The technology of self-driving vehicles has become a part of our everyday life, we can always hear about how they are going to revolutionize our world. It is clear that during everyday traffic, the lives of others are in our hands and vice versa. But how can we deal with the fact that instead of people, machines are making decisions too?

Lately, self-driving cars have been in the limelight, especially now that major players such as Uber and Tesla were involved in accidents. While there was no personal injury in the Tesla case, the accident involving Uber resulted in death.

After hearing such news, probably many will have doubts about the reliability of the technology saying they do not want to find themselves on a testing road on the street. Usually, the news covers the frequency of road accidents; based on Hungarian Central Statistical Office data, hundreds of people die on Hungarian roads every year (this number is 1.3 million in the USA, which makes it one of the leading causes of death). At the same time, we can see that cases like Uber’s evoke entirely different emotions from people since a developing pioneer technology is involved.

According to experts, the technology on its own is absolutely promising, even though it still requires several fine-tuning and development. Basically, fully self-driving cars are not even just around the corner; all we can say is that cars can take the wheel but they are only partially automatized and they do require human intervention. An advancement could be where people do not even have to pay attention passively to the moving car.

However, the positive economic impacts of self-driving cars can easily be identified because traffic can become cheaper, costs resulting from accidents can be cut back, and moreover, time can be saved.


The technology could eliminate a major part of road accidents caused by human negligence; many say that road traffic can become safer in general by self-driving cars. However, a lot of attention is given to cases that pose ethical challenges too. For example, how should a self-driving car approach an intersection? What should it do in a situation where it seems that human injury cannot be avoided? Whose life the car should save, the passengers’ or the others’?

Similar moral problems in philosophy and psychology are simply referred to as the ‘trolley problem’. In the example, the question is whether the driver of a runaway vehicle would be right to sacrifice the life of a passer-by to save five others. A similar question is whether the government should spend more on the diabetes prevention programme or on the public work programme. For such questions, there really is no adequate answer, this is why the situation of self-driving cars is extremely complex.

Of course, developers put a lot of effort into refining the technology, for example, sensors can be further advanced. These cars perceive their surroundings significantly better than humans do thanks to LIDAR technology and other sensors, basically, they can fully map whatever is around them.  Thanks to the sophisticated artificial intelligence, cars are learning while on the road, and even extreme situations and human decision-making can also be modelled in certain instances. If we could advance the technology of self-driving cars even further, it is possible that the chances of errors can be drastically decreased. This way, road transport can become remarkably safer than it is today.

However, it is likely that it will not stop there. A fundamental change of the transport system is also a possibility if we manage to develop cars that are actually self-driving. The idea of smart cities has been getting a lot of attention lately, and as for self-driving cars, in such an ecosystem, every vehicle has to be connected to each other. A system like a new set of traffic rules that takes self-driving cars into consideration could be developed since the current one is based on human drivers exclusively. For example, by changing pedestrian traffic in cities could eliminate numerous risks and maybe traffic lights could also become unnecessary.

Original date of Hungarian publication: 10 April 2018