Hybrid AI to enhance science, engineering and technology

This presentation revisits the genealogy, anatomy and physiology of digital twins, emphasizing the needs for advanced numerical technologies empowering the efficient solution of state-of-the-art physics-based models, while making use of recent data-driven technologies for bridging the gap with the physical reality. Model order reduction, machine learning, data-assimilation and immersive reality technologies will be integrated into the timely concept of digital twin, enabling for diagnosis prognosis and decision making of engineering components and multi-component systems These Digital Twins are then employed in the design and operation of materials, processes structures and complex systems of systems, enabling real-time predictions, optimization, uncertainty propagation, optimization and control, all of them under the stringent real-time constraint. The different concepts and tools will be motivated, introduced and applied in different domain of engineering, from materials behavior to the digital twin of cities and territories.