Andersen

Can Self-Driving Cars Drive Better Than We Do?

Feb 21, 2020
Blog

The global urban transport system, where each car is driven by AI, is not only possible but can be implemented in the next 10 years. Our company’s Analysts have developed the concept of such a system, and now we can tell you how it will look and what is needed to implement the project.

Self-driving cars have not become a common sight on the roads yet, but the way things are going, this will happen in the coming years. But how beneficial will this innovation be? Transport companies and taxi services will surely be able to reduce costs, and the number of traffic accidents might decrease. But the potential capacity of this technology is much broader. What if we approach the issue globally and not simply open roads for self-driving cars but make all cars on the roads controlled by AI? How realizable is it and is it worth it? Our Business Analyst Anton Naidenko presented his own concept of organizing the movement of vehicles under the control of Artificial Intelligence.

Why have all transport under AI control?

If we consider a self-driving car from the perspective of a particular user or a company, we will see a few simple advantages. For a company, the driverless car allows for savings on driver salaries and improved transportation efficiency. For ordinary users, the driverless car gives the opportunity to do what they want during the trip, even eat or sleep. But while waiting in a traffic jam, the driver of a regular car and the passenger of a self-driving car are in the same situation: their hands are free, but they lose time instead of moving to their destination. 

Reduced number of traffic jams and waiting time in them is one of the advantages that a single AI-controlled transport system can give to society. All cars on the road will constantly exchange data, passing it to traffic lights as well, and AI will be able to control the traffic. In some cases, the total number of cars on the road will decrease, as one vehicle will be enough for an average family. One car will be able to drive children to their schools and parents to their offices, then, at lunchtime, go back for the children, and in the evening – for parents. The results are conserving resources and environmental benefits. And perhaps the most important point is a significant increase in traffic safety. If there are only cars that receive signals from a single system, such incidents as red-light running, drunk driving, bald tires in the winter, driving without a license, and so on will disappear.

If every single car is AI-controlled, the human factor will be completely excluded from traffic. Pedestrians will remain, of course, but it will be enough for them to follow the traffic rules. Surely, there are other advantages, which are not obvious now. But the concept has drawbacks too: a danger of system errors in AI and hacker attacks. And this is one of the reasons why the implementation of such a large-scale project could face some difficulties. 

When will we be able to sleep while driving?

According to James Peng, CEO of Pony.ai, fully self-driving cars will appear on the roads within 5 years. The organization of a global AI car management system will take some more time. At the same time, the shifting of all car owners to autopilot cars won’t happen at once, of course. In the beginning, someone will abandon this idea for financial reasons, and someone will assert their right to drive their own vehicle out of principle. However, the full implementation of the system requires every single car to be AI-controlled and connected to this system. Until all car owners agree to such conditions, integration, unfortunately, is not possible.

Despite the fact that, on a global level, self-driving cars will be safer, avoiding accidents that involve these cars is impossible. And then, the discussion of this topic in society will reach its peak, and debates in talk shows, demonstrations, and other signs of the opening Overton window will begin. And after some time, society is most likely to accept the inevitable risks of such a system (provided that globally, there really will be fewer accidents and the system itself will be more convenient and cheaper). And enthusiastic drivers will be seen as savages who are ready to risk their lives and health, as well as the lives and health of others, because of a simple whim.

Who is capable of implementing such a project?

There are at least two options for implementing such a project. The first option is if the initiative comes from the municipality. The system can be implemented, to begin with, in one city, with parking lots and car exchanges for the city visitors. The disadvantage of the system is that manufacturers will have to adapt their cars specifically for integration with the city control system. Not every manufacturer will agree to do so for the sake of one city only, even though such radical changes are inevitable progress, like the transition from carriages and cobblestones to cars and highways. As a result, for the first time, the city residents will have a very narrow choice of cars and may not agree to such a situation.

On the other hand, automakers themselves could be engaged in the development of such a system. This is exactly the case when it’s beneficial for companies to unite and develop a system using shared resources. If several companies (ideally, not direct competitors) participate in the project, this will give users the opportunity to choose, and there will be more chances that residents of one particular city will agree to the launch of this system. At the same time, the support of the municipality will be vital, as traffic lights and other traffic controllers will be included in the system. But the initiative will still come from automakers.

The second option looks more realistic because it’s beneficial to everyone. Except for automakers, perhaps, who won’t cooperate and whose cars, as a result, won’t be represented in this city. Although, in the future, they will have the opportunity to connect to the system – making a certain monetary contribution will be enough. Such cooperation will be necessary at the stage when AI systems appear in all major cities, but they will be created by the joint efforts of different companies. Mobile operators started working according to a similar scheme.

It’s hard to say exactly when such a system will be introduced. People still need to get used to self-driving cars in general, and car manufacturers need to see this technology as a priority. The most important thing is, it’s already clear that implementing this project is quite possible, and everything rests on the acceptance of the idea by society. And whether the roads will become safer in 5 years or in 50 depends on us.

Previous articleNext article