Nowadays, Artificial Intelligence (AI) is already applied to several aspects of modern lifestyle. Machines can now create a completely artificial image of an unreal (but realistic) person by learning fundamental features from pictures of real people. AI is used for facial recognition with important implications in security and privacy-related matter. Machine learning is also emerging as a valid support for drug discovery.
Our program is ambitious and requires combining metasurface designing with machine learning; visual recognition with hyperspectral imaging; humanoids’ controllers with scanning probe microscopy; artificial intelligence with materials synthesis. However, the diverse and well assorted competences of the proponents put us in the unique position to pursue our vision.
Unprecedented; fast; autonomous; optimized. These are some of the keywords describing the methodology that we will develop in this Research Initiative.
AI will be used to enable fast, automated, optical and structural screening of the products of several hundreds of molecular syntheses monthly. With >10,000 syntheses per year and a tight selection of samples, our rate of discovery will be easily pushed-up to 10 new materials per year to be employed in unprecedented devices.
AI will be used for real-time interpretation of single-photons datasets. For example, AI will decode molecular maps (number of molecules per sample position) from bio-samples, by offering a valuable alternative to the inversion of complex models based on coincide photons statistics.
AI will be trained to find optimal metasurfaces design in order to realize unprecedented optical functionalities with ultimate performances.
AI will substitute scientists in running a scanning probe microscope by autonomously determine the best scanning parameters for ultimate image quality.
In a new synergy with humans, new machines with new capabilities will enhance the quality, the rate and the automation of cutting-edge techniques in Materials Science.