Life and death decisions: the ethics dilemma behind autonomous cars

By dpa | 13 October 2016


People looking towards an autonomous self-driving vehicle, as it is tested in a pedestrianised zone near London, on Tuesday. - AFP


MUNICH: An obstacle suddenly appears before a car. The car needs to swerve to avoid it, but to the left of the obstacle there’s a woman pushing a pram, and to the right there’s an old man.

Someone - or something - must decide whether to slam on the brakes or to steer to safety, potentially killing a pedestrian.

It’s an old ethical thought experiment, but it’s one that is increasingly becoming a practical decision for manufacturers of driverless cars.

Automobile executives like to talk about these kinds of scenarios, which bring up a sensitive question: Which decisions can computer-controlled vehicles take and which ones would we rather they don’t?

In Germany, where the major car industry is increasingly looking towards driverless cars, politicians are hoping to set ethical rules for just this issue.

file6ruyo8jywd11e1x7hegp


In the end, there can be no easy answers – and this will be made clear at the first meeting of an expert commission set up by Germany’s Minister of Transport, Alexander Dobrindt.

A key question will be just how much automation a society actually wants, says ex-constitutional judge Udo Di Fabio, who heads the panel, as well as the extent to which the possibilities for “human intervention” will be maintained.

The minister has set two central principles for scientists and association representatives: property damage takes precedence over personal injury, and there is to be no “classification” of people.

In doing so, the German government is already working on a legal framework for robot cars, which would allow drivers to turn away from the wheel to read newspapers, write e-mails or watch movies.

However, drivers should still remain “perceptive” in order to intervene whenever the system requests or requires it. Liability regulations must also be clarified as cars themselves increasingly take control.

Particularly in tricky or delicate situations, Di Fabio expects major moral dilemmas could arise. For example, what ethical priorities could be programmed into computer software “if you only have the choice between different kinds of accidents?”

As a constitutional lawyer, Di Fabio makes no secret of the fact that some debates are, considering the central principle of human dignity, simply “outlandish.”

Would a vehicle decide to crash into a homeless person rather than someone else. Would a car evaluate people according to their social utility? “I do not believe the ethics commission will seriously be able to discuss this,” he says.

Germany’s Constitutional Court had come to a similar decision at the beginning of the year, when the discussion centred around the possibility of shooting airplanes out of the air if they had been hijacked by terrorists for the purposes of an attack. This led to the conclusion that life should not be weighed against life.

Manufacturers are also bracing themselves for potentially fatal situations.

“The dilemma will come,” says Daimler Board of Management member Thomas Weber, on the sidelines of the Paris Motor Show. “That’s why it’s good to deal with these questions as early as possible.”

Companies still express mixed opinions when it comes to the question of who decides in matters of life and death – the only point of agreement is that autonomous driving systems can indeed avoid accidents.

In a white paper drafted last year, Daimler tried to find answers to these questions, but no clear statements resulted. “I am nevertheless sure that we can find solutions,” says Weber.

For his part, BMW Chief of Sales Ian Robertson appears keen to shelve the issue for the time being: “We believe that the driver’s responsibility will remain the fundamental principle for a long time yet,” he says in Paris.

“We also believe that the technology is still somewhat immature for any next steps to be taken.”

Even so, within the industry, the question of who assumes responsibility for the new technology is not up for discussion: from manufacturers’ point of view, it is clear that this can not be delegated to a separate supplier, says Gilles Le Borgne, Vice President of Research and Development for Groupe PSA.

As the German transport minister explains, this discussion is about ethical guidelines for systems that may start to be used in five years or later.

Another factor to be taken into account is that there could be fewer deaths on the streets if human error can be avoided. “This is a great opportunity,” says commission head Di Fabio, but the personal rights of every human being remain a key consideration.

And for especially delicate situations, the overarching question might be even more straightforward: “Should a technology that leads toward such a dilemma be allowed at all?”

Keywords