What happens when artificial intelligence encounters the unpredictability of the real world? That question lies at the core of the work of Marco Pavone, Professor of Aeronautics and Astronautics at Stanford University, Head of Autonomous Vehicle Research at NVIDIA and President of the Scientific Committee of AI4I
In an interview published by repubblica.it on May 16, Pavone reflects on the rapid evolution of Physical AI, artificial intelligence applied to systems that interact directly with the physical world, from autonomous vehicles to robotics and aerospace systems.
“The goal,” Pavone explains, “is to enable robotic systems to act safely and effectively even in conditions different from those anticipated during the design phase.”
The article traces Pavone’s trajectory from Sicily to the Massachusetts Institute of Technology, through the Jet Propulsion Laboratory at NASA and eventually to Stanford, where he founded the Autonomous Systems Laboratory. Across these experiences, a common thread emerges: designing autonomous systems capable of operating reliably in uncertain, dynamic and partially unknown environments.
According to Pavone, the acceleration currently reshaping the sector is the result of several technological trajectories converging simultaneously: advanced simulations, AI models, increasingly sophisticated robotic hardware and unprecedented computational power. “Even people working in the field,” he notes, “would have struggled to imagine this level of progress just two or three years ago.”
These themes are closely aligned with the strategic direction pursued by AI4I, where Physical AI represents one of the most advanced research and industrial frontiers connecting artificial intelligence, engineering, robotics and industrial transformation.
As President of AI4I’s Scientific Committee, Pavone contributes to strengthening the Institute’s international scientific positioning and to consolidating links between frontier research, advanced infrastructures and industrial applications.
The same vision has also shaped several recent AI4I initiatives, including the national forum “Officine d’Intelligenza”, where Pavone addressed the relationship between AI and industrial transformation, and the activities of the CSP–IAS programme, developed together with Fondazione Compagnia di San Paolo to attract top international scientific talent to Italy.
“Physical AI,” increasingly central to the evolution of intelligent systems, is emerging as one of the domains where the interaction between algorithms, simulation, safety and real-world environments will define the next generation of industrial technologies.
