Self-driven cars will face critical situations too, just like normal vehicles. Thus, the need of an ethics to guide them is on the rise

SHARE THIS STORY:

Choosing is often difficult. Especially when you only have a fraction of a second. And when your decision may affect greatly a human life. Car drivers experience this type of dilemma every day all over the world. With self-driving cars becoming an alternative option to regular vehicles, the problem must be extended from humans to car-controlling algorithms. But which decisions should cars make when an unexpected situation happens, and when harming at least one living being, such as a person or an animal, is unavoidable?

 

Self-driving cars can save human lives and help the enviroment, but may also have to take dramatic decisions

Self-driving cars have the potential to revolutionise the traffic of the future. They are expected to save human lives by reducing the number of accidents. Moreover, self-driving cars could help fight climate change via significant fuel saving, a result made possible by a more efficient drive compared to people.

 

But self-driving cars have neither the imagination nor the conscience possessed by human beings. They execute decisions based on what is written in their algorithms. For this reason, developers must decide which instructions should be given to cars in case unexpected and dramatic events had to occur. These include multiple obstacles that are suddenly on the way, and the unavoidable decision of hitting either a child or an adult, a man or a woman, a dog or a cat.

 

The virtual car used for the experiment at the Osnabrück Institute of Cognitive Science (source: Frontiers in Behavioral Neuroscience)

One possibility is to make cars behave like the majority of people would do. A team of scientists of the Institute of Cognitive Science at the University of Osnabrück in Germany investigated this approach and published the results in Frontiers in Behavioral Neuroscience. In their experiment, they asked a number of participants to drive a virtual car in a simulated environment, and to choose whom to spare or hit. The outcome from these experiments provide hints on a possible way to develop self-driving cars’ current programmes.

 

Nevertheless, the debate on the best strategy to adopt will be long. After all, ethics-related issues have always been controversial. And we all know it: choosing is often difficult.

 

Cover image by Automobile Italia