Technology

Physical artificial intelligence ignites the new age of the automobile

The scenario. The industry explores and adopts new technological paradigms to evolve cars for greater efficiency and safety

by Giancarlo Calzetta

26 February 2026

3' min read

Translated by AI

Versione italiana

3' min read

Translated by AI

Versione italiana

Observing the automobile market, the impression is growing stronger and stronger that it is taking on a new identity, moving from being a 'muscular' mechanical vehicle to a hybrid conceived as a robotic, digital and personal entity. In past years, the focus was almost exclusively on the electric transition, in recent months manufacturers have been shifting the centre of gravity towards so-called physical intelligence, and vehicles are becoming increasingly software-defined: the sheet metal is becoming a container for sophisticated algorithms, where computing power and the integration of generative artificial intelligence are the real differentiating factors, taking the place of horsepower, trim and chrome paint.

Autonomous driving is experiencing a pragmatic renaissance, abandoning the hyperbolic promises of the past to present solutions within the reach of current legislation.

It is no coincidence that Nvidia has unveiled the Alpamayo platform (see article opposite), a universe of models, simulations and datasets for autonomous vehicles that allows previously unthinkable complexities to be managed on board. A technology that will soon find its way onto real roads thanks to Mercedes-Benz, which has announced its use on the new CLA, aiming for an increasingly effective level 3 and 4 autonomy.

From China, on the other hand, Geely unveils Full-Domain AI 2.0 architectures that enable the vehicle not only to perceive the environment, but to understand it, taking autonomous parking and urban navigation to extreme levels of precision. These two examples are proof that physical AI, initially 'born' to help robots become more versatile, is beginning to play a decisive role in autonomous driving. Moreover, Arm has also presented its Physical AI division, which is also mainly aimed at managing automata, but which will certainly have spin-offs in the automotive world.

On the electrification front, we are seeing the long-awaited developments aimed at solving the now eternal critical issues related to autonomy and recharging times. The focus on solid-state batteries is very high, with prototypes promising unprecedented energy density and superior thermal stability. And another trend that seems to be gaining strength is that of consumer electronics companies (mainly Chinese) producing cars.

After the pioneer Sony and the various Huawei, Xiaomi and so on, Dreame is also challenging the traditional manufacturers with its Nebula concept, a high-performance electric supercar that is for the moment being passed off as 'a technology demonstration' but is likely to become a real show of strength on the markets in the coming years. And this can only become another spur for European manufacturers thinking of the cockpit as a new arena to fight the war of differentiation. Already in past years we have seen concepts where monitors literally made up the cabin furniture, taking the place of briarwood and alcantara, but now we have the first market-ready products complete with immersive screens and proactive virtual assistants.

Sony Honda Mobility, with the Afeela model, has put on sale an electric suv that harnesses the computing power of Qualcomm to offer a digital ecosystem in which dialogue between man and machine takes place naturally, while Bosch and LG have shown intelligent cockpits equipped with artificial intelligence capable of monitoring the driver's state of health and adapting the interior environment according to his or her emotions or work needs. Finally, the idea of the car becoming a robot is tickling the imagination of designers who are beginning to fantasise about automated and sometimes anthropomorphic car services. Hyundai and its subsidiary Boston Dynamics have shown how 'humanised' artificial intelligence can assist passengers not only inside the vehicle, but also during loading. At the same time, the emergence of (less human, but concrete) robots for autonomous charging such as those presented by Autovoltek is a further hint towards the total automation of ancillary services.

Others, such as the traditional brands Bmw and Ford, have instead focused on an agnostic and flexible approach, integrating AI language models to enhance voice interaction.