- they said for it -- even know if it will be effective because you have to have a turn

over in the vehicle fleet which takes about 15 years. and it just seems that we're changing the nature of driving so radically over that same time period, and these things are so interactive with each other, as somebody nenmentioned, that it probably is better to leave it to the market but within obviously within certain bounds. >> roger, you wanted to weigh in? >> yes, as does everybody. from a mandate perspective, i thought you were going to say -- didn't quite say it, even if we were going to go forward, we could probably be eight years before anything took effect x. if you look at the backup camera man indicate, that was in the neighborhood of eight years and amazing amount of very enlight lightening research that had to be done on top of the research that had gone before to determine exactly how it would be implemented, et cetera, et cetera and all sorts of comment periods and a phase-in, et

cetera, et cetera. what we saw in the obama administration was a shift to a more of a voluntary relationship with the industry. so i would say boy, we should mandate automatic emergency braking at low speed but again, if we pursue a mandated approach, it would be eight to ten years before we were saving life number one. and so they took a voluntary approach to get everybody to commit to -- we're going to put this on the cars by a certain date without the -- there seems to be something fundamentally flawed in the regulatory process, mandating process. where there isn't a flaw i think, i'm not a fleet commercial vehicle industry expert, i think nitsa has more authority to implement mandates in the sector much more rapidly and we might be having a different conversation if the approach had not been to see dsrc implemented in large volume passenger vehicle segments but

in the commercial segment. then today we'd have commercial vehicles with dsrc technology and if you put it in the car you would have the added advantage to be able to detect large vehicles around you, you would immediately have a valuable proposition. but i think there's something fundamentally flawed in the regulatory process to get to a life saving proposition. i want to point out, the time it's going to take radio technologies into the fleet that is all of the cars that we brought, a concern regardless of what radio technology you pick, right? the average american is now holding on to their vehicle for more than ten years. and so it takes time for any new technology to be deployed, where it's semiautonomous, autonomous,

radio based. to me the question is does radio based capability for safety, does that -- is the increment of safety you're going to get off that from semiautonomous technologies that are in new cars today or from autonomous technologies, is the incremental benefit you're going to get from that worth the cost? and what we saw in the nitsa notice of proposed rule making from the end of 2016, they were concluding yes, but i will be the first to say technology moves fast and obviously before you go to a final rule you would have to do the analysis again to find out -- regardless of the radio technology you pick, is the benefit worth the cost. >> 1200 lives would be saved or something. >> the high end was up towards

1400 but that was at 2060. zblo i don't know if that was arrived at assuming a connection between the rf solution and the vehicle controls because -- >> no, that was just assuming -- -- >> a warning. >> which is assuming warnings although the auto manufacturers all say they would integrate that data into their semiautonomous and autonomous technologies. >> the ones that don't act everybodily oppose the mandate. >> i'm saying regardless of what technology -- if you assume a radio technology, it's going to become part of that system. >> i think the main difference between ought mated safety technology and radio technology you can start buying immediately once the auto mated technology is deployed. it doesn't matter what other cars are equipped with. and then in the case of the v2v mandate from nitsa, they were looking at the cost benefit break even point around a decade

after the phase-in began. so a decade after the mandate started to take effect. many argue that was overly optimistic but i think the point is some of the lower levels of automation take emergency braking, incredibly promising right now and starting to see this be deployed in the vehicle fleet today. we can save just as many lives as the probably overly optimistic inpacket analysis from nitsa as dsrc in just a couple of years as opposed to waiting decades to reap those benefits under the vision for v2v, i think -- i don't disagree that in the future we're going to see auto mated vehicles but that isn't what nitsa is talking about right now. and if you look at all of discussion, everything that's in that rule making proceeding,

that's still open at nitsa, very little next to nothing on -- on ought mated vehicle. it's all focused on hazard warnings. finally just to point out, i think there's been even though we're still have this rule making sitting out there open, it's now been redesignated as a long term action which means they didn't expect to work on this for at least 12 months. we've seen other things, maybe not as clear as a regular tri shift but saw the 2018 to 2022 u.s. dot report gut all references to dsrcv2 x strategic plan. and also we've seen guidance pulled back at federal highway administration on vehicle to infrastructure dsrc based systems. i don't think the administration has finished its work on this topic but i think there are in

addition to the redesignation of the rule making proceeding status, i think we've seen other indications from the department publicly that they are moving in a different direction than the previous administration. >> anybody else want to weigh in on what mark just said? >> i might add, sort of regulatory overhang perspective with this proposed mandate still out there, even though it's relegated to the back burner that's problematic for everyone involved. for automotive sector as it sorts out where it's headed with vehicle connectivity technology and problematic from a political perspective and trying to permit the fcc to move forward here and try to finally get to a place where regardless of what the decision is, we get more efficient use of that band before too long. >> i'm interested in the p perspectives what we're fundamentally looking at which i

alluded to in my opening comments, that we're -- for the first time talking about using cellular for an active safety function in the vehicle, which is quite radical, typically the cellular interaction with the cellular industry by the automaker habz for things like in -- automatic crash notification, sure, but safety? that's a whole different group of guys who take a very dim view of the wireless industry generally. and i'm curious, how the rest of the panels think about that fundamental shift. >> that's a great question, mary, do you have any thoughts? >> yeah, as i said in my opening, i think what -- we're lacking a couple of things, we're lacking nitsa administrator who can clearly articulate which way is the wind blowing inside that agency, right? just we don't have one. we don't have anybody sitting in

the decision seat besides an acting person who in good faith is just pursuing the agenda she's been given. but on this question, i think there has been a complete lack of clarity around the business model of a cellular v2x system. who would run it? do the carriers want to accept the liability of being the providers of a vehicle to vehicle crash avoidance system? would they prefer the auto manufacturers to take that liability? what would the business relationships be? for the -- from the perspective is how is the band used, to what end is the greater through-put of the c2 x system, we don't have clarity and i would curious

to know, i haven't seen any clarity yet coming out of the carriers, they are all very interested in exploring the technology to see what the technology can do, which is good, right? but i haven't seen any clarity about how the business how the model actually works. one thing i will say is dsrc, there's a clear business model and it involves the consumer paying nothing once you buy the car. once you buy the car with the radio in it, you don't pay anything any more. there's a consumer angle here that needs to be clarified. >> great point about the liability. i think that's kind of a running theme in 5g. once the network has every part of the consumer's life on it, who's liable when that device doesn't work or when something bad happens. michael wanted to make a point. >> one thing i mentioned in the opening is that fcc

commissioners who were speaking out on this issue the past couple of years, they have been emphasizing there's distinction between access as real time safety. in other words, what you need for safety which has to have low latency, be operable and available to everyone or it is not going to work. remember, this is a technology completely dependent on the network effect on every car having this. and every car being able to talk the same language. there's a big distinction between that and all of the other connected car applications that you might do, whether they perhaps using alternative, you know, radios on the same chip, and what i have been hearing recently from mobile carriers is

acknowledgment of that distinction. it is an important distinction. they're saying oh look, the government is going to have to tell us, going to have to say that every car for safety purposes, even if there's not a mandate, they'll have to say if you have a safety signaling that it must all be the same interface. it has to be inter-operable. they're saying it has to a peer to peer communication, so we're not talking about the way it works on your smart phone. it is not going to go through the cloud, back to each of the competing carriers' different networks. it has to be for safety, basic safety message has to be peer to peer, directly between the cars. and they're saying that can be

the same, use the same 5g radio, but probably something the automakers just put in and it operates and nobody knows who is in charge of it yet whereas all of the other things that you would do will be things that each individual carrier innovates around and competes on and that will probably go back through their own networks for the most part. so it is really in a sense two different things. >> michael, i wanted to follow-up on safety. how much spectrum is needed? >> well, needed and then also, you know, required. what nhtsa has in the notice of rule making, and this is true from the beginning, the requirement that all vehicle to vehicle signaling must be on a

dedicated signal. the band is 75 megahertz, they require that real time safety signaling be on 10 megahertz. they also have a notion to set aside a second channel for first responders because it could be at a higher power level. and in europe what the european union decided years ago, they had a proceeding, they said we think that safety requires 20 mega hertz, but we're going to have a third channel, so up to 30 megahertz as contingency, might be useful to have cars communicating with stop lights and things like that, even though that's not always time critical, it could be on the third channel. so europe is still going with 30, up to 30 megahertz, which is why one of the proposals when it looked like it would be mandated, one of the leading

proposals, cisco had a proposal and qualcomm had a proposal to say let's give safety its own exclusive 30 megahertz at the top of the band so that wi-fi can share the rest, so commercial applications can share between the src and wi-fi. the emergence of cellular throws in a wildcard, new aspect of can we still do it the same way, should we still do it the same way. on one hand, since cv to x is starting from scratch, would be easy to use top of the band for safety, but on the other hand less clear how well they can co-exist with wi-fi or whether that makes sense because each of the mobile carriers have their own 5g networks and separate spectrum. >> and how much control, i'm not

clear how much control nhtsa would have over the 75 megahertz in the situation and how it is divided up. >> just to highlight something about spectrum needs, important to remember for the vehicle to vehicle mandate, again, looking at one 10 megahertz channel for safety signaling. i also want to highlight, there's sometimes misperception this is dot spectrum, this is commercial spectrum governed by the federal communications commission. has an important role. it is not the expert agency when it comes to spectrum decisions. those are things congress committed to the fcc to decide. i think that's why it is time for the fcc to step up, step forward, given changes in the marketplace and start thinking about what the spectrum needs are in the new environment and take a flesh, holistic look at

that. >> just to apply some technology to the statements here, the dsrc technology was designed with vehicle to vehicle channel that is 10 megahertz, that is dependent upon the existence of vehicle to infrastructure channel so they can obtain their security certificates we heard about earlier, right? there's also a control channel at the top of the band. there was a channel state highway departments were going to use for their own purposes. but in a dsrc proposal you can't isolate 10 megahertz, they need something else to communicate to that radio. similarly, when cellular v to x came along, they took that same model but put in cellular technology, right?

so they also need some way to communicate with the radio that's on board the vehicle. whether that absolutely needs to be in the band or could be that communication link could be in another band is a question, it's a valid question. that's how it was designed. it was essentially designed as a cellular version of dsrc. >> interestingly roger had a slide up, i think taken from qualcomm earlier in the discussion where his presentation highlighted various sides. but sounds like there's a direct piece that they're thinking about that band for and already contemplate it operating over the carrier's license system to do the network communication. so again, i think given that kind of unique nate you aure, n

the time to think of those. >> they'll say it is using the same spectrum, can use the same protocols and deliver the same value proposition fundamentally, so that seems to be the intent. >> and we had a white paper from 5g americas released this week that says exactly that. so that's cellular vendors and operators saying it is the same spectrum. >> where does wi-fi fit into this equation? we mentioned wi-fi, several of you have, and i want to focus in on it for a second and talk about where it fits into this spectrum policy discussion. michael, you want first crack? >> no, go ahead. >> i think i pretty much said my piece on this, reiterate that i think we're still seeing great demand, growing demand for wi-fi technologies, seeing increased

consumer demand, congestion on the negotiation and peak hour scenarios, several studies forecasting need foray di additf the spectrum. both of those rely on wide 160 megahertz channels. right now in the u.s. we have one 160 megahertz channel to support everybody. i should say it's not restricted by dynamic frequency selection rules. so under most favorable wi-fi rules, one 160 megahertz channel. to get the first, 5.9 is the place to do that. the only place teed up today. i think this band is important for wi-fi use and to meet growing demand, enable speeds that carriers are delivering that wired broadband providers

are delivering into the home. >> just add that if wi-fi is going to keep pace in a 5g world, then it is going to need wider channels, more spectrum. so we often talk, one misnomer, we often talk about 5g as if it is a mobile carrier technology and network. but it will be even more diverse and decentralized than the current 4g wireless world, so currently we have 4g, but i'm sure all of you on smart phones are actually sending about 80% roughly, what some carriers said, 80% of the mobile device

data traffic is not touching carrier spectrum or network at all, it is going a short distance over unlicensed spectrum via wi-fi. everyone thinks they pay too much, but believe me, you have far less data if everything you were using actually went to the carrier network. but because it is going via wi-fi into the wire line that's within a few hundred feet, about 80% of the traffic, you know, that makes your mobile broadband more available, considerably faster to go indoors and a lot more affordable. when 5g comes along, the same process needs to happen. mobile carriers are using millimeter waves, getting more spectrum to have wider channels for faster negotiation. and we'll all enjoy that, except

that that will be too expensive unless wi-fi can keep pace with that. right now the only place for wide channels is in the five gig a hertz band, which is adjacent to the car band we're talking about. unfortunately the fcc, although it proposed to make as much as 750 megahertz, continuous megahertz available for wi-fi back in 2013, it found among other problems that the military said we can't share with certain kind of radar in the middle of the band. and now we're running into this issue of this band we're talking about today, 5.9 gig a hertz. that's the only place we know of in the upper five band and possibly into six band where you can get wide channels so the

entire 5g system will be robust, both the licensed side and unlicensed side. >> so i concur with a substantial part of what michael said, i will also elaborate further. 5g is the first time the cellular industry is going to have a radio access network, the edge radio, that's agnostic to technology. one of those technologies will be wi-fi. anyone who thinks wi-fi is not part of 5g as a technical matter, misinformed. wi-fi is part of that system and desperately needs more spectrum for all of the demand reasons michael raised. i personally and many of us in the wi-fi industry have become -- it is painful to watch the 5.9 proceeding and 5.9 activities not advance as quickly as we would like because

that spectrum is sitting there largely unused. cisco had a proposal some years ago when dsrc was the only horse in the race to try to share with dsrc. now we're not even sure if dsrc will eventually end up prevailing, although certainly it is sitting in the incumbent's chair. but it is painful to watch. i think at least at cisco we are spending far more time, energy, mind share on opening spectrum at 6 gig hurts because we don't know how to move it along to any resolution, whether it be dsrc, cellular to vx, how much spectrum will it need. there doesn't seem to be any way to move that forward. >> it is painful to see it so

underutilized. i think 6 is interesting. we have some incumbent satellite links there which is maybe a discussion, co-existence, might be discussion for another panel. i think they don't -- why aren't we saying six isn't the one. one reason the proceeding isn't as advanced. they kicked it off with noi last summer. in addition, there may be technical rules that have to be worked out. may be restrictions at 6 that might not be available at 5.9 in terms of ability to operate. still believe 5.9 is the best

option. >> what we are left with is not rationale. it is emotional, political. we spent $700 million on this technology. we have come this far. which takes it beyond a business model discussion as mary is trying to bring it back to and the real technical issues at stake and implications that touch everybody, not just car drivers. so this is what everybody heard is the untold story of the debate, what the wi-fi piece is all about. >> i'm certainly glad we could shed some light on an untold story. any final thoughts quickly from

anyone? >> sir? [ inaudible ] >> all right. so my point was simply wouldn't it be helpful first to establish

we ought to be moving toward objectives of dsrc, whether with dsrc or cellular v to x and then discuss band width second? >> i think the main problem we have now and what danielle referred to earlier, we have regulatory overhang at anhtsa over what vehicle technology will be used but also having negative impacts over at the fcc and on the wireless industry and equipment manufacturers. so until dot gets its act together, and i think it is moving in a good direction at least from my perspective, but until it chooses a path and does so clearly, all of this other stuff is still going to be left unresolved and should be

resolved. >> so i think the logical order of resolution is the one you suggested which we ought to figure out what do we need to do about safety first, then figure out the spectrum consequences. i completely agree with that. one of the issues with cellular v to x, it is still early days. there was recent announcement by ford and qualcomm and some other players that they were going to set up a test bed in california. i certainly hope they share data off that so people can begin to see is this technology living up to the marketing hype we have been hearing about for six months or a year. that would be helpful. and i think it would be important for folks to start telling us what they think the business model looks like so that could be compared. we don't have the information so we don't have clarity around those two issues. i think that's the right logical intellectual pursuit.

[ inaudible ] >> i certainly hear that concern. this is a challenge that federal communications commission faces often in this environment because they're in position of needing to designate band width, allocate band width many years before technology takes off. shouldn't take 20 years, but certainly in order for technology to take off, it needs to have security in place. the fcc needs to make some of these decisions quickly for some of the technologies on the table. >> should add that this is where this got stalled at the fcc, i believe they have been waiting for department of transportation to make at least a fundamental decision about what safety is

needed. it is not so much a choice between dsrc and cellular v to x as the air interface technology because they probably will use about the same spectrum, same amount of spectrum for the purpose of safety, but they really need to decide is there going to be a mandate and what are they mandating. already in europe as i mentioned earlier, they decided 30 megahertz is about as much as would ultimately be needed for the safety side of this. and if they could make a decision like that, then that would allow the fcc to have more confidence in allowing sharing of the other spectrum that might be for nonreal time safety

applications. >> certainly a lot of questions, speaking of which, any more from the audience? go ahead. >> time for one more. >> on the safety angle, what do insurance companies think of all this? do they see this as something that's a big deal for them or do they not see it as a big deal at all? >> the insurance companies are still trying to figure out whether blind spot protection or lane departure warning are reducing claims. they may have just recently come to the conclusion that that is the case. the insurance company, the insurance industry moves very slowly. they focus on history. they don't forecast efficacy of different safety technology. so as far as tesla was concerned, you weren't paying a

fortune because they were taking risks because underwriting was based on historical data. insurance industry will be very careful and slow to react. >> yeah. i don't actually think the insurance tri will play in this debate. you see car manufacturers competing on safety, volvo's tame us goal, get to zero fatalities by whatever year it was, and they're all trying to move in that direction. so i think the auto industry itself is trying in various ways in good faith to try to eliminate fatalities, reduce accidents as a whole by whatever technology means they think is best. >> i heard insurers express interest about this, it is related to how this may eventually interact with automated technology. insurers have been very engaged

on that because not only does it have far more dramatic implications for crash numbers when compared to hazard warnings. it has potential to upend business models shifting liability toward manufacturers and away from end user if the end user is no longer directing the vehicle in traffic. where they talk about this, they're concerned about cyber security aspects but as relates to automated technology, they're less interested in it as a hazard. >> i think we have time for one more quick one. ma'am? >> lauren smith from future of privacy forum. one of the early slides, roger, you had a section that talked about privacy protections that

could come into play that wouldn't be in play for dsrc. heard it discussed often, it was built to be privacy protective but creates risk. why is it that cellular technologies or 5g would be more privacy protective? >> i was going to say dsrc community prided itself on privacy protection. it is an anonymous message, they've had challenges to see if people could find and identify a car from provided data. i believe my sense is they would be equivalent, except with cellular you would have more consumer control. either opt in or out because it would be more wireless service based opposed to mandated protocol that's announcing your

location ten times a second. again, my understanding of the ultimate roll out of dsrc would be completely anonymous and that's my understanding anyway. presumably it should be, you could argue it would be superior to cellular in terms of preservation of privacy. but that would only be in the context of cellular would have presumably an opt in or opt out proposition. but for safety, i suspect it would be the same protocol as dsrc. i will have to check the slide again. shouldn't show a difference i don't think. >> thank you all for a great conversation. and thank you guys for joining us.

[ applause ] thursday morning we visit salem oregon for the next stop on the 50 capitals tour. governor kate brown our guest on the bus during washington journal starting 9:30 a.m. eastern. now a hearing on space and national security. members of the house armed services subcommittee heard testimony from general john raymond, commander of the air force base command and other defense department officials.