3. Conclusion

Mathematical modeling and data analysis using artificial intelligence are two major scientific and technological avenues to help rethink the way agricultural and agri-food systems operate.

By equating knowledge of how plants function, we can simulate their behavior in silico. Multiphysics approaches make it possible to couple geological, biological, mechanical and energetic phenomena to describe the various aspects of crop behavior. The challenge is not necessarily to implement holistic models, whose simulation can be computationally time-consuming and parameterization demanding in terms of input data, but rather to compose a coupled representation of the phenomena we need to understand, anticipate or optimize for a given problem. A generic mathematical formalism has been presented, making it possible to formalize agronomic control and optimization problems....