Audi study on autonomous driving: ethical aspects

A central question about autonomous driving asks: How can people trust machines? The Audi “SocAIty” takes a closer look. In addition to the technological and legal challenges, ethics are one of the focuses of the study.

04/27/2022 Reading Time: 4 min

Graphic collage of concept car interior, driver and digital elements

People will learn to trust machines

Distrust of the unknown is not an insurmountable obstacle. There was widespread consensus on this among experts involved in the “ SocAIty” study. Indeed, distrust has always been a factor whenever new technologies dawned in the human experience. Yet elevators, airplanes, and trains have all become commonplace today.  

Can a driverless car make the right decision in a hazardous situation?

  

Side view through the open doors into the interior of the Audi urbansphere concept

Audi urbansphere concept: The vehicle shown is a concept vehicle that is not available as a production vehicle.

Many people are troubled by the question as to whether a machine can make the right choice in a hazardous situation. But autonomous driving is not the first time this question arose. In fact, it has been a matter of discussion in ethics for decades, as illustrated in the “trolley problem”. This thought experiment asks us to imagine a situation in which one individual could divert a runaway trolley onto a side track where one person lies motionless, thereby saving the lives of five people tied up on the original track. Would this be a criminal act? Should the person rather not act at all? Or did the individual deliberate correctly and act to mitigate the greatest possible harm?

 

With autonomous driving, this discussion has witnessed a resurgence: But experts say the central point of the debate is that a self-driving car would not make its own decision in a hazardous situation, but instead only reflect the software choices its creators endowed it with. It can and will only assume the ethical decisions and values of the people who design it – and apply them without its own interpretation.

In hazardous situations, protecting human life is top priority

The German Federal Ethics Commission began grappling with these questions back in 2017. In its first report, the Commission laid out initial guidelines, but also identified a demand for action and further development in the areas of technology and society. The result of the report are 20 “Ethical rules for automated and connected vehicular traffic”. 

 

One such rule is that self-driving vehicles are only justifiable if they promise a reduction in harm compared with human driving. So accident prevention is the name of the game for the further development of automated driving technology. And in hazardous situations, protecting human life enjoys top priority.  

 

Under no circumstances should the software make any distinction based on immutable characteristics such as age, gender, or physical or mental constitution. Another basic rule: It is prohibited to offset one victim’s life against another. Many experts believe that rules like these resolve hypothetical dilemmas such as the trolley problem. They therefore call for the discussion to turn to questions of practical ethics.  

Side view of the Audi urbansphere concept against a futuristic backdrop

Audi urbansphere concept: The vehicle shown is a concept vehicle that is not available as a production vehicle.

Audi urbansphere concept: The vehicle shown is a concept vehicle that is not available as a production vehicle.

AI4People: An EU Parliament initiative

The European Parliament has also taken up the issue. In 2018, it launched “AI4People”, an initiative advancing ethical standards in artificial intelligence (AI). The initiative was set up to help work out the principles, guidelines, and practices necessary to establish a “good AI society” as well as to make related concrete suggestions for companies and the economy at large. 

 

So far the conclusion is clear: Protecting human life is the greatest good. In keeping with this, autonomous vehicles are only ethically justifiable if they lead to fewer injuries and fatalities compared with human driving – in other words, a positive balance of risks.  

 

The fact is that, according to the German Federal Statistical Office, more than 85 percent of all accidents resulting in personal injury in Germany in 2020 were caused by human error while driving. And the WHO reports that worldwide, one person dies on the streets every 24 seconds. Against this backdrop, the experts in the Audi study see in an autonomous future a new kind of reliability and road safety.

“In automated vehicles, the computer is always awake”

  

View through the panoramic roof into the brightly lit interior of the Audi urbansphere concept.

Audi urbansphere concept: The vehicle shown is a concept vehicle that is not available as a production vehicle.

The answer to the question of whether machines (or algorithms) can drive cars more safely than people is therefore: Yes, under certain circumstances. In a familiar environment with clearly defined parameters, the technology is highly reliable. “In automated vehicles, the computer is always awake, the system never stops working,” says Director of Smart Mobility at Fraunhofer FOKUS Ilja Radusch in the study. 

 

And Oliver Hoffmann, Member of the Audi Board of Management for Technical Development, agrees: “I am convinced that highly automated driving will make our streets safer. On the one hand due to the advanced sensor technology that an automated vehicle uses. In this context, Audi relies on a variety of different systems such as radar, camera, and lidar. This means driving situations are assessed with significantly greater accuracy, leading to automated warning, braking, and avoidance functions that adapt situation by situation. On the other hand, we use V2X (vehicle-to-everything) technology – this means connected driving with car-to-car as well as car-to-environment communication, which also includes other road users such as pedestrians and cyclists.”

Zero tolerance for errors: Not a 100-percent possibility in mixed traffic

But is a zero tolerance for errors, the “Vision Zero” called for by some, even possible? Experts in the study suggest it will never be 100 percent feasible. Even with autonomous driving, people will remain the greatest factor of uncertainty.

 

One challenge the study identifies for the coming years is “mixed traffic” – that is, traffic characterized by both self-driving and traditional vehicles. While vehicle safety will continue to improve, new types of accidents might occur in mixed traffic, such as when human-driven vehicles go over the speed limit. Programming autonomous cars to factor in such human risk factors and to react in the interests of all road users’ safety remains a technical challenge.

“Self-driving experiences” might convince skeptics

But how to convince the intractable skeptics? The study concludes that one of the most important means of persuasion is to demonstrate the advantages and personal benefits of self-driving cars, such as the time saved or the increased comfort. There might also be enormous potential for societal acceptance in other areas, such as in inclusive mobility for persons with disabilities. One concrete way of introducing users to the technology today are so-called “self-driving experiences”. Examples include trials of people-mover concepts in in German cities like Berlin, Hamburg, or Karlsruhe and testing the Volkswagen Group’s automated valet parking.

The “SocAIty” study 2021: The most important insights
The “SocAIty” study 2021: The most important insights

Initiatives 12/02/2021

The “SocAIty” study 2021: The most important insights

From the appropriate legal framework to ethical questions to digital responsibility: the 2021 “SocAIty” study from the &Audi Initiative examines the overall societal dimension of autonomous driving.

Read more

Initiative &Audi

The Audi initiative for a responsable use of new technological developments such as autonomous driving and artificial intelligence.

Close

Audi worldwide

Models, products and services – switch to your country / sales region website and discover the regional diversity of Audi.

    Close

    Close

    You are using an unsupported browser to access this website. To get the best User-Experience while visiting the website, please use the latest versions of Chrome, Firefox, Safari or Edge.

    Thank you very much for your visit.
    Your Audi Team

    The international Audi website

    Discover Audi as a brand, company and employer on our international website. Experience our vision of mobility and let yourself be inspired.

    Audi of America: models, products and services

    Explore the full lineup of SUVs, sedans, e-tron models & more. Build your own, search inventory and explore current special offers.

    Switch to audiusa.com

    Error report

    The feedback form is currently unavailable.
    Please try again later.

    You may deactivate your ad blocker to view the feedback form.