Conclusion

The use of mesoscopic approaches to solve industrial problems is still in its infancy. However, a wide range of possibilities is open. Not only are codes becoming more and more powerful, running on more and more powerful machines, but the role of the experimental numericist, who exploits computational methods to solve physical or industrial problems, is becoming increasingly established within universities and companies. On the other hand, the artificial intelligence revolution is opening up new avenues of development. Finally, combining atomistic, or even quantum, and mesoscopic approaches, at the heart of multiscale vision, will enable us to better understand certain phenomena that are otherwise difficult to apprehend.