screeria -- nigeria. -

hearing. manhattan institute hosted a discussion about education and the future of the african-american community. today the brookings institution hosted a discussion about driverless vehicles. engineers, legal analysts and state department officials participated. this is an hour and a half. >> good morning, everyone.

the u.s. department of state and it is my pleasure to welcome all of you to the discussion on bringing driverless cars from markets and introduce you to the the deputy assistant secretary of state for science, space and health, doctor jonathan margolis. as a member of the service, dr. margolis oversees policy programs in the areas of international science technology, international health and body of defense and space and advanced technology and also conducted courses at the university of maryland, american university and the foreign service institute on these dr. margolis began his career as anti-american association of incident of science policy fellow and worked in the numerous bureaus in the positions of the intersection of science and diplomacy and with that i will turn the floor over to doctor margolis. >> good morning, everyone. my name is jonathan margolis and

i do work at the department of state as the deputy the deputy assistant secretary for science, space and health. it's my pleasure to be here today at the brookings institution and i would like to thank brookings for putting together this distinguished panel to talk to us today about autonomous vehicles. let me also thank my colleagues from the embassy pf germany here in washington for their work in bringing together the panelists today. the topic for today is one that is near and dear to the obama administration. i think most of you know that they placed a critical role on science technology and innovation. it underpins many societal goals whether security, economic goals, environmental, health, its epicenter of them. and today science is global. many countries around the world are increasing their investments

anticipate that's true in oecd countries. and the countries with aspirations to develop further further and bolder economies and in the united states we are facing a situation where we have to start thinking about how scientists can engage not only domestically but globally so they can get access to the best minds that may be outside of the borders and institutions which are here. in the state department we have a concept we call science diplomacy. we think about using science in the diplomatic goals and diplomacy to advance scientific goals. this is very much a goal to use science to address some of the global challenges that may exist and the answers may come through international collaboration. this is part of what secretary kerry referred to as the shared prosperity agenda way of raising economic growth in countries

around the world. in my part of the state department where i state department where i work in the bureau of international environmental and scientific affairs, we maintain over 50 bilateral science and technology cooperation relationships with other countries and the goal is to advance the science and technology as a part of the broad agenda of the state department might have for building relationships overseas. one of those countries is germany and we are here today in large part because two days ago, experts in the united states and germany met to discuss the cooperation as part of what we call the u.s./germany joint committee on science technology cooperation. that's one of the ways the advanced science diplomacy and in a meeting i described the identified areas of joint cooperation, one of them was in mobility and so the topic we are discussing today is on the minds of those of us engaged in the

cooperation between the united states and germany. emobility is a central piece of the innovation strategy and the discussions that took place earlier in the week between the two governments discussions that will take place today fits squarely with president obama's prioritization of advanced vehicle technology as part of what will be the soon-to-be released national strategy for innovation. this technology i think you all know in the interest here today could be a transformative technology. experts talked about this in terms of not only the technology spinoffs but also reducing traffic fatalities and it says so here i'm not sure, but the panel will discuss that i'm sure about 2050. doing so will require a major investment not only in r&d but other issues. in the obama administration

budget for 2016 it doubles the request for investment in autonomous vehicles and proposes a number of pilot programs to prepare the roadways for safe introduction of the vehicles. clearly that is a u.s. priority. but if you think about this in the international context, the technologies involved the applications will go far beyond the borders so we need to cooperate with others at every stage in the development process particularly those countries that have major automobile exporting capabilities and then manufacturing capabilities. the major manufacturing companies in countries such as germany, japan, united states, others as well where the large manufacturers in the world need to cooperate on the development of the technologies and also the fusion of the technologies in the environmental health and safety regulations that are going to be part of making this technology as transformative as it can be and this brings me to the next point and that is the enabling environment that will

be necessary to create to make the technology work. we need to work together internationally to make sure the new market takes advantage of the best approaches to innovation technology, safety issues and property rights protection, a whole range of things to make this possible. the individuals on the panel today are in very good hands and i have the privilege to be with the group at the beginning about what we were discussing politics that are covering and your in for a real treat of folks that we have cover a wide range of capabilities not only from the development of the technology and the vision that we are trying to get to using those caught implications for society so they can apply outside of the government context and ultimately the international implications of how the mechanisms might be necessary. so with that is kind of a preview of what we are going to

do and let me say again how delighted i am to be here to express the state department's joy at co-sponsoring this event and i will turn the floor over to ben to moderate the session. [ applause ] >> thank you all for coming out on a beautiful morning when we were normally be outside to talk about driverless vehicles. i want to do this in a discussion oriented way so the panelists all agreed to dispense with opening presentation statements and we are going to go right into a conversation. i am going to introduce the individual panelists as i direct initial questions to them so as to not gum up your time with lovely all available information

into handouts that were given to their full bios. we will take audience questions and so as you have a question, please do wait for a microphone because we are being recorded on broadcast and please when you speak, introduce yourself and say who you are in the form that would be recognizably a question so to start with the manager and outreach innovation policies, i want to start by defining the terms a little bit. we all say the words driverless carvers as though we are all talking about the same thing and as though there is a generally

recognized definition. this morning i was driving in a non-driverless car my son said where he needed to be this morning and i have to explain to him with what a driverless car was and actually it's hard because certain levels have been with us for long time, anti-lock brakes, cruise control, various features and yet the phrase driverless car connotes a sort of total autonomy and so my question is what the heck is a driverless car, and what is it in practical terms today and what are we aspiring for it to be in the future? >> that's a heavy first question, but before i answer i want to thank brookings for tackling this issue and the other panelists for being here because this is definitely an

issue that will only grow over the next several years and a couple of decades and i think it is much bigger than people recognize and i hope some of the remarks will get that point across a little bit on the international context. but to answer the question or to attempt to come a driverless car maybe a little bit of a misnomer. what we are talking about is an automated vehicle and there are different levels of automation. some of which as you said are here that are here today in cars especially in some new cars. we are looking at currently in some of our mercedes-benz vehicles technologies like cruise control, but adaptive cruise control, active brake assistance, these things that can take over if you are having an emergency as a driver in the

car that can brake for you to avoid an accident, things like that. so looking forward, taking that as a current situation coming looking forward to '25, 2030. we are looking at a level of automation on the scale it could be a fully automated vehicle by that timeframe which would mean that there still would be a driver but the definition of driver would also change just as the definition of a car has changed in the past 50 years and 100 years and will continue to change as we move forward so we are looking at a driver that is more of a manager, vehicle manager and on the heavy-duty side we are looking at a logistics manager rather than a truck driver and that role will be for monitoring and being able to step in if there is some sort of a glitch or problem in the system. but the way that we see this moving forward for the autonomous vehicles in the future is kind of on the passenger side of the mobile lounge that will allow passengers and drivers to have a

more relaxed environment as they are getting from home to work or day care or wherever you need to be but still be able to take over on a saturday morning when you really want to drive on a nice country road. on the heavy duty side also as many of you know we produce freight liner trucks etc. and just this week we released our free play and inspiration which is a fully autonomous 18 wheeler heavy-duty truck which has a license now to drive in nevada with a logistics manager behind the wheel for testing purposes and there are two of those on the road now. so this has implications but looking at the logistics that can have huge implications as well. so i hope that answers the question of this progression of autonomy as the years go by starting with where we are today.

>> it does beautifully and i want to follow-up on the fully autonomous truck because i suspect a lot of people in the audience walk in here without a sense that there are fully autonomous vehicles driving around nevada. did you know that? because i didn't when i walked into brookings this morning. so tell us about the truck. what is a fully autonomous 18 wheeled vehicle and how autonomous is fully autonomous and what happens -- what does it do? >> i don't want to go to into the details but i will say for those of you that studied logistics management, freight, these kind of issues, this has the implications of something like this for the trucking industry and freight management in the united states and other

areas is huge. the truck will have a driver not called a driver that will be more attuned to where other trucks are, where they need to go, what the fuel economy is looking like. certainly in the future the powertrains for the vehicles are looking at fully electric or hydrogen so this has implications as well. so the future we see this certainly has a large part due to platooning and this automation will definitely help with that. the fuel economy, as i said, it is going to be greatly impacted by these kind of technologies such as the cruise control and they can travel exactly the same speed regulated not by a

person's foot on the pedal but rather the computer within the truck that has enormous implications. so that kind of a system moving forward i think is going to change the trucking industry and the jobs of what it means to be a truck driver and what it means to be a manager and how we get goods in this country from place to place especially because trucking accounts for a very large percentage of the gas usage in this country. mpg is much lower on a truck and on a passenger car. and it's also contributing on the road. all of these issues can be impacted by a level of autonomy moving forward. >> carlos is the head of reliable automation control the

corporate technology fields. i'm interested in your sense of the research and development necessary to accomplish the vision that we just heard. what are the missing pieces and what do we need to develop that we haven't developed and in the spirit of the sort of international side of this conversation, how much of that is u.s. technology and how much of it is not u.s. technology and what is the interaction between the two? >> first of all, thank you for the invitation to this interesting panel discussion and the road to automation is more

and more automated systems in our daily lives. so we have it in a military area already somehow and the next step on the road. so the car industry is recently the driver for the technologies and other areas as well and i want to address for the research first of all the safety and accessibility of the system and when is the system safe enough to some figures like the probability for failure before automated driving in the future

to more or ten minus, 11 or 12. but nevertheless there is a probability for failure. every system can fail and we have to deal with that but on the other hand, the safety of the system. how are we improving that. the calculation very simply shows 1 billion kilometers on the road is not enough for drivers to show the system is fulfilling these conditions. so we have to find new ways for testing and this cannot be done on a national level it has to be done on an international level. so the way that we come to that is modeling -- so it isn't the way that we did in the past just driving several miles and stop. the paradigm changes.

the second point we may not see it in the next few years. it will take time. we have a way for automation and more functionality which will help the driver avoid accidents and it would increase so when we accept that the driver is a backup for the system, how hard is the driver coming back, what is happening to the driver during that time and is he still able to take over the system and how do we provide the driver if there is a possible flaw so there is an open question and the answer is ten seconds up to 15 minutes.

you see how big the variance and we have to find the right way how we can solve this problem. the second point is security and privacy. as soon as the car is driving automated on the road come manufacturers and researchers expect this is done with a lot of environmental information. but the data could be corrupt. who is libel at the end and how do we guarantee and who is liable at the end if it isn't correct and how are we dealing with such problems? the fourth point is the social and ethical impact we have. it's also a research arena. so what is happening if a machine is causing an accident?

drivers are causing accidents. this is a big difference. is it different around the globe? when this is happening that the machine has a choice between two bad situations every situation is causing accidents but it's making a choice and it can be followed up because it is programmed. at the same could be done by humans that's different. and forth, one of the capabilities of the autonomous systems is that they are learning so they are not systems that have a fixed state and are delivered to the customer at the end, it will change over a lifetime and this is still happening when we have the automated driving because assuming this car is using -- it would change over time so there

is a learning system. it might be in the car or outside of the car but the system would never remain the same as it was in the starting point and how are we dealing with that and guaranteeing that the systems are still safe? >> we are going to return to those momentarily because i have a million questions about that. but i want to hear from the fellow at the new america foundation and the author of the book about electric cars and so this is an interesting element that isn't the first time we have had or tried major transitions from one form of technology to another. some have been more successful

than others and some have been promised for long periods of time without materializing in a way that we would imagine. one of the lessons for the earlier transition in the current set of the transitions and specifically when you think about the experience and how they make it electric rather than internal combustion engine based, how much does that make you optimistic or pessimistic about grand promises of removing human agency or minimizing human agency from the driving in the first place? >> thank you for having me here and i'm excited to be a part of being that this discussion. before i get right into the question i want to talk about

what i think about when i think autonomous vehicles such as my experience of growing up driving around in a car. i had a ten brothers and sisters and as you could imagine that is quite a brood to manage and so my mom would be sitting in the front driving around england rearview mirror trying to look at us and yell at us while she was in the front of the car and managing all those things at the same time. so i can imagine an autonomous system being somewhat more safe than the alternative. [laughter] i would also say happy mother's day mom. you kept us safe all those years and we appreciate it. getting back to the book and about lessons from electric and autonomous vehicles i think the first thing i would say is these massive societal transitions don't just happen. they are driven by policy and it's important to remember that. and my book is an international

examination of what policies are effective and what policies are now and how they can apply strategically over the course of decades they primarily look at china, japan and the united states which are the largest producers in the world cumulatively they produce produced more than 50% of automobiles in the global economy. and it comes to kind of an interesting conclusion which is that while all three of these huge economies and the governments that manage them or are regulating them or pushing for electrification during the period stretching from 2007 to the present day there's another factor that happened to have a much bigger influence on electrification and ended up being about technology driver for the entire industry and that was the state of california and

the reason was the state of california has an institution called the california resource board that made it a priority to drive automakers towards electrification and they did this over the course of decades and they used a specific set of market-based tools to make that happen and what they did is they told automakers if you want to sell cars in california, then you have to make a percentage of them electric which is a mandate. that is kind of draconian and expensive and inefficient. what if they don't have particular expertise? that could force them to develop a whole new set of technical capabilities that maybe they don't want to invest in that moment. so to make the system much more

efficient they overlaid a market and allowed the automakers to buy and sell credit they were awarded when they sold an electric car so then what you have is a market driven mandate and that resulted in something much more efficient than if you have the mandate behind the electrification of programs. so this is applicable to the concept of autonomous vehicles that they are also different from electric vehicles. leng be electric vehicles are great. everybody loves the tesla model as it is maybe the best in the world at this point in time, but the truth is what we are aiming towards is a whole set of social goods related to its climate, energy security and other things like that. as you can talk from my introduction i can see many reasons why an individual might want to have an autonomous

vehicle so they are going to be much more powerful market drivers for automation than there were for electrification. nonetheless, that doesn't mean we don't have to think strategically about the regulatory aspects and industrial policies undermining the transition from a standard vehicle we all drive ourselves with and so i think we should take the lessons from california and realized there's a small corner of the global economy of the act strategically and apply policies over a long period of time can end up setting the agenda and that's why forums like this and bringing together united states and germany to talk about the global cooperation on standards and on a road map for

autonomyout going forward is so critically important. >> last last, sonya smith is the professor of critical engineering at howard university and i want to ask you about this international cooperation aspect jonathan margolis alluded to it or talked about it very directly in the introductory remarks and several of the co-panelists have alluded to that as well. when we think about auto development historically, we don't think of sort of international cooperation. we think of detroit or places in japan or germany but these kind of local engineering parts where the cars are built in the regional terms. why is this different and what is the consequence of having a significant global cooperation dimension to this or being in a

negative sense not having it? why does the global side of it matter? >> thank you and i would echo my panelists in thanking you for asking me to be here. i think as the educator on the panel i need to emphasize that part of the answers to these questions be it research and driving the technology further has to do with incorporating students and faculty in global collaborations. we are one of the institutions of the partnership in the advancement of collaborative engineering education and it's is a partnership among general motors and others to catalyze projects in education and so to answer these questions it's important to not only involve researchers at the companies and faculties but also students as

well. one of the things we do is collaborate on the global design competitions and this involves teams of students and faculty from across the globe. the team at howard university is partnering with one in germany and the university in sao paulo and toronto and new mexico state and we are on a global design project to answer these kind of questions. we exchange interns. we have an intern here. it is a very rich experience and the students are the ones that are going to be the early adopters and the drivers of the technologies. to get back to the question of why it's important to have a global perspective as opposed to the original engineering aspect

when we talk about autonomous vehicles, the subject spans all disciplines. it not only spans policy it spans engineering and creative design and atmospheric sciences. when you get all of these disciplines together in the educational arena, it definitely enhances the educational experience for not only the students here in the united states but also globally. >> i want to push you a little bit on why is this more true in the area of autonomous vehicles than other vehicles or is it just the scope of the research necessary to do these things is so vast?

why is it we see these international collaborations companies and students in this area but we didn't see it -- maybe we did but we didn't talk about it when it was earlier generations of cars? >> i think there were different types of collaborations early on that as we move, as the vehicles become more complex, the issues become more complex you need a collaborative and a global approach to solve the problem. it's not just from the mechanical engineering and software from computer science. you also need a policy involved in that and it's extremely important so we need to reach out to colleagues and disciplines that we might not otherwise incorporate in a design or engineering type of process. >> i want to talk about safety. a bunch of people praised safety

and we started with the possibility of the promise of the zero accident fatality road systems. that sounds totally fantastical except that in defense of this completely fantastical hypothesis i want to point out that this year it is safer to fly in the united states than not to fly in the united states. all kinds of accidents happen in your home, you are actually safer on a commercial airplane than you are not traveling. [laughter] >> the more interesting it gets. so the possibility of truly radical de-escalation of the violence associated with roads you don't have to get to zero before you get to something

that's really attractive and a policy matter. on the other hand, as we heard, the deaths that do happen will happen because of machines. and as we see in the military context, when machines cause human death as opposed to people, we get really uncomfortable. so i want to float this to the panel in general if you have a regime, technologically and policies in which many fewer people die but they are caused without human agency and by programming decisions made remotely or system failures, is that a win or something that we will have great social difficulty accepting? >> i will try to tackle that first.

this is another example of one of the issues that need to be discussed across the country borders. this is something that is cultural which as we know is different in different cultures around the world as well as the technology. and this is something that the auto manufacturers are looking at right now and we haven't had as much before on other technical issues on this kind of technology. as someone mentioned before, the issue of the inevitable crash and what the computer programming should look like for something like that and the liability in the situation like that. the ethics how to program these kind of technologys is tantamount, paramount. so this is something we are going to be looking at already

this year. one of the board members has convened a group of academia as well as technical experts and representatives from the eu and the german government to come together to discuss the ethics of this every couple of months moving forward because it's going to have big implications. >> when i drive down the street -- and this has ever happened to me but you think about it, you are in a situation where there is no good option. somebody's is no good option. somebody's going to get hurt no matter what you do and you make a split-second decision and there is the policy behind the decision but it amounts to i'm going to kill that person, not that person or i'm going to risk that person's life to protect my own life -- there is no global policy about that. but when you replace my mind with a computer you have to have a policy about that and it's

called a code and it amounts to a policy level decision on the part of the programming to kill this person instead of that person. what are the ethics of that? >> the difference if we are able to reduce the number of accidents on the road first of all that is a big advantage. we have to admit that as well. the second thing is as soon as the machine is causing an accident with full transparency we know what is behind it. at least you can stop the process. you have a split-second decision and who is really able to judge in the split-second what is the

best choice we have here and the program made a choice based on the system and as long as the process is transparent you know what is happening. so now we have to discuss what kind of rules are behind that. we have to do it somehow, somewhere. what is better if you have the choice, a or b because of an accident. that is something we have to go on and then make this decision. >> it's a very important issue and has to be addressed. it's an issue that lends itself to a lot of washington pending in a hemming and hawing. [laughter] >> we need to think about this really carefully about 30,000 people die every year in auto

accidents. my father died in an auto accident. his father died in an auto accident. it affects us all very personally and if we have a process of potentially eliminating 90% of those that is a win-win and backs the personal example my father died because of a mechanical belt jim. there was nothing that he could do about it. we were in the mountains in he was driving and there was a chemical malfunction and he gone off the side of the road. his father died because of human error, someone driving drunk and ran into him. you have the process of eliminating 90%, 95% of accidents or fatalities on the road through autonomy i think much more important than we have to get machines to make these decisions. let's save 27,000 lives. >> i agree with you completely

and i am very adamant about reducing the deaths. when i raised the argument in the context of robotic weapon systems, there is a mobilized constituency of the human rights activists who take the view that if you can have a dramatically greater compliance with the laws of the war as a result of the weapon systems, they object to this because you have removed human agency from the task of killing. and so i wonder if those of us that believe that actually aggregate debt reduction is a great good that should trump a lot of other goods are going to

have a tough hill to climb in terms of social acceptance of the idea that removing people from the chain of command that leads to death is somehow presumptively suspect; what do you think? >> again, you have to step back. on a policy perspective and the logical perspective, it seems like they are apples and apples. i think from a human perspective they look like apples and oranges. you are talking about robots that put people and things we are talking about here today are robots with transport people and try to keep them safe. i understand why there would be a constituency that would mobilize against robots to kill people. there will be those opposed to the autonomous vehicle but at the end of the day we have to meet some social decisions and i

think the weight of the public opinion is going to be on the side of the safety. >> do you finds, sonia in these educational and development partnerships, they use the safety and ethics issues are kind of front and center of what people are working on or are they kind of background issues that are concerns but mostly there's just a lot of excitement about the technology? >> there is a lot of excitement about the technology but i also look at it from a different perspective saving lives isn't only through the autonomy and going to be achieved in reducing death and auto accidents. autonomy is also going to affect mobility and access to other services that will save lives, health care is one of them. so i think that in the

aggregate, autonomy will not be looked at in the same lens as just saving lives through the automobile accidents. those are the issues we look at through these partnerships. >> so what is an example of, you know, if there isn't merely reduce traffic accidents, we've heard about the environmental benefits with respect to fuel efficiency, but what is an example of the great social goods and reduced human suffering as a result of autonomy that is not fewer traffic accidents? >> we don't have specific evidence that an example that has been talked about is access to healthcare in rural communities and people who are

disabled. they may be able to get access to the healthcare services much quicker through and off on his vehicle van waiting for their friend or having to walk or get to a service so that is one life-saving result, not a perfect design point that it definitely is an interactive benefit in the autonomy and i think again if you look at not only the lives saved through reduced auto accidents but also lives saved in personal mobility and access, i think the autonomous argument is quite different from the drone strike. >> i would like to add one thing to that which is a couple of years ago i was at the doctor's office and when i was coming out there was a gentleman on an electric scooter, severely disabled and we started talking with him and is that you have a cool electric car and he said

yes but what we are really looking forward to is autonomous vehicles because there are so many things they can allow us to do that we can't do on our own right now so that is a really compelling argument. >> one issue that autonomous vehicles raise is that data collection. these are incredibly sophisticated sets of many sensors, all of them collecting a lot of data and processing a lot of data in order to make decisions. the data collection gets controversial and has different cultural baggage and different countries and different degrees and this strikes me as an area where the harmony might be quite difficult particularly between germany and the united states

such as many of you know fight about the data a lot not just in the nsa context but also principally in the business context. so now you have 15 years from now a giant fleet of the big data machines roaming around every city. what are the prospects for the regulatory harmony on the subject across the borders and what are the prospects for the cultural queasiness associated with the idea of our cars as personal surveillance? >> first we have to differentiate a cluster of machines running around the country. they are collecting a lot of data about what is happening

with this data? is the car relying autonomous but also drive -- there's a separation from outside because everything that is generating can forget immediately after the privacy issue. so they are not quite clear. what is the right path for the future? the fraction thinking we need a lot of data from around the car, from city centers and so on they then have immediately the problem of what the privacy and security of this data. data from the car can corrupt the car and corrupt the driving situation. again safety issues.

secondly is sending data outside of itself, whose the owner of the vehicle, who can use that vehicle. is it the driver, is it the car owner, is it society? these are all the questions we have to discuss and this is a cultural issue around the globe not in every country the same. one thing is for sure. as soon as data is available, it will be used and it will be used for business. it was always the same story in the world. it will happen here again. so as soon as data is outside of the car available, it will be used. that's for clear. >> i want to underline that. i think this is a perfect time to explain why this kind of collaboration between governments, between companies, and between legislative bodies and corporations makes so much

sense. it is really important. of course informed by academic bodies as well. we do a lot of research have universities around the world and have for decades. obviously most companies do interface with government at different levels but regulatory bodies diplomatic missions executive, et cetera. so these kinds of collaborations are necessary because we need to make sure that moving forward technology is advancing as fast as societal adoption society's willingness to accept this kind of technology, and at the same parallel speed as laws and regulations because we want to make sure they're harmonized. we have an opportunity with this technology to do so. obviously cars have been around a really long time. invented the car almost 130 years ago. at that time there was no need at all for germany and the u.s. to be talking about safety

regulations in automobiles. now there is overwhelming need for all the countries to be discussing this together. so we have new technology and it needs to be discussed and we have opportunity to do that across country lines as informed by experts with the technology and privacy, data experts, something we can't overlook. as the process moves forward, there's no choice but to work together. >> and i would say autonomous vehicles are one small facet, but significant facet of the changing relationship between data and vehicles. we are moving towards a world of connected vehicles where our cars will have huge amounts of media and internet connectivity, an era where you're going to have connections between vehicles and infrastructure in the city to tell the car how

fast to go what kind of traffic situation is going. that's a new stream of data. cars will be connected to each other, vehicle to vehicle technology. that's something that's under development and we have whole regulatory structures built around it as well. finally, you have the uppermost tier of automated vehicles. so i think we have to think carefully about data and transportation and how we're going to manage that merging relationship. >> all right. so several of you have alluded to as distinct from safety issues what we might call cyber security issues with respect to you have all of this data you have the systems that are autonomous, both the data streams and systems themselves that are operating vehicles like any complex computer system are subject to attack. as we have seen or in the case

of data corruption. as we have seen in other areas that we have systematically networked and become very dependent on these become very attractive targets for hackers from the lowest grade to the most state sponsored. is this a situation where we are now creating dependencies that we will eventually turn around and say oh my god, how did we come to give the north koreans control over our traffic safety? just to use an example i'm sure has nothing to do with the news. i mean, why shouldn't we be worried? how worried should we be about cyber security implications of having basically our cars as you know the operations of our

cars as network instruments, both in their own internal systems and data that they're collecting and relying on for safe operation. >> from my perspective, i actually am not so concerned about this. i think it definitely needs to be done right and this is another area where we need to be talking between private and public sectors across country borders to really effectively regulate these issues, make sure we are thinking of them in the right way. i know in our company we have basically internal hackers who are tasked with trying to hack our vehicles all of the time. it is amazing, as i said before some systems in passenger cars today are already very advanced and already maybe starting to inch up on the scale of automation such as these lane

assist and adaptive cruise control, braking i mentioned, et cetera. these kinds of issues, even in some of our cars we have systems which will notice if you're getting sleepy not by looking at you by how you're sitting and driving compared to when you were driving when you first started the car and will flash a coffee cup on the dash to tell you why don't you take a break now. those kinds of systems are collecting data on you as a driver and the way that that data is stored is certainly cause for debate. but we think that i can only speak to our company, but we are -- because we are based in germany, we are probably the most stringent -- subject to the most stringent data laws. we are confident that moving forward that's the view we'll

take. >> talking about this data privacy and security again maybe we have to differentiate between data which are necessary to drive autonomously and not all of the data in the car. we have a lot more data as was mentioned already. so this kind of entertainment data is just a free flow of data, and every user can decide what he wants to do with that. this is actually according to some agreements between the provider and user. and data we have in the car to drive the car somehow autonomous in the future it happened already in the past. so that's not a new issue. we know that already. the research arena it is known. it will happen. this is for sure. to make first of all the right choice, what is really necessary

for the automation of the car and small amount of data is, it is more difficult to tackle. if you need everything when we get around from the car, every camera system in the city then it can be corrupted easily. if it is just a small amount of data or no data that the car is collecting everything by himself, keeping it for itself, then we have not such a big issue. as i said cars were corrupt in the past anyway. you can pack the car and put wrong data in it. in the future, we will see with autonomous, we will see a different kind of car in the future. quite sure you can load your own apps down to the car you have your special driving functions and so on. this will be the future of the cars, not just this autonomous driving. this is just one of the functionalities.

so as soon as we bring new software into the car, it can corrupt the car another way. this we have to protect. we have to be very precise. it is not all data it is just a few. and fewer this amount is better we can protect it. nevertheless, you have to have the public discussion, what will happen if if something is corrupt. >> for sure. this is why we see so many redundancyies redundancies. >> and not just cars so many facets of the economy where this is a huge issue the energy system, i came from a two year appointment to department of energy, we think of this all the time, the systems are connected to networks. if you have someone break through the control system of a nuclear power plant or more traditional coal or gas-fired power plant or into the grid that can cause some pretty huge problems. so this is definitely a big

issue and i think it is an economy wide issue. >> i just read a report about commercial aviation how secure aviation systems are. the answer seemed to be that they're quite secure, as long as they don't interface with the public, the public facing computer systems on airplanes that are not secure at all. the trouble is we have now inter laced them. people like to watch flight data on their computer screens, so the two systems now interact in ways that create vulnerabilities for the other systems. let us go to audience questions. if you have a question, flag me and please wait for the microphone. we have a lot of questions. the gentleman over here i saw first. please introduce yourself and please keep questions brief. >> thank you.