Proceedings of the Standing Senate Committee on
Transport and Communications
Issue No. 16 - Evidence - May 9, 2017
OTTAWA, Tuesday, May 9, 2017
The Standing Senate Committee on Transport and Communications met this day at 9:30 a.m. to continue its study on the regulatory and technical issues related to the deployment of connected and automated vehicles.
Senator Dennis Dawson (Chair) in the chair.
[Translation]
The Chair: Honourable senators, welcome to the Standing Senate Committee on Transport and Communications. This morning, we are continuing our study on the regulatory and technical issues related to the deployment of connected and automated vehicles.
[English]
Appearing before us today are representatives of two consumer groups. From the Automobile Protection Association we have George Iny, Director of the Head Office in Montreal. From the Canadian Automobile Association we have Ian Jack, Managing Director of Communications and Government Relations; and Jason Kerr, Director of Government Relations.
[Translation]
Welcome. I know that you are used to being here. You appeared last year for the study on Bill S-2. You enjoyed it so much that you wanted to come back. Mr. Jack, the floor is yours.
Ian Jack, Managing Director, Communications and Government Relations, Public Affairs, Canadian Automobile Association: Mr. Chair, thank you for inviting us to appear today to talk about our concerns and Canadians' views on connected and automated vehicles.
[English]
Most of you will be familiar with our brand. Founded in 1913, CAA today is a non-profit federation of nine clubs providing more than 6.2 million members coast to coast with emergency roadside services, automotive insurance, rewards and travel services. CAA has also, since the beginning, advocated on issues of concern to its members. Today those issues include road safety, the environment, mobility, infrastructure and consumer protection.
Approximately one in four adult drivers in every province is a CAA member, and today our comments will provide an important perspective, that of drivers, as it relates to connected and autonomous vehicles.
[Translation]
It's important to study connected and autonomous vehicles together, since the development of both is so interrelated, but it is also important to keep in mind that they are different things. When we refer to 'connected cars' — these vehicles have already been developed and are spreading. And virtually all vehicles produced today include some form of what we can call automation, ranging from cruise control to parking assist. Fully autonomous vehicles, however, may take anywhere from five to 30 years to become prevalent.
[English]
Similarly, you would be remiss not to examine the future of electric vehicles in your study, but again, EV is not a synonym for AV, nor for connected vehicles. In the future we are likely to be driving autonomous, connected vehicles powered by a non-internal combustion engine. But each of those developments, while linked, will proceed on its own timeline and presents it own opportunities and challenges for consumers, companies and policy-makers.
When it comes to the connected cars we have on the road today, they are already amassing volumes of data. Many in the technology sector now believe a single vehicle will soon generate 1 gigabyte of data per second. In May 2015, Tesla reported that it had collected 780 million miles of driving data in the previous 18 months and that it now adds another 1 million miles every 10 hours.
[Translation]
At CAA, we believe one of the pressing, current issues in the area of connected and automated vehicles relates to transparency and consumer control over vehicle data.
[English]
In a February 2017 CAA national poll, 49 per cent of Canadians said they were not aware of the range of data being collected by their vehicle, and when it comes to the sharing of that vehicle data, nearly 90 per cent of Canadians agreed that the consumer should decide who gets access to their vehicle data.
Interestingly, the same poll also found that nearly 70 per cent of Canadians think that in-car data technology is very or somewhat beneficial, while a strong majority are also concerned about it. And yes, those last two numbers add up to more than 100 per cent. The explanation is that a sizeable number of Canadians are able to see both the benefits and the risks of in-car technology. Those who express concerns are not just Luddites. They include people who enjoy adopting new technologies.
[Translation]
Overall, the poll found that 81 per cent of Canadians feel there is a need for clear, enforced rules to protect the privacy of their personal information when it comes to vehicle data.
[English]
CAA has long held that vehicle owners should be informed about what data is being collected and be able, within reasonable limits, to choose with whom they share it. It must not be a take-it-or-leave-it approach where in order to enjoy the benefits of in-car technology the owner must abandon all rights to privacy. Only your phone, after all, will know more about you in this brave new world than your vehicle does.
In this context we note with approval Privacy Commissioner Therrien's comment before this committee last month that his office is interested in the development of a code of practice as it relates to privacy for connected cars and his call for federal leadership that is proactive rather than reacting on a complaint basis when it comes to this issue. However, we should not confuse the funding of a third-party report into this issue, or a commitment to leadership, with the idea that the consumer interest has already been taken care of.
We would also note that a bill currently before the U.S. Senate titled the "Security and Privacy in Your Car Act'' would direct the U.S. administration to establish federal standards to secure American cars and protect drivers' privacy. This would include a privacy rating system for consumers; clear and conspicuous notice to the driver about what data is being collected, if it's being transmitted or saved and how it's being used; and the ability for a driver to opt out of data collection without interfering with their ability to use navigation tools. It's an interesting piece of legislation.
[Translation]
I'm conscious both of time and the fact, honourable senators, that your brief is wider than just data privacy, so let me close with a little more on the autonomous vehicle.
[English]
When asked in our polling about the benefits that would come from AVs, only half of respondents cited having more time for leisure activities during their commute as their number one interest in this new technology, and this runs counter to what many in the field feel is a primary benefit. The top benefits cited were improved accessibility for people with mobility issues and addressing driver fatigue.
[Translation]
There is no doubt that autonomous vehicles will be a boon for road safety. CAA supports the responsible development of autonomous vehicles and believes connected cars are already here and will soon become ubiquitous. And we salute the timing of this study, since it is now, and not after the fact, that policymakers should consider the federal role and the ramifications for citizens of these developments.
[English]
There are many questions to be considered in addition to privacy, from consumer acceptance, to readying consumer infrastructure at all levels, to moral and ethical concerns, to implications on the insurance industry and, in due course, solving how autonomous vehicles will coexist with so-called dumb cars on shared roadways.
We thank you for the invitation to speak today and look forward to your questions.
[Translation]
George Iny, Director, Head Office, Automobile Protection Association: Mr. Chair, the Automobile Protection Association (APA) is a non-profit public interest organization with offices in Montreal and Toronto. We serve all of Canada. The association was founded in 1969 by two Americans who came to settle in Canada and were previously working for Ralph Nader's team, in the United States, which was one part of the associations that emerged at that time of social dissent. There are a number of consumer associations in Quebec, a number of which still exist today.
The APA is dedicated to promoting consumer interests. We support consumers in class actions. We are occasionally a plaintiff; we even have a hearing before the Supreme Court at the end of the month as part of one of our actions. The APA also evaluates the selling practices of businesses through anonymous field investigations often commissioned by the government. In addition, the association has a new vehicle testing program conducted in collaboration with Protégez-Vous magazine, which some of you know.
We are a small association of about a dozen people and we try to improve our presence a little with the help of industry experts who often, very discreetly, join us voluntarily.
We are here today to talk about some issues, because your investigation is very broad. We know that other experts will come to testify in the coming weeks, and we don't necessarily want to address the same issues as them. We will try to complete the picture by identifying some concerns that may have not yet been discussed.
[English]
As Ian said earlier, to a certain extent, many of the vehicles we drive today have a form of limited automated driving. Cruise control is one example. All vehicles by law have today a stability system — although the owner isn't aware of it — that will take over if it senses that the vehicle is heading into the weeds. So the vehicle will actually intervene — the owner may not even be aware — to try and straighten it out. So even if you do run into a pole, for example, because you went off the road, you will most likely hit with the front of the car. Even in that regard, that has contributed to injury reduction.
These devices were developed by old-style car makers who were already making the vehicle itself, and they're designed as add-ons to be integrated with the systems on the vehicle. They weren't considered revolutions. They were considered just new features or new technology. The legal system for those features didn't change. Even though there are cars today where you can push a button and the car will park itself, it did not require a big rethink on insurance coverages, and the driver of the vehicle is still the primary party responsible for the control of the vehicle.
If we're going to go to much more autonomy in vehicles and introduce what appears to be a much more rapid timeline, we looked, and what we see is that predictions are as early as 3 years from now to as far as 30 years from now. It's very hard to predict when you're looking at a disruptive technology exactly when it's going to arrive and exactly at what speed the rollout will happen.
There is no question that if we do have that, then we have to look at the legal framework or protections we have now, and also how much of that could be adapted going forward. To some degree, the future ends up standing firmly on the shoulders of what went before, and if we notice there are gaps, we will be in a position to address them proactively, and that's our purpose here today. My purpose is to bring some of those issues to your attention.
There are also other forms of transportation that do have a high degree of automation and involve a certain amount of vehicle-to-vehicle communication.
The safest transportation device for humans is generally recognized to be the elevator. Of course, the elevator is fully automated today, but it wasn't always the case. Elevators talk to each other on rudimentary systems to ensure that they don't all arrive on the same floor at the same time. It may seem ordinary to walk into an elevator and push a button. That wasn't always the case. Some of you may remember human operators in elevators.
The other area is possibly aviation, where bringing in experts from other disciplines facing some of these same issues might provide a bit of a road map.
In terms of the human driver, we have a poor reputation behind the wheel. If we were starting today and someone came out with a transportation device that could travel upwards of 200 kilometres an hour that you would have full control of, we probably wouldn't license private individuals to use it, if you think about it. Auto collisions are a leading cause of injury and death among the young. Human factors — the numbers vary a little bit from one study to the other — are present in over 95 per cent of collisions, notwithstanding what we think. The experts will say, "Well, even if the roadway was slippery, you were driving too fast for the conditions,'' so there is definitely a human error issue.
What's worrying experts is that despite the improvements in occupant protection — the vehicle has become a form of egg crate; you're protected on all sides — we're not seeing the reductions in fatalities or injuries we had hoped. It would appear after significant declines over the last decade that we have possibly hit a point where those improvements are not coming. The thinking behind that is that we're not paying attention, that there are more things competing for our attention behind the wheel than there used to be. Two-way communication was already installed in vehicles before it was fully studied or perhaps before it was understood how risky it might be. I'm talking about user communication via the cellphone, so that technology presumably, knowing what we know now, would have been designed so that you don't use it in a car, period. There is no significant evidence, for example, that hands-free operation is any safer than holding it in your hand. It's more an issue of attention and divided attention than the loss of one hand on the wheel.
The other concern, of course, is that we're encouraging people to use other modes of transportation, and pedestrians and cyclists are more present on the roads and therefore more present in collisions. An intelligent vehicle that is looking all around itself all the time carries the potential to reduce collisions dramatically.
If we're looking at not a major rethink but simply adding technology to the existing vehicle in which the driver remains in control, the early data that's available suggests that those vehicles will have 30 to 40 per cent fewer collisions. They will look and feel and drive. You may already be driving one if you have a high-end European car. Many vehicles sold in the last year already come with these features. The vehicle will warn you if you are too close to something in front. It may be able to warn you if there is a pedestrian, and it will brake if you don't intervene, or actually tug the vehicle back into the lane if you are not reacting. Those vehicle features have the potential to reduce collisions by 30 or 40 per cent already. That's apparently what the data are showing.
In the world of public health, a 30 per cent reduction is staggering; it's huge. You fight for small increments, so that is a very promising start and fits within the current legal framework.
Federal law requires a car manufacturer to give a notice of defect when they know there is a problem with the vehicle, and they need to inform the government and the owners of the vehicles. Soon, we hope, thanks to the work you did last fall, they might have to do the recall for free. Currently that's what we have now, and it results in almost all cases in a no-charge correction.
It also requires that companies provide the government with the means to retrieve and analyze information created or recorded by the vehicle or its equipment. This is important because vehicles today already record on-board data prior to a crash. The last maybe 30 seconds, sometimes the last couple of minutes, is information that is already registered on board the vehicle, like the speed, whether the brakes were being used, the change of velocity. Companies didn't always share this information with the government even when it was in the vehicle, but it is an obligation.
Both federal and provincial laws require that the person with the care and control of a motor vehicle, whether it is in motion or not, is the primary party responsible for its safe operation. This is predicated on a driver-behind-the-wheel model, not a fully automated vehicle.
To go back to the law, when the law might have applied a similar analogy to this, you would have to go back to before the automobile, probably to the horse and buggy era. Of course, horses were partially automated. The animal didn't always listen to you. Occasionally it was frightened and there were damages caused in those days. There are old cases where it had to be determined who was responsible, for example the owner saying the horse acted spontaneously, autonomously and wasn't in their control; and of course the injured party arguing that the owner of the property was responsible for it. While we're coming up with a framework, we may have to go back and look at how those issues were dealt with before. They may be able to provide guidance as to what works and what doesn't.
Carmakers already own millions of vehicles that we drive every day. Those are, by the way, leasing contracts. In theory, as the owner they could be liable and suffer a loss if the leased item is destroyed. Car manufacturers have figured out a way around that. They require that you maintain adequate insurance. They determine the level of liability you have to hold and the collision coverage minimums. Those are the conditions for having your lease. If you have a self-guided vehicle going forward, the guidance system might have its own form of insurance that the user would be forced to pay for; it's possible.
I'm raising this today only because we have seen arguments to say that the legal framework is either insurmountable or isn't able to handle this. However, there are analogous situations already and there may be other areas where transportation is involved, such as aircraft and elevators, where this has been considered.
Finally, there is privacy. When you buy a new vehicle and look at the back of the owner's manual there is a clear statement saying you own the electronic data recorded by your vehicle prior to a collision and no one can get it unless you permit them to use it or it's under the authority of a court. The difficulty is that is not all the data your vehicle might be collecting. For example, if you subscribe to certain concierge-type services with your vehicle, you might be asked to sign, at the time of delivery, a release or consent form that does allow sharing or use of private information. The only way you're going to get to benefit from that feature is if you sign or check off a box that states, "I accept.'' We have, on the one hand, the industry recognizing explicitly that you own the data on the vehicle. We also have other data that you are saying you're okay to share. The concern going forward will be whether the current protection —
The Chair: Can we go to questions pretty soon?
Mr. Iny: I'll make it quick.
The data needs to be available to government and Transport Canada needs to build capacity or tool up going forward. We believe that a team currently of perhaps three to five people with the appropriate expertise and resources would be helpful to prepare for this. This would include engineers with the required computer coding and other specialties — not necessarily moving deck chairs around to borrow people from other departments but going and get the competencies. That would allow you to investigate when vehicles that are on some form of automated guidance are involved in a collision and to do it using the latest skills and technologies and to participate in research.
Also perhaps form a working group — the government used to have this. It was called national public safety organizations. It was multi-disciplinary and included the police. It could include representative carmakers, certainly the guidance system makers. The idea is that this group might be able to identify emerging issues and help develop strategies for dealing with them before the vehicles are actually put on the road.
The perception right now is that the area is changing too quickly for regulation to be effective and that it would be out of date very quickly. However, there are other tools, either a memorandum of understanding or another device, where you have the important players at the table and the commitment by them to meet certain standards. I think that's very important before the vehicles are put into the hands of private parties.
That's it for me.
[Translation]
Senator Boisvenu: Welcome to our guests, and thank you for your presentations.
Mr. Jack, in the past decade, has the appearance of semi-autonomous cars changed your relationship with clients? We used to wait for CAA when our battery died. With technology, has CAA's role been changing over time?
Mr. Jack: Thank you for the question.
[English]
There is no doubt that there is a long-term business challenge for all of us in the automotive industry figuring out the way forward. In the short term, we haven't seen any impact on the business at CAA. We continue to grow by every year by 100,000 or 200,000, regardless of the economy or developments in the vehicle, so we're very happy about that.
The top two or three reasons people call us are because the battery has died, they've locked their keys in the car or they have had a roadside issue with the tire. Two of those three we can expect over the next five to ten years to largely disappear. I don't think people will have keys in the next five to ten years. That's a simple technological improvement we see in many vehicles today, and we expect batteries to continue to improve.
On the other hand, in this country we will always have — for the next couple of hundred years anyway — winter. As a result, batteries will continue to die and people will continue to have issues.
It is a very good question about the future, not just of ourselves but as well, for instance, of the automotive after- markets, that is, the people who repair vehicles, especially the mom-and-pop shops as opposed to the dealerships. Access to the software to repair vehicles is a challenge. We already see that Tesla does software updates from the factory that alleviate a lot of issues. As vehicles become more and more smartphones on wheels, there will be more and more software solutions as opposed to ball-peen hammer and wrench solutions to issues with vehicles. I think that's a little bit of where the future is.
[Translation]
Senator Boisvenu: Mr. Iny, you raised two points in your brief that caught my attention. The first one is about building Transport Canada's capacity and the second one is about clearer federal-provincial-territorial collaboration, including the municipalities. Those two aspects have also been raised by other stakeholders, along with the fact that the federal leadership in this area seems fragmented. Can you tell us more about those two aspects?
Mr. Iny: On the first point, when a vehicle is operated autonomously or partially autonomously, in an emergency situation before an impact, it is important to analyze the program and have the tools to ask the vehicle where the information is stored. The current legislation expects the information to be contained in the vehicle, but right now, it can be stored in the cloud or outside Canada. We would like Transport Canada to have competent staff to carry out those analyses and communicate with the vehicle properly. The businesses must also understand that the expanded access will include all the information compiled by the vehicle, and perhaps even by other vehicles if the impact involves more than one vehicle. That was sort of our concern.
About a decade ago, we witnessed the situation when Toyota had vehicles that could record the conditions before impact. Allegations had been made, but Transport Canada was not aware of it and, when it did become aware, it did not have the tool to communicate with the vehicle. That should not happen again in the future. For instance, the culture of a company like Uber is not very open to sharing its information, even with transport companies that may be dealing with it. It insists on confidentiality. We want to make it clear that this rule does not apply to safety.
In terms of your second question, yes, if we look at the history of the automobile, the first driver's licences and traffic permits were issued by the municipalities. You would go to the city hall, and you knew either the mayor or someone influential and you would obtain the right to take a carriage on the roads. The same thing is happening now. The municipalities have innovated. Some provinces have started to intervene, but the ideal would be to have standards and some agreement, instead of the lowest common denominator.
Senator Boisvenu: I would like to provide some publicity for your magazine, Mr. Iny. This is the Protégez-Vous magazine, which is very beautiful. I encourage you all to get it and see whether your vehicle is a good buy.
Senator Cormier: Gentlemen, thank you for your presentations, and especially for the work that you do for consumer protection.
Actually, according to a survey conducted by McKinsey & Company, it would seem that attitudes about connected or automated cars vary depending on age, location and income. My question is for Mr. Jack. Do you know whether your members have this same concern? Does the rate of adherence to the concept of automated vehicles vary according to age, location and income? If so, what would be the challenges for the full deployment of automated cars in rural areas, in particular?
Mr. Jack: Younger people are more open to technology, but we were surprised to see that the numbers were not very different. We saw a difference of about 10 per cent when we did our survey of younger and older people.
[English]
Our feeling would be that there isn't a huge degree of difference on that issue. One of the points I made in my opening remarks is that when we dug into the data and isolated out the people who are positive about the technology, still I think it was something like 63 per cent of those people had some concerns.
[Translation]
The people with Apple Watches.
[English]
Even 60-odd plus per cent of them see the other side as well. To us that was very interesting.
[Translation]
Senator Cormier: Mr. Iny, in terms of the consent of individuals to provide access to data, in our daily lives, we are constantly asked to provide consent in order to be able to access a service. In your view, what role could the federal government play to raise public awareness about this issue? We tend to provide our consent easily, which can clearly have a negative impact on the flow of data.
Mr. Iny: In our opinion, the classic notion of consent, which is a free agreement negotiated between two parties that know all the risks of the situation, does not apply here. It is a false consent when someone is asked to check off a little box to have access to a new app on their device, whether for a car, a telephone or another device. The person wants to have access to the commodity and to be able to use it. The little box is often not just limited to consent to the exchange of information, because there are many other things that go with it.
In our view, there must be some guidelines and, first of all, priorities in terms of information sharing. Clearly, information collected for public safety or protection purposes or to prevent criminal activity will not be treated in the same way as information collected for purely economic purposes. There will be a hierarchy and rules. That would be the ideal and it would apply in spite of what the consumer may have consented to without thinking when getting a new application.
[English]
Mr. Jack: If I may just add, what we have always talked about at CAA is the notion of choice for consumers. They probably shouldn't have any choice over anonymized safety data. Transport Canada, I would think, would be very interested in the notion of a real-time recall database, where they will be able to see in real time that a particular vehicle is starting to have brake issues, let's say. There is absolute public value there, public security issues as well, as George said.
In terms of other information being collected, how the individual drives — and, again, we're talking about the next, let's say, five to ten years where that data is being collected and is transmitable. But before we get to fully autonomous vehicles, which people may or may not own individually, there should be consumer choice. So we have used the old- fashioned notion of the pre-sets on your radio as an image for people. Rather than a yes/no, "Do you want all of the advantages of this system but, by the way, we're going to take all of your data and do with it what we want?'' as George says, that's not a real choice. We think there needs to be "I'm prepared to share this and this, but maybe not that.''
As we said, of course we recognize fully that people should have the right to be anonymous. The bulk data about vehicles on the road should not be completely anonymous, though.
[Translation]
Mr. Iny: We would like to see it resolved beforehand, not afterwards. We have seen how difficult it was to deal with certain issues, such as the Do Not Call List and spam, because we were not farsighted from the outset when those practices started. The ideal is to start the work now and bring the industry people together around the table so that we fully understand all the issues and find ways to achieve a good balance in terms of protecting privacy.
[English]
Senator Galvez: Thank you very much. I have been a member of CAA for more than 30 years.
Mr. Jack: Thank you for the publicity.
Senator Galvez: I know that you are very useful to the community that drives.
This issue is very interesting and important and we have discussed many aspects, but with you I would like to discuss two aspects. First, the transition between an autonomous vehicle with a driver and increased electrification of the vehicle, and we now have a series of safety issues. One day it will be driverless, and I want to know how long and when and how this should happen. We are in a transition right now.
The other question is about driving data privacy issues. I have divided this into three parts. There is data that serves my own benefit, because my car is learning how I'm driving. If it tells me, "Oh, you're not doing this right; you should do it this way,'' of course I will give permission right away to use my data for my learning and better driving.
Then the company wants this data. I don't think I want to give that data to the company to use for benefits and other things. As a consumer, I want to make the choice of who is using that.
Third, someone can propose that if I consent, then this data will be used for the public benefit. I will accept that if you can assure me that this data will be used by a university in data mining, megadata, big data calculations. Maybe then I will say okay.
I was at Dalhousie, the only university working with Tesla, and they take this data. If it's for the benefit of everybody because it's for research, then I think I will accept.
If you discuss this with your consumers, maybe it will interest them to see these aspects.
Mr. Jack: Yes. Shall I proceed?
The Chair: Yes.
Mr. Jack: Well, as you'll hear from almost all of your witnesses, nobody has a handle on exactly when this transition will happen. It's one of the great challenges. It's also why we think it is very smart for a committee such as this to be looking at this issue now. I think we can guarantee that we won't all be driving autonomous vehicles two or three years from now, so you have time to do your study and to put your recommendations out and for governments, both provincial and federal, to put the processes in place to be ready.
One of the unknowns is whether we will end up with safer roads in that interim or whether we'll end up with more dangerous roads, frankly, when we have a mix of vehicles on the road. I think it's fair to say that they won't be any safer. Whether they are less safe, we'll see, but you'll have human drivers judging robots and robots trying to judge human drivers at intersections. It's a very challenging transition that I think we face. It may actually push the tipping point a bit earlier. There may come a point where regulators have to consider mandating the majority of "driverful" cars off the road. We're not in that position yet, but we may get there. There may have to be buy-back programs or something because it's going to be a very big challenge.
One thing that may aid that monetarily, to bridge to your second question, is the notion of the value of the vehicle and the data being collected about it. So we can talk from a privacy perspective and from the perspective of Commissioner Therrien, who I know testified before you a few weeks ago, of an encadrement of the industry, and we are supportive of that.
Another school of thought, slightly more libertarian, is looking at this and saying, "There's value in that data.'' We have seen work from a consultant who, very roughly, has calculated and proposed one possible future where you might buy a private vehicle for $40,000 to $50,000, but if you give away all of the data, you might get the vehicle for $5,000 or even for free if you start calculating the value of the data in the vehicle.
Conceptually that is an interesting way to look at where the future may be. Frankly, it may be worth an automaker's time to buy back your old vehicle and give you an autonomous one, in the same way the telecom companies love upgrading your phone, supposedly for free, because they figure that you will use more data, and they've got their hooks into you at that point. We may actually be looking at a future that is somewhat like that. There is clear monetary value in that data. Some people may choose to sell it.
Mr. Iny: If you look at the timeline for implementation of new technology in auto, it will be anywhere from 12 to 30 years before half the vehicles on the road have a particular feature, for example, air conditioning, automatic transmission. The reason why it can take this long is that, first of all, a vehicle has a lot of value. So the existing fleet, even if it doesn't have the latest features, is not something you're going to throw out — like your old big television when you bought a flat screen or a VCR when you bought your first DVD — because it has value and we're invested in it, making payments on it. The traditional model is a relatively progressive implementation of the new technology.
The question would be this: Is it possible at some point down the road — this is what Ian is talking about — that the vehicle will have so little value because it's not electric and not connected that, in fact, that would be disrupted? Otherwise, that's pretty much what we could see. There will be a long transition period where some vehicles will have different levels of automation in them.
The Chair: Supplementary to Senator Galvez, Mr. Jack, when you made your introduction, you talked about sharing of that vehicle data. But the reality is we are seeing that there will be a sharing of vehicles. If you're going to be sharing a vehicle because you're in an urban setting and you have the vehicle from 9:00 until noon and somebody has it from noon until 5:00, that data is in the vehicle and is not yours anymore. Where do you draw the line between the personal information that Senator Galvez is preoccupied with protecting and the vehicle-related data that is less personal? The vehicle might go to McDonald's, but if you don't go to McDonald's, it won't bring you there. Where do you draw the line on data sharing?
Mr. Jack: We don't know where that's going to land right now. As the automakers progress, the technology companies, in the development of these vehicles, are trying out a few different models. They were very interested in putting Wi-Fi behind the dash and having the ability to transmit from behind the dash. Now there seems to be a bit of a movement toward having people port their smartphones and running the vehicle in the future through the smartphone.
I would expect that in a future of car sharing, whoever's running that company is going to want to have an interest in both. In order to get into the vehicle, you are going to have to have some kind of an ID. It will probably run off of your phone. It will be able to know that it's Senator Dawson who is in the vehicle. You're quite right that some of the repair data — the oil level is down; the oil needs to be checked — is vehicle specific, not Dennis Dawson specific.
I think there will be two streams of transmission, but I cannot imagine that the ultimate owners of these vehicles, whoever they may be, are not going to want to know who's in them. Therefore, that will give them the possibility of tracking the information that's collected in the vehicle during the time that you're in it.
As well, there is the Airbnb example where somebody says, "It's me and my elderly mother coming to Montreal for the weekend,'' and it turns out a frat house descends with 16 guys and trashes the place. One would imagine that the owners of vehicles for that reason as well are going to want to know that it actually is Senator Dawson and his very well-behaved wife who are in the vehicle and not four guys driving to Bishop's for the weekend.
The Chair: Thanks, Mr. Jack.
Senator Bovey: I am very interested in all of this, as you know, and I was really interested in the CAA survey. My question is: How is the system going to gain the confidence of the public? How long do you think it will be for that general acceptance?
In looking at this and other things I've read, there seems to be a great deal of contradiction. People like the idea. They're excited about it. They're afraid of it. They trust it more than they trust other drivers. They like the accessibility, but what's going to happen to a wheelchair in an autonomous vehicle? There seems to be a lot of give and take with about 60 per cent of people being afraid of it, if you average the two years, and the real concern about accountability in the event of an accident. How do we move from now, where there's excitement but fear, to general acceptance?
Mr. Jack: Well, I think that people are wise. There are no answers to most of the questions you just raised, and so the fact that people are wondering about them and a bit disquieted about them makes perfect sense. I think time will take care of most of those questions.
We're sitting here in a government panel, and so I would suggest that there's room for a gentle guiding hand from policy-makers in helping in that area. That's, again, why we're very happy; this is the perfect time to be doing this study right now. You're not going to be able to answer every question out there about exactly what government and the industry should be doing, but I think some guidance and a clear explanation of what the issues are that need to be dealt with in the coming years and what the expectation is of government and of industry will be very helpful in doing that.
Senator Bovey: So what would you say the key question is or the key piece of advice we should give the government to move this along? What's the one major thing you think government should be doing?
Mr. Jack: I think this will be one of the most transformative technologies that we have seen, and so governments, at both the provincial and federal levels, need to be taking it seriously.
I had a thought earlier as George was speaking. He mentioned insurance. Auto insurance is largely a provincial issue, but I hope that the committee will see its way through not to tell the provinces what to do — never that — but maybe to list some of the issues that provincial policy-makers should consider.
I think this is the first deliberative body in our country to take a serious look at all the angles of this issue. That would actually be valuable because the car, apart from the safety of the new vehicle, has largely been a provincial jurisdiction. We see with the autonomous vehicle that all of a sudden we're looking at spectrum issues. That's federal. We're looking at international issues, because you don't want your car to die at the U.S. border. That's a federal issue. We're looking at interprovincial issues where the federal government has a role to play. I would suggest that for the first time in history, the motor vehicle is becoming at least as much of a federal issue as it is a provincial issue.
I see signs that the federal government is starting to take this seriously. I know they have a DG-level committee starting to look at the issue. A gentle nudge in that direction from you would be very helpful.
Senator Bovey: Thank you.
The Chair: Thanks for making our job a lot easier.
Senator Runciman: Just looking at the Earnscliffe survey done for you at CAA, these are prompted questions, I assume?
Mr. Jack: As in?
Senator Runciman: The questions that you have responses to. The polling firm is posing that range of questions.
Mr. Jack: Correct.
Senator Runciman: They are selecting from the range presented to them.
I'm looking at improved accessibility for people with mobility issues like the elderly. That's just one area, but we had a gentleman here from the Victoria Transport Policy Institute.
Mr. Jack: Yes.
Senator Runciman: The bulk of witnesses who have appeared before the committee have been pretty optimistic and enthusiastic, while expressing some concerns.
He was a different witness and raised a lot of interesting issues. We talked about the disabled, for example, and he raised the issue about getting in and out of the car, those kinds of issues, which I think aren't addressed in a polling exercise like this.
He talked about autonomous taxis. A taxi probably requires a cleaning on a fairly regular basis; people get sick and all sorts of things in taxis.
Mr. Jack: You have heard from Barrie Kirk.
Senator Runciman: So it seems to me that it might be a useful exercise at some point to pose some of those questions as well during this polling exercise — to raise some of these issues that people perhaps haven't given any thought or consideration to. That's just an observation. I think he raised some pretty valid points that the committee will have to take into consideration. There's no doubt about it.
When we talk about things like GM's OnStar, what happens to that data now with the automakers? Are they selling it? Are they retaining it, even after the subscription expires? Do you have any information on what's happening today?
Mr. Jack: Well, I'm so glad you asked.
Three or four years ago, the then-chair of GM said he looked at telematics as a $2 billion a year greenfield revenue opportunity for the corporation. That is still more dream than reality for the automakers. In my more puckish moments, I think of them as Gollum with his Precious down below; the pursuit and the gathering of "the Ring'' is the important thing. They haven't quite figured out what it's for or what to do with it.
They are sitting on mountains of data right now with hopes of monetizing it, but that hasn't really turned out to be the case right now, largely because they are automakers, not technology companies. We can expect Uber, Google and the other technology players in the sector to be a little bit quicker off the mark in figuring out how to monetize this.
I'll just throw a couple of things at you. You may have already seen this, and I'm sure it's part of your research, but the KPMG 2017 auto survey asked a bunch of interesting questions. One was, "What do customers want in return for their data?'' Eighty-four per cent of consumers want direct monetary benefit in exchange for their data, and 45 per cent of automotive executives believe they need to make no offer of anything in exchange for the data. That's automotive executives who were surveyed not by us — not by crazy radical leftists — but by KPMG. So there's a clear dichotomy. There's a hope on the part of the industry.
The other thing I'll throw out is that the CEO of Intel last year said: "When it comes to the car of the future, data is literally the new oil.'' And I would never accuse The Economist of plagiarism, but into my In Box last night popped the new edition, and it actually talked about car data as the new oil as well.
There is lots of value there. They haven't figured out how to convert it yet, but like the Saudis with their oil, I'm sure they'll find a way.
Senator Runciman: I recall seeing something from Mr. Ford, the head of Ford Motors, with dollar signs in his eyes with respect to what the future holds.
Mr. Jack: With respect, senator, we would say that's fine, as long as the consumer knows about that, has some say in the matter as to "yes'' or "no,'' or has some say to be compensated.
Senator Runciman: I'm not sure if it was Mr. Iny who referenced hacking and the fact that 75 per cent of auto executives have no countermeasure strategy with respect to hacking. I don't know if that indicates complacency or that they are looking at the profit side but not the significant concern of many consumers.
Mr. Iny: It's the culture. Car theft isn't new, and for several decades, the key was a flimsy item costing a few cents, and the industry opposed, having to spend as much as $4 a car for an electronic key.
They didn't pay for the cost of joyriding; that was supported by the justice system. They didn't pay for the cost of replacing the car, which was covered by insurance. So they are to some degree insulated from that.
Hacking is a new form of theft. I'm not including terrorism, because that's outside of my knowledge, but obviously if someone could just program your car to leave your driveway in the middle of the night without even having to hire a human to do it, it's probably safer, because the car will drive better than the person you hire to do that kind of work.
But, yes, it is possible that theft rates might increase. What's clear is that the industry hasn't adequately planned for this. The disappearance of the manual key has made it easier for the people with the right technology to steal a car, because you no longer have to actually damage the vehicle or have someone pop it; you can do it remotely. You still need someone to get in and drive it.
It is a concern.
Senator Eggleton: I'd like to go back to the question of the transition period and the integration of more and more. We're in an automation-connected vehicle era. It's gradually happening, but at some point we'll get to what is called level 3 of automation, which is described as an automated driving system that performs all aspects of driving with the exception that the human driver will respond appropriately to a request to intervene. Level 4 and level 5 even go beyond that, to more autonomous driving.
I don't know when we'll get to a level 3 — maybe soon, maybe in a few years. I don't know who will buy those first few vehicles. It might be people who have a fair bit of money, because they could be quite expensive. I don't know about that.
Then the question is: How do we make it safe for people to drive in this mixed field of vehicles, with some of the kinds of vehicles we have today? Some people keep their vehicles a long time. There are country roads; there may be more of those than city roads. Then there are level 3 autonomous vehicles.
One of the previous witnesses suggested that maybe we should have a dedicated lane on highways. That might be the start to it. He also pointed out that that would be serving affluent customers, whereas those who pay for this road through their taxes — there would be a question of social equity.
What about that? How do we prepare for this mixed driving situation, and do we need things like dedicated lanes?
Mr. Iny, you mentioned the elevators, but the elevators are in elevator shafts on tracks; but not these cars.
Let me hear from both of you your suggestions about this integration period, because this could be quite challenging.
Mr. Iny: First of all, it would be easier to program the automated vehicle to recognize human behaviour than humans to recognize an automated vehicle, perhaps. I don't know which would be more effective.
The people introducing the new technology would have to recognize that it will be operated in a mixed fleet. Hopefully we will have the ability to monitor and intervene if that activity isn't safe.
I do think a dedicated lane is probably a good way to do the rollout, but segregating the transportation system isn't ever going to happen that way I don't think. You could have it initially, as we already have segregated lanes for certain classes of vehicles, and perhaps those could become lanes for automated vehicles. If we are looking at a 30-year phase- in, then those are good questions to ask.
Certainly one way around it is to require both the human and the automation, which means the driver will be prompted every few seconds to put their hands on the wheel. This is what cars today do, the ones that have a near-level 3 capability. They do ask you to still pay attention.
Mr. Jack: The Tesla crash in the States was a cautionary example of a near autonomous vehicle and human error because the human was too confident in the automated features they had.
As George says, to me the reason this is probably going to take 30 years is not because of the technology. We're almost there now and I'm sure within five years we will have the technology. It will be all the other factors, whether they are infrastructure-related or the very challenge you lay out about how we integrate the two at the same time. That's what's going to take a long time and it's going to put the brakes on the industry. I think people will slowly get more and more used to the technology and that will be helpful, but that suggests a graduated rollout.
It is most likely we will see fleet operations going autonomous first. I can see a segregated line on the 401 for autonomous trucks. Those of us who are driving personal vehicles will start to see that over time, and then eventually we will calm down about that and the technology will become more mass-market at that point.
As I'm sure you heard, there are already examples on closed campuses of autonomous buses. That is an obvious fleet operation.
We will see limited delivery services next, and from there it will slowly move out into personal vehicles, not because the technology won't be there, but because it will be a play thing for the rich, for the next 5, 10, maybe even 15 years.
The more likely future is the fleet. That will get people used to the idea and it will move on from there.
Senator Eggleton: Should taxpayers be paying for all of this when just certain businesses or people of means have these kinds of vehicles?
Mr. Jack: On the commercial side I guess the answer is yes. There is an economic benefit if we're talking about a truck lane on the 401.
I think industry will end up paying the lion's share of this. We can look out 50 years, and who knows what the world will be like? I don't see a world in the future where municipalities, provinces and the federal government will have the money to implant sensors under every kilometre of every road in this country. Therefore, for this technology to work at all, it will have to be vehicle-to-vehicle, and the government is not developing the vehicles; the private sector is. If they really want this to work, they will have to figure out appropriate protocols. The federal role would be to ensure the spectrum is there, for example, all the work that the federal government does right now for the telecom companies. So that's at an added burden. We're moving up to 5G and who knows if there will be a 10G in 10 years, but that makes money for the government.
I don't see a large government role in infrastructure. Certainly it's going to be more on the policy side. I'm sure you have already heard questions about how many vehicle kilometres will be travelled. The number will dictate municipal planning. Will we need parking garages in 20 years? What will we do with these spaces? All of these questions will be interesting and have an impact on the tax base, more so than an expectation that government will be spending billions to develop the infrastructure.
Jason Kerr, Director, Government Relations, Public Affairs, Canadian Automobile Association: I attended an OECD international transportation forum round table in December. We seem to always be kicking around the idea of a segregated lane of traffic for autonomous vehicles. Another option for consideration was one raised at that round table, that the dumb vehicle will be parked at the outskirts of the downtown core. In the core, you will have only autonomous buses and taxis on certain streets, essentially shuttles that take everyone around the main core. Then you can think about trains coming into the downtown core and people being manoeuvred on autonomous vehicles for the last mile.
It's not just the concept of a dedicated lane, although I agree that's the first way that we will see it. I think we'll see the platooning of trucks, shuttles and taxis, mixed with trains that are in some cases already autonomous because they are on tracks. That will perfect the technology over the years and make people feel more comfortable using the technology.
Mr. Iny: The taxpayer will ultimately pay, maybe not in the form of taxes, but in some other way. If we benefit from this, hopefully it will bring us a benefit greater than the cost.
The last time that these huge infrastructure investments were made was with the arrival of the motor vehicle, when they had to introduce signalling in cities at the corners, which didn't exist before. Road signage had to change dramatically and finally roads as well. There is a tremendous cost there.
It's too difficult to predict what could happen going forward. But the consensus or the belief is that, as Ian said, the spend would have to be on the vehicles this time around. Society will not be as willing to support the cost as we would have been in the 1920s and the 1950s when those big interstate projects were approved.
Senator Eggleton: When we deal with public policy on this and the different government levels, should we be updating current laws or rewriting new laws to accommodate autonomous vehicles, particularly in this transition period? What we have is basic. It deals with safety and the things that government is concerned about. Should we stick with that and perhaps make some amendments to take into account automated vehicles or is it time to rewrite the whole thing?
Mr. Iny: I'm a believer in building on what you have already and what already works. Law tends to be very predictable when it's based on something that happened earlier and less predictable when you're starting with a clean sheet. If you can find the roots of what you need in what is already there, then you should build upon it.
This was touched on earlier, but it's probably changing so quickly that writing regulations right now might prove not to be up to date. There might be other ways this can work by having an active engagement with players, not just with conventional auto companies but these new technology companies, where you would use other tools, such as memoranda that are good for five years. Ideally you would be doing it proactively before the devices are already on the road, at which point you will have a lot of resistance from companies on what they consider interference in their work.
Senator Runciman: I'm curious about on-road testing. What role do governments play with respect to autos coming on the market? We've talked to Ralph Nader in the past and Volkswagen recently where commitments are made, but then it turns out in many instances they were not accurate, to say the least.
What kind of processes are involved now with government approvals? I know that, for example, the Tesla semi- autonomous vehicle had the fatal accident in Florida and they found it was a problem with their auto pilot. What is the process and what's it going to look like going forward?
Mr. Iny: The current thinking is that a human has control of the vehicle and you need to get an exception or exemption for research purposes if they don't. If we're going to something that is semi-autonomous, you could probably do that within the existing framework because the human still has to be paying attention and the vehicle has fail-safes that will turn the system off if it judges that you're not paying attention at all. That's where we are right now.
The next leap is the reason we're here today.
Yes, I think we should consider that somebody might try to cheat or might simply not have tested something thoroughly enough because they want to get a jump. We have a system in place for that. It works well most of the time. It needs to be tooled up to deal with those situations when they occur.
I hope we won't get those spectacular failures like the one you mentioned with VW or General Motors where, for more than a decade, the companies have more or less the certainty that no one with oversight will catch them.
Mr. Jack: It won't surprise you to hear that the Canadian approach is typically more cautious in terms of allowing these things. It is a bit of the Wild West in a couple of U.S. States right now, which I suspect they are rethinking.
I would offer one other thought that ties together a few things we have been talking about recently. It's the increasing privatization of ground transportation, if you will. Whether we're talking about private companies collecting data, whether we're talking about Uber in Innisfil, Ontario, now providing the transit service at a subsidy, this is ride-sharing. Again, this is the question of car ownership or not, not autonomous vehicles per se.
It used to be that the only people who knew any data about you were the traffic police or the Ministry of Transportation when you got your licence and insurance. Now an awful lot of that is moving into the private sector. That's our point about data privacy.
I would say, therefore, there is a role in public policy — and, perhaps, for the Privacy Commissioner — in ensuring that that data is being properly looked after. We're talking about vehicle to vehicle instead of the traffic lights that the municipalities look after.
The other thing is that from a federal point of view, and perhaps to Senator Eggleton's point a little bit, most of these regimes are set up municipally and provincially for testing. If there is a federal role, it might be to continue to expand on making sure that we do some of the R&D in this country.
I read on the front page of my Globe and Mail today that another leading Canadian researcher in artificial intelligence has been bought up by Uber and is now moving to the States. This is on ongoing phenomenon that's happening on the science and tech side of things. That's something for the government to be aware of and think about.
[Translation]
The Chair: I would like to thank our witnesses for their input today.
[English]
Honourable senators, tomorrow we will hear from representatives of the Innovative Vehicle Institute and the Online Network-Enabled Intelligent Transportation Systems.
I would also like to inform you that our analysts prepared two documents about automated vehicles in the United States and other countries, which have already been circulated to members. If senators have any questions about these documents, you can contact the analysts directly.
I will adjourn the meeting, but I would like to ask the members of the steering committee — Senator MacDonald, Senator Bovey and Senator Boisvenu — to remain in the room so we can have a meeting of the steering committee.
Thank you very much.
(The committee adjourned.)