Proceedings of the Standing Senate Committee on
Transport and Communications
Issue No. 18 - Evidence - May 30, 2017
OTTAWA, Wednesday, May 30, 2017
The Standing Senate Committee on Transport and Communications met this day at 10:00 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, this morning, the committee is continuing its study on the regulatory and technical issues related to the deployment of connected and automated vehicles.
[English]
I am pleased to introduce our witness, Mr. Ross McKenzie. He is the managing director of the University of Waterloo Centre for Automotive Research, WatCAR, one of the top automotive academic research centres in North America.
The University of Waterloo was also the first organization approved to operate an autonomous vehicle on public roadways in Canada under the Ontario Ministry of Transportation autonomous vehicle pilot program.
[Translation]
Mr. McKenzie, thank you for joining us this morning and thank you for inviting us to visit your research centre. Please give your presentation and then senators will ask questions. The floor is yours.
[English]
Ross McKenzie, Managing Director, Waterloo Centre for Automotive Research (WatCAR): Thank you very much. I'm happy to be here today.
Thank you for your patience for my tardy arrival. A crew member this morning was ill and couldn't make the flight. The passengers were ready to go but there was not enough crew, so thank you for tolerating the delay.
The Chair: We don't have automated planes yet, and we hope we never will.
Mr. McKenzie: Actually, we do, but only once they're in the sky. Autopilot, anything with autonomous vehicles, we'll get into that.
It is my true pleasure to be here this morning and join you. I thank you for this opportunity.
There are some diagrams through the opening remarks that I have to offer this morning, so I invite you to follow along, if it pleases you, to track those as I present my opening remarks.
Globally, a rapid in-vehicle technological convergence is taking place within passenger vehicles. The convergence accelerating these advancements in vehicle connectivity and autonomous driving is happening at a rate considered near normal for the information and telecommunications technology sectors, but at a rate of unprecedented speed for the automotive and intelligent transportation systems sectors.
There are four distinct stages in vehicle development that will culminate in fully autonomous vehicles being realized on Canadian roadways: connected, automated, highly automated and, ultimately, fully autonomous vehicles.
As we move forward, the connected and partially automated vehicles on our roads today will soon be truly automated. Then these vehicles will become highly automated and ultimately fully autonomous. The time it will take to achieve this, however, is the subject of much discussion.
The answer to "how long?'' depends on what you define as an autonomous vehicle. If "autonomous'' means getting around your city, the city that you live in, then perhaps 10 to 15 years is the likely time frame. If "autonomous'' means driving from one city to another over streets, highways, freeways and exit/entry ramps, then 20 to 30 years is the likely time frame.
What is known, however, is that the increase in on-board vehicular computing power is rapidly increasing and will continue to expand at a near exponential rate. Vehicles continue to receive an ever-increasing volume of data from more and more external sources, as well as an increasing number of on-board sensors, as shown on the diagram at page 2: connectivity, GPS, vision systems, radar, LIDAR, wireless sensors, autonomy, machine learning, mobile security, vehicle control, safety and power management. To achieve autonomous driving, there is a need to process all of this information instantly, in real time. The processed data is then used by the vehicle's computer to make operating decisions for deployment through an ever-expanding number of Advanced Driver Assistance Systems technologies, or ADAS, all on the vehicle.
The autonomous vehicle revolution is being driven by enhanced vehicle connectivity known as V2X communication, where "X'' represents everything from the driver and passengers through the vehicle to other vehicles, known as V2V; to roadside infrastructure, known as V2I; and through the IT cloud to the Internet of things, known as V2C.
In Canada, we are embracing this revolution through a natural synergy that emerges from the overlap of strong, established IT and automotive clusters in the same geographic space. In IT, the 110-kilometre Toronto-to-Waterloo innovation corridor is the second largest in North America, after California's Silicon Valley.
In automotive, the 425-kilometre corridor from Toronto to Windsor assembles more vehicles than any other state or region in North America. The overlay of two strong, established clusters in IT and automotive assembly is the only one of its kind in North America.
In Canada, we are engaged in connected and autonomous vehicle development, leveraging our ability to draw upon the significant strength of these two overlapping clusters.
I'll apologize to the French-speaking senators; the diagram at the top of page 4 is the only diagram I did not have an opportunity to translate, so I do apologize for that.
Currently there are two leading sources working full time on autonomous drive development in Canada: the University of Waterloo and BlackBerry QNX here in Ottawa.
We have a strong, successful track record of collaboration between the private sector, government and academia. Whether in intelligent transportation systems, ITS, information and communications technology, ICT, automotive assembly or automotive parts production, in Canada we appreciate the diversity of perspectives and are leveraging this to advance the understanding, application and deployment of technologies.
This fall, Canada will host the twenty-fourth intelligent transportation systems world congress in Montreal from October 29 through November 2. As a member of the board for ITS Canada who was involved in a successful bid to bring the world congress back to Canada for only the second time, I know there is anticipation from across the country for this event. The world congress will profile, promote and celebrate rapid technological advancements, not only in ITS but in connected, highly automated and autonomous vehicles from North America, as well as Europe and Asia.
I sincerely appreciate this opportunity to address the committee today. I welcome your questions and I look forward to a productive discussion.
The Chair: Thank you very much, Mr. McKenzie.
Senator Galvez: Thank you so much for your very interesting presentation. We have been listening to experts in this field for quite a while now. I realize I was asking very technical questions, but last week was a break week, so I had time to reflect. I'm an engineer, and I understand how thrilled you are about this new area and how passionate you are in this competition. I feel the adrenaline from you.
However, now I'm a senator, a citizen, and I need to bring down to basics what are the social needs and priorities, and how this emerging industry is going to answer to these needs. The needs are mass transportation, getting rid of congestion, efficiency of these cars — because 90 per cent of the time the cars are in the garage — and green energy. This is what we need. We no longer need fossil fuel cars. Finally, we need a reduction in resources of the materials in the car.
I understand things have changed since Apple, and Apple doesn't do any more needs analysis. The guy says, "I'll tell you what you need,'' instead of society saying, "This is what we need.'' I understand there is some change, but please enlighten me with your analysis.
Mr. McKenzie: There are two big opportunities, in my opinion, that autonomous driving and autonomous vehicles hold for society. One is to augment mobility systems that are in place, whether that's a bus transit system or an urban metro subway system in a larger urban centre. The way that's going to be realized is with people who don't necessarily want to take the subway.
Take my mother, for example. She'll turn 80 this November. Two years ago she asked me, "Will I ever be able to use one of these autonomous things you're working on?'' Her mother lived until she was 95, and her grandmother lived until she was 102, so I think chances are pretty good.
But she said to me, "You know, I've driven enough; I don't really need to drive anymore. So if a vehicle that I didn't even own could come and pick me up, take me to the dentist, bring me back home, maybe I wouldn't need a car; maybe I would just need to use a vehicle that could be shared with others.'' This is my mother, from watching the news and picking things up from extensions of Uber-type applications.
There are people who will not use mass rapid transit because they're either not interested, like my mother, or they perhaps have some need — whether they might be handicapped or they're going to go shopping and have a bunch of groceries that they don't want to lug them home on the bus — where I think these types of vehicles will serve to fill a void.
In the process, as they operate autonomously on our roads, they will not run red lights. An autonomous vehicle doesn't look down to reach for its coffee cup in the centre console and miss a stop sign, which ironically I did on my way to give a talk about autonomous vehicles in Stratford. I was ticketed by the Stratford police for failing to stop and rightly so, because I was in the wrong. But if I had been in an autonomous vehicle, I wouldn't have missed that stop sign.
There would be less congestion because those autonomous vehicles will drive conservatively; they will leave space; and they will not rush through an intersection, perhaps not be able to clear it and congest by blocking traffic going the other way.
In terms of the efficiency of use and more green energy, I think that's a natural evolution that's going to happen and is happening now in vehicular transportation, regardless of whether it's autonomous or connected.
You've got people like General Motors with their electronic bike, or e-bike, initiative, working on that project in Oshawa. The main target of that is something called first-mile/last-mile. It's a gap in the system whereby people live just a little too far from the transit stop. Let's say you're in Toronto or Montreal, and that transit stop would get you, if you could walk to it, cleanly to the commuter train to take you into the downtown metropolitan area.
Coming back to the individual, they'll drive a car because it's a little too far to walk, or maybe the weather is inclement. If they had the ability to get on another mode of transportation, like an e-bike, and take that and lock it up in a bike parking stop at the GO station or the metro station, imagine the vehicles that could be removed from the commuting exercise in an urban area, let alone the parking spaces that wouldn't be needed anymore.
As far as social needs and priorities go, I don't pretend that connected and autonomous vehicles will be the answer, but they can certainly contribute in many positive ways to an appropriate solution on some of the points that you've articulated, in my opinion.
Senator Galvez: Thank you.
[Translation]
Senator Saint-Germain: Thank you for your presentation; it was really very interesting. I refer to page 3, where we find your table explaining the autonomous vehicle revolution and the issues of interconnectivity. This leads me to talk to you about governance, especially during the transition period. You say that, 20 or 30 years from now, most vehicles will be autonomous, but there will certainly be a transition period.
You are very reassuring about the technological and engineering aspect. But that is not the aspect that interests or concerns me the most. My concern has more to do with this whole dimension about managing public infrastructure, especially the connectivity with roads, and the communications issues — the V2I and the V2C — as well as the issues of legislation and regulation, whether they are administered federally, provincially or municipally.
Do you think that the technology will be available but that the legislation will not and, therefore, the implementation could be delayed? Is there some risk in that respect?
[English]
Mr. McKenzie: I think it's possible that technology could be ready sooner. However, before I get into the detailed answer to your question, I'd just like to step back from the opening comments. Perhaps I misstated this or misspoke, but even if every vehicle is autonomous, not all vehicles will operate autonomously. Much the same as the taxi driver who brought me to the airport and drove the vehicle today, there will always be vehicles that will be driven manually by human operators on public roadways.
[Translation]
Senator Saint-Germain: Let me modify my question: is it possible that the technology will be ready but there will be issues in regulation and in adapting infrastructures, especially in the hybrid context where autonomous and non- autonomous vehicles will be operating together? There is an issue there and you see where I am heading. How do we operate in parallel, but by means of a coherent program that will lead to governments and the industry arriving at the same destination?
[English]
Mr. McKenzie: I do see where you're heading. I understand.
[Translation]
Senator Saint-Germain: I am autonomous.
Mr. McKenzie: Wow!
[English]
The challenge for the technology is those drivers who will manually operate their vehicles. If we were looking at a situation today where the roadway infrastructure that you are forward thinking on and the related governance issues, technology being ahead of the curve, if I can phrase it that way, if we were only dealing with vehicles that were going to have autonomous operating modes, then yes, we would have a significant problem. But we're not likely to ever have that. There have been some proposals for isolated areas where only an autonomous vehicle can operate. Quite frankly, if we had that, we would have autonomous vehicles today, because the biggest challenge for autonomous vehicles isn't to operate on its own on a road autonomously around other autonomous cars. The biggest challenge for autonomous vehicles is to be able to safely interpret and react to the random behaviour of pedestrians, bicyclists, motorcyclists and human operators of other vehicles.
I have no scientific data to back up these numbers, but my opinion is 20 per cent of the challenge is autonomous driving, and 80 per cent of the challenge is getting the autonomous vehicles sufficiently advanced to handle the random acts of humans that they have to interact with because we're putting the autonomous capabilities on top of the existing infrastructure of manually operated vehicles, pedestrians on our roadways, cyclists and motor cyclists.
To your point about technology getting ahead, I can only see that happening if a municipal jurisdiction took an initiative to declare a certain space within an urban environment or an urban/rural environment, if that's what their responsibility of oversight covered, and declare that space autonomous vehicle eligible. The legislation, as it stands right now, is only in Ontario can you operate autonomous vehicles on public roads and only with a human operator in the driver's seat ready to take control at any time.
Unless jurisdictions, whether they be municipal or provincial, are going to define areas where this is not an autonomously exclusive zone but an autonomous testing and operating area, only then will there be any potential for the technology to leapfrog the governance.
There are municipal initiatives under way. I'm familiar with one in Ontario. An organization known as the Ontario Good Roads Association formed an automated and connected vehicles committee with invitations only to some of the more progressive jurisdictions, such as the city of Brampton and the region of Peel. There are about 10 members. They are just starting small and they intend to grow it organically.
They are addressing it from the perspective you have articulated, which is, what do we need to do to be ready for this on our city streets? From a provincial jurisdiction, if it's not a provincial highway or roadway of responsibility, the challenge and the responsibility both fall to the municipal jurisdiction.
I can't speak for the rest of the country. I'm sure there are discussions, but in the absence of an AV pilot program in other provinces, it's less of a priority.
They are moving along at a reasonably parallel pace right now, partly because the municipalities are being proactive and partly because the technology to handle random human behaviour is going to take a considerable amount of time to achieve for fully autonomous vehicles.
Senator MacDonald: Thank you. I want to speak to you about security. We saw this week where all the planes in London were shut down. Heathrow was shut down because of a lack of security.
BlackBerry QNX is here in Ottawa and, even though BlackBerry has done a very poor job at marketing, it's a well- known and established fact that BlackBerry has some of the most secure devices in the world.
Do they have the security necessary to ensure — and can they transfer this technology — that somebody couldn't hack the system of a fully automated vehicle or a fully automated system? Imagine the chaos if somebody hacked these systems and these vehicles that can do everything without us worrying about them. All of a sudden they can't do anything without us worrying about them. I would like your thoughts.
Mr. McKenzie: Within any operating system, software-driven application — whether it's a smartphone, a vehicle, or all points in between, on the cloud and through the Internet of things — cybersecurity, and especially the realm of mobile security, is something that is being actively and aggressively pursued and investigated. There are technologies out there that can facilitate detection and prevent hacking at a very basic level simply by monitoring the power that is being consumed.
Think of it as something similar to a heartbeat monitor and sinus rhythm and this is what it does. You have a system on the autonomous operation that monitors that power consumption and looks solely for variances. Whenever it detects a variance, it puts up a red flag and asks, "why has it done this? Is this correct?''
There are very simple methodologies that are actively being deployed. I'm familiar with some of this because of work going on at our institution, but also through conversations and discussions with some of the assemblers. General Motors, for example, has an entire cybersecurity unit that is doing nothing but thinking ahead and tapping into the mindset of would-be hackers by employing some of them.
You're never going to have it completely, 100 per cent fail proof, but there are strong levels of awareness of the potential for disruption. We saw two years ago somebody quite easily hacked into a Jeep. That led to a global recall for a certain series of five or six years' production of Jeeps because they were able to hack into the on-board system just driving along side, and that wasn't even an autonomous vehicle.
I strayed a little bit, I think. I probably rambled, and I apologize for that. If I didn't answer your question, I could add more if you have a supplementary.
Senator MacDonald: I do have a supplementary. You referred to the monitoring of data use to detect whether something is being tampered with. Are there already examples of where the monitoring of data and usage have been obscured by those people who are tampering, as in it's not possible to measure it because it's been obscured?
Mr. McKenzie: Are there examples where it has been tampered with? Absolutely, yes, there are. The Jeep example I just mentioned would be a very basic one.
You have the system on a Tesla right now. When that software update came down at nobody's request, it updated everybody's operating system. They got in their vehicle and it said, "You now have autopilot,'' which is a grossly misnamed vehicle feature function, because it's not. It said that by using autopilot you agree to sit in the passenger seat, pay attention and have your hands ready to take control of the wheel at any time. Everybody accepts that and off they go.
It didn't take 24 hours and there were pictures of people sleeping while on autopilot, driving through the streets of San Francisco. There were pictures of people who hopped out of the driver's seat and were sitting in the passenger seat, with nobody in the driver's seat, and it was operating in autopilot mode.
You can only be as vigilant as you trust people to be, ultimately. It isn't just the people who could potentially hack; it would be the people who will abuse the privilege of what they're being given in the opportunity to deploy that also presents a challenge.
[Translation]
Senator Dagenais: Thank you, Mr. McKenzie. What you have told us about the components of these vehicles of the future indicates that there will be a good number of companies involved in development and manufacturing. To what extent will the companies making the components be Canadian companies? How many jobs could be created with these new technologies? To what extent will the vehicles being built here have Canadian content?
[English]
Mr. McKenzie: The involvement of Canadian companies has the potential to be significant. I will refer you to page 4 and the diagram at the top. Looking first at the purple dots, which are labelled "automotive original equipment manufacture research and development,'' auto OEM R&D, down in Windsor, we have long-standing facilities of both Chrysler, now FCA, and Ford.
In Oshawa, at the other end of the red ellipse, we have the long-standing General Motors technology centre. However, that GM technology centre just doubled in size in the last 24 months. There are now over 300 people employed there, and they actually did the retrofit of Chevrolet Bolts that are now being tested and deployed on the Warren, Michigan, R&D campus of General Motors. They utilized suppliers from Canada — to what extent, I don't know — as well as suppliers and technologies from other countries, as well.
In Toronto, the second purple dot with G is a new General Motors Markham campus that currently employs just under 300. That's from zero staff less than 10 months ago, and it is ramping up and will ultimately employ up to 750 people in pure software programming and development applications for autonomous vehicles, infotainment, on-board passenger options, as well as connectivity.
In Waterloo, General Motors opened a "technology outpost,'' they call it, just a little over a year ago, where they are looking to mine the ecosystem from an IT perspective, as opposed to the automotive perspective that their operations in Oshawa across the street from the production line have generally done in their tech centre.
Just three months ago, I'm sure you're aware of Ford's move to hire 400 former BlackBerry engineers who were working purely in the infotainment space. They are now working in offices here in Ottawa. There are about 200 here in Ottawa, 100 in Waterloo and 100 in Oakville at Ford's headquarters who were formerly at the BlackBerry campus in Mississauga.
I don't know if you can say those are jobs saved, but there is a recognition among the "Detroit 3,'' in my opinion, that they have an opportunity here in Canada to undertake software development and related programming technologies and requirements, because Canadians tend to stay employed with their employer longer.
In Michigan, from the triangle that runs from Detroit over to Ann Arbour and north to Flint, there are about 375 R&D centres in the automotive industry. Any time anyone in one of those centres has some degree of accomplishment — maybe they register a patent, maybe a product has launched and they were part of the team — they get recruited heavily and are frequently relocating out to Silicon Valley. Once they go to the valley, they tend not to come back. That's not to say Canadians don't also go to California, but fewer go and more repatriate.
General Motors initially and now, in my opinion, Ford have recognized this. This is one of my reasons for the emphasis on the overlaying clusters, because while there is some opportunity in parts and material supply, the real opportunity is in programming, enhancing the on-board operating systems and enhancing the control systems with the increased computing power being put on vehicles.
How many jobs? How many companies? We are at a very opportunistic moment in the evolution of this sector in the automotive and mobility systems space. We are well positioned to make a difference and be a player. We are starting to see some of that, but it will take a continued effort and commitment by both the federal and provincial levels of government, in co-operation with entities like the Automotive Parts Manufacturers' Association, which I know has presented to you, and other jurisdictions to continue to build the awareness that we have the talent and the ability to contribute to this vehicular mode of transportation's evolution — or revolution, depending on your perspective.
[Translation]
Senator Dagenais: Can we say that the A1 will become a kind of technological laboratory, like F1 is?
[English]
Mr. McKenzie: Can we say that we would become a technology lab? Is that the question?
[Translation]
Senator Dagenais: We know that Formula 1 is recognized as a laboratory for automotive manufacturers. A lot of technological trials go on there. Can we think that the automated electric vehicle, the A1, will play the same role in your area?
[English]
Mr. McKenzie: Yes, I think we can. For proof of that I will refer back to the diagram and talk about the blue dots labelled "Automotive IT R&D.''
Google has operations in both Waterloo and Toronto that are involved in various elements that contribute to android auto and the Google maps function that is so intuitive now, utilizing a degree of artificial intelligence. Apple has some operations not solely focused so much on research in Toronto, but here in Ottawa they have something they won't publicly disclose. It's across the street from QNX, it's called Project Titan and they've hired a few senior engineers from BlackBerry's QNX division going back a year and a half ago now. It's strongly rumoured within the industry that they're working on a car and whether they would assemble and build that consider is highly doubtful. However, it's also strongly rumoured that Magna is bidding earnestly to be the vendor of choice in Apple's eye to potentially build those vehicles.
We've also seen partnerships with Google and Fiat Chrysler, who are outfitting Pacificas in initial trials of these vehicles which were assembled in Windsor and then retrofitted across the river in Michigan at some FCA facilities. An initial trial of 20 was so successful that they have now gone on to a further commitment of another 100. Then you have Cisco Systems with an innovation centre in Toronto that has a portion of its mandate focused on automotive development.
Not on the map, Microsoft purchased a Waterloo start-up known as Maluuba, which has relocated to Montreal to be in the heart of that AI cluster rapidly emerging down the Ottawa River from us.
There are all these things and they are interconnected. As long as we're leading in technology or actively involved, there are pieces of it from all these different sides that will play a role in connected and autonomous vehicle development.
Senator Bovey: Thank you; I find this very interesting and exciting. I read an article recently that talked about the fact that the 1919 World's Fair was already dreaming about autonomous vehicles, so nice to see that 98 years later we're catching up.
I was intrigued with your comment about the processed data being used by the vehicle's computer to make operating decisions for deployment. It seems to me, in an article I read by Hod Lipson, a robotics engineer, that he talked about data being the fuel to teach the car how to do what it's going to do. I think we've been focusing on data as questions of privacy and security. Can you talk more about data becoming the fuel for the automobile's decision- making?
You talked about clusters and the work that's going on, do you see an opportunity here now to form a consortium to make application to the new supercluster granting program that was announced last week?
Mr. McKenzie: I'll talk about data first and then conclude with a response on the supercluster question.
An autonomous vehicle never drives down a road that it has not been manually driven on. When it manually drives down that road for the first time and it subsequently drives down that road again, whether it's in autonomous or human-operated mode, it's making the most detailed map you can imagine. It notes where the signpost is that's holding the stop sign up and then uses object-detection technology to validate that that's where it is with radar the next time it goes down that road. With the ever-increasing numbers of sensors and the amount of information taken in by it, it's taking a picture and building a map far more complex than we can imagine.
Yes, data is absolutely crucial. And in the instance of that data you raise an excellent point: The data collected is collected by the vehicle and maintained on the vehicle so the vehicle can learn and the next time it goes out it draws on that from itself. It's not in any way an issue with the validity of that data or the authenticity of it because it collected the data itself.
As to opportunity under the new superclusters initiative announced last week, I think the opportunity lies in pursuing connected and autonomous vehicles under an umbrella of advanced manufacturing. I think to put all our eggs in the connected and autonomous basket when so much of what is being realized is a segment of the advancements being made in artificial intelligence and machine learning, let alone improved production processes and enhanced methods to integrate more software and digital functionality into traditional mechanical operations, I would say advanced manufacturing would be a better route to go as opposed to a pure connected and autonomous for a supercluster proposal.
Senator Runciman: Touching on what Senator MacDonald raised about security, I'm wondering if you have some views with respect to the role of government in terms of that issue. You mentioned Tesla making these changes and then announcing them to owners after the fact. We know that there are cars on the road. I believe Tesla was announcing they were going to start running cars between Los Angeles and New York, if they haven't already started, and Uber is doing this.
I'm wondering about the role of government in terms of regulation, because there are references in the past to the trust factor with respect to the auto industry. We've seen what happened with General Motors, with Volkswagon and others, allegations with respect to Fiat Chrysler. I'm not a big fan of increased regulation by government, but I think there is a critical role here with respect to these vehicles going on the road and I'm thinking more of security from the terrorism perspective. We've seen autos used in a number of incidents and not just in Europe but in Canada as well. I'm concerned about leaving this up to the industry and saying we have a standardized approach this, but I think there has to be a more intensive role played by government. Do you have a view on that?
Mr. McKenzie: I would agree, at least from an initial stage or initial phase of setting it up, to make sure that it's structured to satisfaction for the safety of society, that it's structured such that there is accountability and not just some token effort by industry because they're saying that they're self-regulating when in fact they are just doing the bare minimum they have to do, because we've seen that happen in other instances in other sectors.
Overall, in the medium to long term, once something is set up and established, it might be simplifying it a bit but to make it no more of an oversight than you have speed monitoring on highways now. You trust the public to go the speed limit and then you have random moments where police officers are monitoring people's speed and may or may not be issuing tickets. The system is established and everybody knows where they need to go for safe operation of their vehicle and, for the most part, you can trust society to do that because we drive safely every day. So, for the medium to long term, less oversight but initially, to be sure it's structured right for society as a whole, I would agree with your approach.
Senator Eggleton: The science marches ahead in new technology before public acceptance, before governments get in to regulate and all that, and of course there is the period of time of transition through systems. We talk about the end state with autonomous vehicles, but there is a long time in getting where they will mix and mingle with other kinds of automobiles, the ones we're used to.
How valuable was it to do more of these tests in off-road zones, perhaps an industrial campus or some zoned area that could help test the vehicles and build up public confidence that they can operate before you totally mingle them into regular traffic out there in the world?
Mr. McKenzie: I think it was very valuable. It was a strategy that Google deployed initially when they first had what was known as the Google car, which has now been succeeded by something that looks more like a Volkswagen Beetle. The original Google autonomous car was a Lexus SUV. It was restricted to a confined area in Mountain View, California. That number of miles travelled — Google always talks about the millions upon millions of miles their vehicles have logged — the vast majority of them are in that space. The vehicle gets to know the roads so then it can focus better on the other random things that could come along, whether it's a pedestrian or a human operator of another vehicle. Demonstration zones or areas serve a definite purpose because you can test things out in a confined area and find out what works and what doesn't and then from that point of learning scale it up for broader deployment.
Also useful, just in a general sense, we had rented a parking lot last November and December so that we could do autonomous vehicle testing on snow. Roads get salted or cleared, and we brought in the snowplow to clear it eventually, but we never salted it so we could test in these adverse conditions. That's one of the distinct advantages we have in Canada, in my opinion, is to do that authentic, cold-weather testing, but not just up in some isolated area of the Far North where it is really cold but in an urban environment that has people and population and activity.
Senator Eggleton: Does your operation WatCAR do any of the demonstration zone type testing?
Mr. McKenzie: Not as of yet. There is a demonstration planned between the Automotive Parts Manufacturers' Association and the City of Stratford, and we will participate in that.
Senator Eggleton: I might add, one thing you pointed out that is very exciting is the IT corridor of Toronto- Waterloo being the second largest in North America after Silicon Valley, and that the automotive sector assembles more vehicles than any other state or region in North America. Better not tell Donald Trump that.
This overlay, as you point out, brings two strong capabilities together here. There is also the question you raised about roadside infrastructure, or V2I. Can you give some examples of that and who will pay for that?
Mr. McKenzie: An example of that would be an automated intersection where the traffic lights at the intersection have a camera on top of them or some type of object detection device that can know when a vehicle is coming up.
Currently, especially in Ontario, there are inroad sensors. If you pull too far forward, you miss the sensor and you can sit at a red light forever or if you've crept into the sidewalk or if you sit too far back. These will look down and they can factor in a more immediate change in response in any kind of weather conditions to automate the operation of the lights to control the vehicle. To your earlier point with respect to the environment, it reduces idling, maintains a more consistent flow of traffic and eases congestion as well.
That is just one. There are roadside infrastructure sensors as well, and there is a bit of a — not a debate on right now but a bit of a parallel challenge within the industry because technology is advancing so fast. You can get information from other vehicles, and the need for roadside infrastructure isn't quite as strong as it was five to seven years ago.
We've even progressed to the point now where, let's say, I'm driving ahead of you by two kilometres and I hit a patch of black ice and I'm only encountering slip on my traction control system on the passenger side wheels; it's only in half my lane. That information now, today, can be bundled up, put into a packet and conveyed to a vehicle going the opposite direction on the road, carried down until it catches up to you, unbeknownst to me as the driver or the driver in the opposite direction, and then that information gets put on your vehicle and allows, when you hit that point two kilometres down the road, your vehicle to be proactive with all of its on-board technology, such as traction control, instead of having to be reactive.
So there is huge potential with vehicle-to-vehicle and vehicle-to-infrastructure communications to improve the safety and operation of vehicles on the road.
Senator Eggleton: Will that be paid for by industry, by the manufacturers, including the roadside —
Mr. McKenzie: Yes. And the traffic lights deployed will be part of the upgrade when cities or municipalities renew their lights, and it will be at a lesser cost, so no additional cost.
[Translation]
Senator Cormier: Thank you for your presentation. Two things interest me: access to these autonomous vehicles, and the transfer of knowledge and skills in the new technologies. You told us about your mother at the beginning of your testimony. I am going to tell you about mine, who is 93 years old. She does not live in the Toronto-Waterloo corridor, or in Montreal; she lives in one of the most beautiful regions of Canada, northeastern New Brunswick, which is rural.
We have heard an awful lot of testimony and examples from urban areas. I wonder if you have thought through the issues of introducing these autonomous vehicles into rural areas. In terms of knowledge and skill transfer, you are from the University of Waterloo; so you are an expert in the subject. How can we be assured that the knowledge and skills will be transferred to the provinces? What role could the federal government play in its collaboration with the provinces, the industry and the universities to ensure that access to autonomous vehicles and the transfer of skills and knowledge in the new technology become a reality?
[English]
Mr. McKenzie: I'll speak briefly to the access to autonomous and the rural element of that.
I think there is strong opportunity in, if I could call it such, semirural. I think of my cousin who is a grain farmer on the Prairies in Saskatchewan. He's on a gravel road that you take two other gravel roads to get to. That's rural to me.
But in a semirural area, like your mother, for example, autonomous shuttles that have four- to six-person capacity and could operate in a pooled environment, maybe there is opportunity where every Tuesday we go to the grocery store or every Wednesday we go to the seniors' centre to play bridge, whatever those are. The ability to deploy in those rural or semirural environments I think will be significant because there aren't large issues of volumes of traffic or multiple roadways. It won't take a lot for an autonomous vehicle to learn.
There is a company from France, ACCA Technologies, that is in the process of setting up an engineering operation here in Canada. They are working on that kind of a four- to six-person shuttle, as are others within North America.
In terms of the transfer of skills and what the federal government could do with provinces to encourage universities and colleges, I come back to my point about our history of collaboration and working together. I think if everybody just keeps talking and asking questions like you are, those opportunities will present themselves.
For example, the University of Toronto and the University of Waterloo were announced last month to be two of the eight teams participating in an autonomous drive competition. It's a newly formed student competition for universities. There are three schools in Michigan, two in Ontario and in three other states — I think they are North Carolina, Texas and Virginia. We are already looking at that because of the interest it has generated, and that's in addition to the other student competition teams we have, some of which are in vehicles, offering a modified artificial intelligence course specific for application of what these students are going to be challenged to do in that competition.
As long as the dialogue continues, I think the opportunities will continue to present themselves.
[Translation]
Senator Cormier: Am I to understand the federal government should make sure that this conversation continues? Is that what you are saying?
[English]
Mr. McKenzie: Yes, I would say they could go a long way and they currently are. The Department of Innovation, Science and Economic Development, the Department of Global Affairs Canada and Transport Canada all have active roles, and we liaison with various branches in each of those, as well as Natural Resources Canada, as it relates to vehicles and the ecodeployment of vehicles, let alone connected and autonomous opportunities.
[Translation]
Senator Galvez: Mr. McKenzie, we can see today that, with vehicles like the Prius and these connected vehicles, vehicle automation is possible. Driving is possible because the infrastructures do not need to be changed. These vehicles drive like any others. However, according to the expert testimony that we have heard, my impression is that, when we move to the stage of complete automation, we will need infrastructure that is very different, very expensive and very intense in terms of information technology.
If it is only to transport one single person, my mother, for example, or a very small group of people, who is going to have to pay? Not them. How do you see that? Will it be a user-pay model? Will the industry have to pay? Who will have to pay to use the infrastructure, in your opinion?
[English]
Mr. McKenzie: The road infrastructure, at least in the next five years, doesn't need to be changed. From five years on, it needs to be enhanced, whether it's through vehicle-to-infrastructure communication capacities, or things within the roadside infrastructure that facilitate vehicle-to-vehicle communication. For example, there is the demonstration zone that's planned for Stratford. Stratford has ubiquitous free wi-fi throughout its downtown core. It was installed primarily for the benefit of tourists, but the plan in that demonstration zone is that they're going to utilize communication technology through the Internet, as opposed to utilizing cellphone technology, and antenna placement on cellphones has done a lot, as well as DSRC, for technologies that are currently deployed. So, it will be testing that next generation of communication, because the wi-fi that's ubiquitous in downtown Stratford is probably ahead of the curve on where society is going. Who's going to pay for that? Well, the City of Stratford is paying for that one, but infrastructure as it's factored into cellphones or communication technology with data plans is all going to improve. The infrastructure on roads will be enhanced. I don't think it will have to be altered or be a separate parallel investment to be purely autonomous.
To your point about who is going to pay, in the instance where people ride-share to play bridge at the seniors' centre, they're going to pay for that service no differently than they'd pay for a taxi or pay a little less to get on the bus, but they will probably be prepared to pay for the convenience of that shuttle coming to their door than having to walk out in inclement weather, wait for the bus and not get dropped exactly at the door.
Senator Galvez: Thank you.
The Chair: Mr. McKenzie, thank you very much for your presentation today. We hope to organize a visit to Waterloo next autumn.
Honourable senators, tomorrow we will hear representatives for the Global Automakers of Canada and Toyota. Next week we have three sessions. On Monday we're going to Kanata. On Tuesday morning, BlackBerry is coming to talk about what we saw the day before, so we will have stimulating questions. On Wednesday, we will have the Canadian Vehicle Manufacturers' Association, with Ford, GM and Chrysler coming in.
Tomorrow night, we will have to decide what to do the following week. The last week is becoming busy. Do we want to invite witnesses and not have time to see them or adjourn and not have the opportunity to see them? You all have caucuses today, so we'll have a better idea of where we'll be at the end of June. Thank you very much.
(The committee adjourned.)