AIxtreme

ABOUT THE PROJECT

The AIxtreme project aims at providing an AI-based, physics-inspired solution for forecasting extreme weather and space weather events. The basic idea of the project is that natural disasters at Earth caused by extreme weather events can be avoided, or at least mitigated, by efficient forecasts often based on satellite data. On the other hand, the performance of Earth- and space-based technological systems, such as terrestrial communications links, power grids, and satellite operations can be significantly affected by the impact of extreme space weather events. 

The AIxtreme project is aimed at seeing the forecasting problem as an overall problem that affects the entire heliosphere, i.e., the area under the influence of the Sun, and to provide an AI-based, physics-inspired computational framework where to formulate and implement weather and space weather prediction tools. In this perspective AIxtreme intends to combine the strengths of deterministic and data-driven approaches: the novelty of this new approach lies in the inclusion of the physical knowledge, typical of the deterministic approaches, into the learning process typical of a learning machine. 

Following the chain space weather - weather, the main project is then split into two complementary sub-projects. 

• AIxtreme-I: Physics-driven approaches in concert with AI-based methods to predict extreme space weather events. 

• AIxtreme-II: Mechanistic approaches in concert with AI-based methods to predict extreme weather events.