There is a moment in the history of the automobile that almost no one saw coming. Today, brands from all over the world, from the most traditional to the youngest, compete for the same goal: a car capable of driving itself. It is not about selling more technology for its own sake. Autonomous driving promises to change the way we move, reducing accidents, recovering lost time and, along the way, transforming cities. But how does such a car really work, and how much is left before we see it on the streets?
What autonomous driving is
In a few words, an autonomous car is a vehicle that drives itself without the need for a person to keep their hands on the wheel. To achieve this, the car needs three things: see what surrounds it, understand what is happening and decide what to do in a fraction of a second. It is like a human driver, but with sensors instead of eyes and a computer instead of a brain.
It is worth clarifying that not all “smart” cars are autonomous. Today most already have aids such as automatic braking or lane keeping. That is driving assistance, not autonomy. The difference lies in who has the final responsibility: if the human is still at the wheel, the car is an assistant; if the car takes over everything, then we are talking about autonomous driving.
The five levels of autonomy
To put it all in order, engineers divided autonomous driving into five levels, from 0 to 5. At level 0 the car only warns; at levels 1 and 2 it already helps with the accelerator or the wheel, but always with a human watching. Most new cars today are between levels 1 and 2.
The real leap begins at level 3: in that case the car can drive itself under certain conditions, such as a highway, although the driver must be ready to intervene if necessary. At level 4 the vehicle is capable of driving without human intervention in specific areas and situations, for example on a delimited urban route. And level 5 is the utopia: the car drives itself in any place and situation, without even a steering wheel.
How an autonomous car “sees”
The most fascinating thing is the perception system, because the car has to build an image of the world in real time. To do that it combines several technologies. Cameras, placed around the vehicle, read signs, traffic lights and lanes. Radar, the same system used by airplanes, measures distances and speeds, and works even in rain or fog.
And there is lidar: a sensor that fires laser pulses and measures how long they take to return, creating a three-dimensional map of the environment with millions of points per second. It is, in essence, the most precise “eyes” that exist for a car. All that information is combined in a central computer that decides to accelerate, brake or turn. The whole process happens in milliseconds, much faster than a human.
Why it matters: safety and time
The great promise of autonomous driving is safety. The World Health Organization estimates that around 1.2 million people die every year in traffic accidents, and the vast majority are due to human errors: distractions, fatigue or speeding. A machine does not get distracted or fall asleep, so supporters of this technology believe it can save hundreds of thousands of lives.
There is also the matter of time. The average driver spends tens of minutes a day behind the wheel, time that, with an autonomous car, could be devoted to reading, working or resting. And in the future, driverless cars could circulate in shared fleets, reducing the number of vehicles on the streets, congestion and parking.
What is needed for it to be fully real
With so many advances, why do we still not see fully autonomous cars on every corner? The answer is that the challenge is not only technological, but also of trust and regulation. Although the systems are already very good on highways, driving through a city full of pedestrians, roadworks and unpredictable situations remains an enormous challenge: every street, every climate and every country is a different world.
In addition, the laws are not yet ready in most countries. Who is responsible if an autonomous car has an accident: the manufacturer, the owner or the software? These are questions that governments take time to answer. That is why experts talk about a gradual arrival: first fleets of driverless taxis in specific cities, then delivery vehicles and, much later, fully autonomous private cars. It will not happen overnight, but the steering wheel that does fewer and fewer things already gives us a hint of where we are heading.
