As things stand today, fully autonomous cars—vehicles with an autopilot system that can drive without human intervention—are likely to occupy an increasingly important niche in global transport by the middle of this century. This is largely due to their greater environmental sustainability, reflected in a lower carbon footprint than conventional vehicles. This is partly because driverless vehicles are expected to be 100% electric, given the critical role of electronics in their operation. But also because autonomous vehicles drive more smoothly than many humans—minimising braking and acceleration—which translates into greater energy efficiency and lower consumption. In fact, the US Department of Energy estimates that the smooth driving characteristics of autonomous vehicles reduce energy consumption by 15-20% compared to human driving.
In addition, thanks to the constant information it receives from other computers, a driverless vehicle can choose the best route with the lowest energy consumption and interact with other autonomous vehicles to travel in formation with a minimum safety distance—known as vehicle platooning—taking advantage of the suction effect that reduces the vehicle’s drag and therefore its energy consumption. In fact, it has been calculated that platooning could reduce energy use by between 3% and 25%.
Brainteaser 1
This brainteaser features three vehicles with distinct characteristics departing from the same origin to reach the same destination, yet they follow three alternative routes, each involving covering different distances.
Just relax in the car and forget about stress
Apart from their sustainability, another major benefit of autonomous vehicles is that they promise to reduce stress levels and increase user comfort and convenience on the road, according to surveys of drivers of semi-autonomous vehicles, where partial control of the vehicle is handed over. Paradoxically, however, this can be counterproductive for the environment, as the same surveys have shown that this increased comfort also results in more journeys and thus a higher carbon footprint. A survey conducted in 2021 found that users of semi-automated cars drove an average of 8,000 km more per year than drivers of conventional vehicles.
Brainteaser 2
A question of confidence
To achieve this vision of autonomous transport, driverless cars will have to overcome the limitations they still face today, which are mainly economic—they must be affordable for most users—but also social and moral, meaning that these vehicles must inspire a level of trust in the user that makes them willing to relinquish total control of the vehicle and become dependent on the machine.
It is relatively easy to program an autonomous vehicle to drive better than a human in ideal conditions of free-flowing traffic and surrounded by other autonomous vehicles with which it can communicate. But how will it react to the unforeseen and emergency situations that will inevitably arise in mixed traffic? For example, sharing the road with unpredictable human drivers, cyclists, electric scooters and pedestrians who are used to jumping and running traffic lights, jaywalking and looking at their mobile phones. Can we trust autonomous vehicles to make the best decision from a human and ethical perspective in these extreme situations?
With the aim of designing and training AIs to make decisions that are in line with our morals, several programs and experiments have recently been developed based on the massive collection of human decisions faced in the anomalous and/or extreme scenarios and situations that may occur in real driving. In this line, one of the most talked about initiatives is the Moral Machine Experiment, a study based on the classic trolley dilemma, a thought experiment that poses the following scenario: “Imagine a runaway trolley with no brakes is hurtling down the track towards five workers. You can’t warn them and you can’t stop the trolley, but you can pull a lever that will divert it onto another track. There is another worker there, but he is alone and equally oblivious to the danger. Should you pull the lever?”
However, as many experts point out, the trolley dilemma and its variations do not represent a real, or at least plausible, scenario in which participants really feel involved, nor do their responses. It is therefore not enough to define a moral code of human driving. More finesse is needed.
At this point, and bearing in mind that one of the main advantages of driverless vehicles, as explained at the beginning, is their environmental sustainability, the question arises as to whether it is appropriate to also include the environmental factor when creating a driving ethic for autonomous vehicles. More specifically, how should a driverless vehicle react when faced with a dilemma where there is a potential for serious environment damage?
Brainteaser 3: Defining an environmentally humane morality
To begin to answer this question, we pose an open-ended task—meaning that there is no single or correct answer—by presenting a diptych of dilemmas of this type, in which an autonomous vehicle is confronted with three extreme situations and the reader must choose the best or least bad solution in his or her opinion:
Comments on this publication