- FramekworkI am currently an Associate Professor in the Department of Industrial Engineering at the University of Trento, where I bring over a decade of experience in advanced control systems and machine learning techniques for autonomous system. My research bridges the fields of applied mechanics, control theory, and machine learning, with a focus on developing intelligent and adaptive models and controllers (see the Presentation of the research). Specifically, I aim to integrate classical model-based principles with data-driven techniques to create innovative solutions for model and control physical systems.
A major milestone in my career came in December 2024 when I was awarded the prestigious Italian FIS2 grant for the project Neu4mes: "Structured neural network framework for modeling and control of autonomous systems." This grant supports the development of nnodely, a groundbreaking open-source Python framework that combines the flexibility of neural networks with the rigor of traditional model-based design. nnodely facilitates the creation of model-structured neural networks, a novel approach that integrates physical principles and control theory into neural network architectures. These neural network are designed to require fewer parameters and less training data while maintaining the flexibility and power of machine learning. This makes them particularly well-suited for real-time application and autonomous systems where the energy consumption is the key issue.
At the core of my research is the ambition to merge the strengths of machine learning and model-based control, crafting self-learning strategies that draw on the best aspects of both. This interdisciplinary approach continues to drive my work, pushing the boundaries of what is possible.
Gastone `tonegas` Pietro Rosati Papini - last updated on