The autonomous car will have, above all, to avoid personal injuryAutomotive News
Stand-alone truck that ‘works’ in Colorado (USA). It protects the operators that circulate in another that goes ahead.
Germany develops the first ethical code that regulates its response in case of conflict
Imagine the scene: 2040, an autonomous car is involved in an inevitable accident at a pedestrian crossing. Why did the car run over that person and not the one next to him? What is the basis of the machine to draw a maneuver that can be lethal? And above all: Who carries responsibility?
The situation is still a long way off, but in Germany, a nation with a strong spirit of foresight, it is already a subject raised by the Government, which has just presented an ethical code for autonomous vehicles that seeks to shed light on these dilemmas.
It is a pioneering document in the world and has been drafted by 14 experts in subjects such as ethics, law and technology. They are not binding, but consists of 20 points that try to establish the framework on which the cars should be programmed without driver.
A few months ago, the law has been drafted that allows its circulation and the motorist, for example, occasionally withdraw hands from the steering wheel. But always attentive and willing to take control.
The text should help to solve complex situations, such as in the case of imminent and inevitable outrage. How should the machine operate in case of unavoidable personal injury? Experts insist on not comparing potential victims, ie that no preferences are established for characteristics such as age, sex, or physical or mental condition.
The most important thing, they say, is that this situation is not reached. The report finds that autonomous driving only makes sense if it causes fewer accidents. And in the case of one that can not be avoided, material damages or animals will always be preferable to people.
To do this, the document requires all driving data to be stored – although they can only be used if the driver gives permission – and emphasizes that it should always be clear who is responsible for driving in the vehicle at all times. Something that for the Mexican Raúl Rojas, professor of artificial intelligence of the Free University of Berlin, does not raise doubts: the manufacturer of the product must assume the commitment of security.
“The autonomous car must be insured by the manufacturer”, he assured EL MUNDO. Rojas is at the head of a laureate team of scientists developing autonomous cars since 2006. Under his command, the Free University of Berlin has developed several pioneering prototypes that have been examined with magnifying glass by the industry. Its latest creation, the AutoNOMOS was tested in Germany, Mexico and the USA, becoming the first autonomous vehicle with experience in three countries with very different driving conditions.
The professor emphasizes that today the driver is the one who has the responsibility, but in an autonomous model it is the product that responds to any problem. Thus, this car will have insurance that protects the manufacturer company but not necessarily the driver, who will simply be a passenger, says Rojas.
Experts insist that the computer can not decide for itself. It will be programmed to stick to the laws and not to interpret them. “It seems trivial, but a lot of people are asking if the computer can decide on its own. What stands out from this ethical code is that computers will not have that capability. ”
PABLO BARBERO – Berlín -El Mundo – 25 AGO. 2017 (Translation Soft)