THE STANDING SENATE COMMITTEE ON TRANSPORTATION AND COMMUNICATIONS

The Standing Senate Committee on Transport and Communications meets today at 9:30 a.m. to continue its study of technical and regulatory issues related to the arrival connected and automated vehicles.

Senator Dennis Dawson (Chair) is in the chair.

The Chairman: Honorable senators, the committee is continuing its study this morning on the technical and regulatory issues related to the arrival of connected vehicles and automated.

We have with us this morning Denis Gingras, who is a professor at the Laboratory of Vehicle Intelligence at the University of Sherbrooke, as well as Tony Zhijun Qiu, who is also a professor, but at the Faculty of Engineering of the University of Alberta. Last year, when the committee members went to visit the University of Alberta, it was Mr. Qiu who welcomed them. I wanted to thank you from the bottom of my heart.

I welcome you. I invite Mr. Gingras to make his presentation, and then I will give the floor to Mr. Qiu.

Denis Gingras, Professor, Vehicular Intelligence Laboratory, University of Sherbrooke, as an individual: Honorable senators, I would first like to thank you for inviting me and giving me the opportunity to express my opinion on this subject.

As you know, mobility is extremely important and fundamental for the development of our society and for our civilisation. Everyone will agree on that. However, paradoxically, the road transport system as we know it today is quite inefficient, and this is a bit surprising. But if we examine the various parameters, we realize that it is an extremely inefficient system. Let us first think of the business model and the economic system which is based on the individual ownership of vehicles; there are definitely too many vehicles on the roads. I think trucking agencies all agree on this. Then there is population growth and urban concentration. This therefore results in an overload and overcapacity of our road infrastructures.

In addition, freight transport has shifted from rail to road infrastructure to meet “just in time” demand using trucks and road trains, resulting in effect of increasing the overload of our road infrastructures.

Vehicles and cars are always single occupancy up to about 80 percent, so the payload in a vehicle is on the order of 5 percent. 100 compared to the total weight of a vehicle. It is thus not really effective. The vehicles are used on average one hour per day; 23 hours a day they are stationed. They are useless and they rust. Vehicles are still equipped with combustion engines that run on gasoline and fossil fuels and have an efficiency of around 25 to 35 percent. 100. It’s really not good. Also, the vehicles are still driven by human beings who are the cause of about 90% of the accidents. 100 of the accidents on our roads.

For 100 years, since the invention of the automobile, incremental improvements have been made, but still today, for the majority © of the car park, the cars are still quite unconscious, stupid and blind. Fortunately, we have been experiencing a technological revolution for about two decades; I have worked in the field for 35 years.

I would like to talk to you about three axes. The first relates to the electrification of propulsion systems. This essentially affects the ecological aspect, but it does not change the paradigm of our terrestrial mobility. The second axis concerns driving automation, and the third major axis concerns connectivity.

In terms of technological developments, we have the opportunity for these three main axes to develop new solutions that will hopefully improve the efficiency of our ground mobility. . We would like driving automation to ultimately lead to fully autonomous vehicles in order to be able to eventually migrate from an individually owned system to a mobility service model. ©. This allowed us to significantly reduce the number of vehicles on our roads.

Then, for road safety reasons, it would eliminate the driver from the equation and thus reduce the number of collisions and fatalities.

Major impacts will occur, not only on the technological level, but also with regard to insurance companies, legislation, the after-sales market, maintenance, as well as for daily living habits, such as the need to obtain a driver’s license, vehicle registration and so on. This will significantly change our lifestyles.

It will also promote the “last mile solution” which helps to promote better integration of the different modes of transport and improve the efficiency of the interface of the systems. different transport souls which, currently, work rather in silos and in autarky one with respect to the others.

To achieve fully autonomous vehicles, however, there are enormous challenges to overcome that are far from being realized. Safety with an autonomous vehicle is essentially based on two main parameters: robustness and reliability. They are two totally different things. In terms of robustness, an autonomous vehicle is a robot on the road and it operates in an open environment, not only because of mobility. Typically, for a vehicle traveling at 100 kilometers per hour, the vehicle’s environment changes every three seconds. This is already huge and therefore requires real-time applications and systems that are very efficient, and have very fast response times.

Moreover, there are a large number of parameters like weather, traffic conditions and road conditions. All of this leads to many driving scenarios that not all of us can foresee or consider. You can’t program an autonomous vehicle the same way you program a traditional system as we know it today. This requires an extremely robust system to deal with all these variants.

Then comes the complexity of the systems. Typically, a vehicle today contains over 40,000 parts and components. There are more than 100 million lines of code and a hundred microprocessors. This then requires validation and testing methods that are once again completely different from what is usually done. This requires techniques based on so-called stochastic algorithms, sampling methods, a bit like those found in large industries such as microelectronics, for example. We cannot make a deterministic and systematic validation of all the possibilities of the system. You have to go by sampling and, at best, what you get is an idea of ??the performance of the system on a probabilistic basis.

This brings us to a major challenge, namely artificial intelligence. Over the past year and a half, you may have noticed that major automakers are investing billions and billions of dollars specifically in the area of ??artificial intelligence. This is not a surprise, because the most important technological challenges relate to this aspect. We are far from having robots that have cognitive capacities, capacities of perception and consciousness, as well as decision-making capacities equivalent to those of a human being. Computers are great at speed, logic, and combinatorics, but humans can do things that machines still can’t. The proof is that, when a human being is born, it takes years for him to develop his motor capacities, his cognitive capacities, his capacities of perception, of awareness of oneself and one’s environment. It is for this reason that a driving license is given only to a young adult who is at least 16 years old; you don’t give a driver’s license to a 5-year-old child.

Here are some observations regarding connectivity. Connectivity basically enables the exchange of information. This exchange of information can be used at two levels: firstly to improve the redundancy of information available in intelligent vehicles and, secondly, to provide complementary information to that which is available. in smart vehicles. Redundancy is important because it allows us to validate information that we already have, to increase our level of confidence in this data, and therefore to improve the probability of good results. decision-making when any driving situation arises, for example when it comes to avoiding a collision or an obstacle.

The complementarity of information allows to have an extended perception, namely that certain information can be obtained by means of communications which go beyond the distance at which our own sensors embarked in our vehicle can perceive things. Typically, the sensors embedded in a vehicle have a perception of a distance of approximately 100 meters or 200 meters at best. With telecommunications, we can get information which comes from sources which are at much greater distances.

Communication makes it possible to envisage and design mobility solutions where intelligence would be distributed and no longer concentrated only in vehicles. We can, at this time, exchange information between road infrastructures and vehicles. For example, if a city has intersections where the risk of accidents is higher than in other places, we can exploit these intersections, make them intelligent and, suddenly, give people a chance. vehicles, sometimes a little more stupid, to obtain the information and take appropriate decisions to avoid accidents.

Here are some of the recommendations that are offered: first of all, I believe that we need to develop strategies, in Canada as elsewhere, because this is an international problem and not just Canadian. With the participation of governments, automobile manufacturers, road transport agencies and research institutions, a comprehensive development and deployment strategy must be put in place, akin to the concept of cities. smart ©s.

On the economic level, the merger of the automotive sector with information technologies opens the door to numerous business opportunities for Canadian companies, for the development of a new generation. generation of companies, start-ups, SMEs, which will be multidisciplinary in nature and which will combine information technologies with the context of mobility.

Finally, as a teacher, I believe that the curriculum of our education systems must be the subject of a major reform at the college, professional and university levels. , so that we can train a new generation of professionals and a highly qualified workforce who will be able to face the challenges of complex systems and multidisciplinary problems .

It’s your turn to share your insights with us, Mr. Qiu.

Tony Zhijun Qiu, Professor, Faculty of Engineering, University of Alberta, as an individual: Thank you for inviting me. I am very happy to be here to present our work to you and give you our opinion on connected and automated vehicles.

I would like to start by defining, technically speaking, what connected and automated vehicles are. In the case of automated vehicles, it could not be clearer: it is about vehicle technology. When it comes to connected vehicles, there is sometimes some confusion, because we don’t always know what we’re talking about exactly.

To me, connected vehicles are like trying to digitize the infrastructure to make it compatible with technological changes in vehicles.

Until now, we were talking more about the integration of vehicle technologies. In my opinion, this is an apt appellation, because there are different reasons for wanting to convert a trendy vehicle.

We can see how it is possible to use ICT, or information and communication technologies, to digitize the infrastructure. In a connected vehicle, the first step is to apply ICT to make it intelligent.

Last September, when you came to visit my lab, we spent about forty minutes observing a demonstration vehicle in Edmonton. As I anticipated, the Connected Automated Vehicle sought to create an integrated, open, data-driven platform for a future transportation system . For example, some municipalities find themselves at an impasse, because the municipal council receives several proposals — public transit, fire trucks, vehicles for winter maintenance – but treats them separately. Also, because there is a lot of talk about infrastructure.

When I talk to officials, they always wonder why these things are so contested. They are unable to integrate them.

That’s what connected vehicles do. Then, we must digitize the infrastructure and essentialize the management so that the different vehicles are all connected to the same platform. That’s why this platform is said to be integrated, open and data-driven. Connected vehicles also allow us to make efficient use of current infrastructure. We can come back to that during question period.

Automated and connected vehicles also make it possible to use the technological boom and the automotive industry as leverage.

Certainly, they aim to reduce the errors committed by human beings.

In Edmonton, since 2012, the University of Alberta, the University of British Columbia and the City of Edmonton have joined forces. We installed 30 roadside equipment and a number of vehicles to develop this technology and see how it could help federal and provincial governments better manage traffic.

If you weren’t here last year, come see us the next time you’re in Edmonton, and we’ll show you everything we do.

I will end by saying that connected and automated vehicles must be adapted to the Canadian reality. The cold climate is obviously a huge factor to consider. I don’t know how other countries do it. In Edmonton, we tested our prototypes at -30 degrees. We never got to -40, but last year we did some testing at -30, which gave us great data for future policy and related studies . Winter is long in Canada, so the conditions here are unique.

Traffic is mostly urban in Canada compared to other developed countries, which creates situations not seen anywhere else.

I would like to finish by talking about investments in public transit infrastructure. Canada spends a lot of money on roads, but it still has a long way to go to use infrastructure more efficiently. Automated and connected vehicles, for example, could help increase capacity. Right now it’s 1,800, but connected, automated vehicles could push it up to 2,100 and even 2,400. We need to spend our money more wisely.

In terms of next steps, connected vehicle regulations are obviously at the top of the list. We also need a policy on radio frequencies, with the approval of the Canadian Standards Association, because our infrastructure is now digital. We have to get permission from the association.

.

new cars use in-car computers to make driving safer.