Research: Analysis of Behavioral Patterns in Cooperative and Non-Cooperative Collective Dynamics Using Physics-Informed Neural Networks
Raoul Vetere graduated in Pure and Applied Mathematics from the University of Tor Vergata, where he also completed a master’s degree in Space Sciences and Technology. He later earned a second master’s degree in Computational Sciences at Roma Tre University, focusing on the application of physics-informed neural networks to physical and optimal control problems. During his academic career, he completed an internship at Thales Alenia Space, contributing to the development of machine learning models for advanced aerospace applications.
He is currently Chief Artificial Intelligence Officer at Involve Space, where he previously served as Chief Technology Officer. At Involve, he developed and implemented sophisticated Physics-Informed Neural Networks (PINNs), optimizing the control of stratospheric balloons in stochastic environments. By integrating data-driven predictions with physical laws, he enhanced operational management in the stratospheric environment.
His primary interests are in mathematical analysis applied to optimal control and the study of cooperative and non-cooperative differential games. He focuses on the conditions for the existence and uniqueness of solutions to these games and the development of numerical algorithms to solve them. Additionally, he works on advancing Physics-Informed Neural Networks to generalize classical and viscosity solutions of Hamilton-Jacobi-Isaac equations using deep neural networks.
Through his work, he has combined applied mathematics, artificial intelligence, and space technology to develop innovative solutions to complex problems, particularly in the management and optimization of stratospheric balloons.