Major automakers as well as giants in the technology industry (eg Google) are prepares cars with Full Self Driving systems. The first such vehicles have already appeared, but engineers and researchers are working hard to improve as much as possible the intelligence but also the capabilities of these systems to drive them as much as possible right and above all safe for everyone.
Both for the passengers of the self driving vehicle and for the other vehicles on the same road as well as pedestrians. In addition to the various technical issues that need to be solved for the creation and operation of Full Self Driving vehicles, a theme has also emerged as the “trolley problem”. This is a matter that moves in the sphere of morality.
The “trolley problem” was formulated in 1978 and is a moral dilemma. According to this, five people are in danger of being killed because a trolley is heading over them. People can be saved if the driver pulls a handbrake and the trolley goes into the side street. On the sidewalk there is a man – if the trolley goes to there, the man will be killed. The dilemma is therefore whether the driver chooses to kill one or five people.
This problem should also be answered in the case of self driving cars that will be challenged at some point. Let’s look at a few examples:
1) A self driving car carrying passengers is sure to hit pedestrians if it does not change direction – but if it is sure it will fall on a wall or a pillar and will probably injure or kill the passengers of.
2) A self driving car without passengers will come out of its course and either fall on a pair of elderly people or on a young child.
3) A self driving car comes out of its course and will either turn on another car or a motorcycle or fall into a cliff.
What decision should the car take in every case? Is there a moral answer to the question of what the vehicle should do when it comes to the possibility of a road accident?
Read more :
- Elon Musk doubts about self-driving Lidar sensors utility
- Α boom in Full Self Driving cars sales by 2030 expected
- Kubota reveals autonomous robot tractor of the future
The philosophical approach.
The answer, as many say, can not only come from engineers or experts in the field of artificial intelligence, but other specialties, like the philosophers, must be involved. As quoted by qz.com, a group of philosophers headed by Nicolas Evans, a professor of philosophy at the American University, Mass Lowell, develops algorithms.
The team believes that these algorithms will allow the stand-alone vehicle to respond properly to a number of possible scenarios it can face on the road. The team has received $ 560,000 funding from the US National Research Foundation to create various kinds of “trolley problems” and to show that it can respond to the moral dilemmas that arise each time in a stand-alone car.
For researchers to do it, they try to turn moral theories into a language understood by computers. The philosophers of the so-called “utilitarian school” believe that all lives have the same moral weight, and so an algorithm based on this theory will be programmed to give the same value to the lives of the passengers of the vehicle it controls as well as to life pedestrians.
There are many who believe that it is everyone’s duty to protect himself and so when an autonomous car is not programmed to cause willful damage to other people, it is acceptable to try the vehicle to do whatever it can to avoid a collision with another vehicle or going off the road to protect his passengers – even if the actions that the vehicle will do endanger other people.
The group of philosophers who develops the algorithms for the time being does not take a public position on which moral theory is the right one. He hopes that he will not have to decide which option the car will have but will create a framework where decisions on what the vehicle will do in every difficult case will be taken either by the manufacturer or the manufacturer same customer.
Researchers do not yet collaborate with any automotive industry and they intend first to complete a series of tests before they reach a company. Evans and his team are waiting to see the results of the decisions the vehicles will take in the different scenarios that were designed. Researchers are waiting to see if a moral theory will lead to saving more lives than another, or if the results will be more complicated.
“It’s not just the numerical issue, that is, how many people have died, and who were chosen to die and who to live. It is possible that two scenarios will result in the same number of deaths but not in the same individuals. The difference between Scenario A and Scenario B may be that the first one to die over 50, while in the second scenario, those losing their lives are under the age of 30. We must therefore start a debate in society not only about the level of risks we will accept from the use of stand-alone vehicles but also about which people we accept to put at risk. If some moral theories state that passengers of an autonomous vehicle need to be saved while other pedestrians, there should be a debate on what option the autonomous vehicles will take. There should also be a new debate on the structure of the road network and the possibility of a greater distance from pedestrians, ” Evans said.
Unexpected risks and applications
One more serious issue for the philosophers team is whether the software of self driving vehicles will be completely safe or will be able to take control of these vehicles by some capable hackers who will be able to “tease” the algorithms . There are also technical hazards that need to be discussed, such as the case where an infrared laser is used to interfere with various system sensors, and so the vehicle comes out of its way and cause a traffic accident.
There is also the issue of communicating autonomous vehicles with each other and whether they will interact smoothly, especially if they operate with different artificial intelligence systems. There may be positive aspects with the generalized use of stand-alone vehicles. It can, for example, significantly reduce road accidents, especially those that lead to serious injuries and deaths.