2023 will go down as the year of artificial intelligence. Everyone is talking about it and has talked about it. Everyone wants it, some fear it and others praise it. It's a worldwide craze that affects many areas, including the automotive industry. How many times in recent months have we talked about AI and mobility? Countless.

Today, some people are using the term "artificial intelligence" to describe certain aspects of their technologies, although the time has not yet really come. For some, AI is HAL 9000 from 2001: A Space Odyssey, Skynet from Terminator or any droid that has appeared in a film, from Star Wars to Blade Runner. Today's reality tells us that these are computers connected to the web and educated by billions and billions of pieces of information.

So we're still a long way from science-fiction scenarios, but something is happening and the road ahead seems to be mapped out: artificial intelligence is going to take over more and more of our cars. Here's how and where.

True autonomous driving

Autonomous driving has been discussed for some time. Some manufacturers still believe in it, while others seem to have thrown in the towel, either for good or until the time is right. We are now seeing the first level 3 examples, i.e. with limited possibilities of letting the car manage the steering, brakes and accelerator. Artificial intelligence could be the real turning point.

Cars capable of driving themselves thanks to increasingly sophisticated and precise sensors (radar, cameras, lidar), whose data is processed in real time by an AI capable of making decisions in different scenarios. From city traffic to long motorway journeys. Systems capable of learning as they go along, so that they can adapt to different situations as effectively as possible. Systems with predictive functions, based on information from the satellite navigation system or external sources.

Questions remain, of course, such as liability in the event of an accident (in the UK, this will be the responsibility of manufacturers) and the choices to be made by the AI in the event of danger. This is an issue addressed by the online platform Moral Machine, developed at MIT (Massachusetts Institute of Technology), whose aim is to create a moral code to be taught to artificial intelligence.

Nissan ServCity

A prototype self-driving Nissan taxi in London

Advanced infotainment

By now, we're used to getting into cars equipped with increasingly large screens. Digital instruments, screens for front and rear passengers and augmented reality head-up displays are commonplace; in the future, they will be even more powerful thanks to artificial intelligence.

Auto makers are already trying their hand at advanced voice assistants with the integration of ChatGPT on their models, capable of performing complex actions to manage all aspects of the car, even automatically, with smoother and more efficient interactions. Again, the software will be able to evolve over time, learning which music tracks to play depending on who is driving. OTA (over the air) updates will also become even more precise and timely, keeping the car in step with the times.

Prevention is better than cure

How many times have we set the car in motion only to find that it has broken down unexpectedly? The headlights come on and you have to go to the garage to repair the damage, at considerable expense. In this case, artificial intelligence will help to constantly monitor the car's state of health, with comprehensive checks to prevent any malfunctions in good time.

The predictive functions of artificial intelligence help to reduce costs and risks, because a car that doesn't work properly is not a safe car.

BMW Proactive Care

BMW Proactive Care

The best route

If you regularly drive an electric car, you already know: GPS is an essential tool, because it includes charging stations in the itinerary. In some cases, it even automatically reserves the recharging station, pre-cooling the batteries at the right time so that they are optimally ready for charging.

Thanks to artificial intelligence, this aspect will also undergo a major evolution, with real-time recalculation of the route in the event of changes in traffic, weather conditions or driving style. Battery management will also be even more precise.

Always connected

We live in a world of acronyms, and one of the most widely used in technology is IoT, the Internet of Things. An acronym that identifies every object connected to the network. Not just smartphones, but also household appliances, lamps, alarm systems and cars. A global network to communicate with everything around us, thanks of course to artificial intelligence.

The IoT is the foundation of smart cities, those much-vaunted cities where traffic lights, buildings and vehicles are connected and communicate with each other. A wealth of information to sort through to avoid traffic jams and accidents. A sort of summary of the four points mentioned above: only cars capable of driving themselves and equipped with an advanced infotainment system can enable the birth of cities capable of managing themselves thanks to AI. Vehicles that communicate with other vehicles (the acronym in this case is V2V, Vehicle To Vehicle) to signal their presence or obstacles along the route, traffic lights that anticipate red lights, and so on. All managed by complex software that adapts according to conditions.

Mise à jour OTA du système Mercedes-Benz MBUX

The Mercedes infotainment system

The dangers 

The points raised above are somewhere between science and science fiction. They are technologies in development, sometimes still immature, but they are part of the plans of carmakers and the big names in technology. But as hundreds and hundreds of books and Hollywood films have taught us, artificial intelligence also entails risks.

We certainly don't have to fear the machines revolting, but relying on increasingly complex software doesn't always solve problems. Sometimes it even creates them. This is the case with cybersecurity: the billions of pieces of data circulating on the network need to be armoured and hacker-proof. The risk is not only that your personal data will fall into the hands of malicious people, but also that your vehicle will be stolen or - this is a borderline case, of course - that you will no longer have control of it because someone has managed to break into the steering, accelerator and brake management system remotely.

This is why a cybersecurity standard for cars, called ISO 21/434, was created in 2021 to guide manufacturers in developing secure, intrusion-proof systems.