supermaterials

The applications of the artificial intelligence (AI) are presumably unlimited. The rapid development that this technology has experienced during the last five years, and of which we are all being to a greater or lesser extent, invites us to assume that we are probably not yet entirely aware of the impact it will have on our lives. The AI ​​is already being used in the medical diagnosis by images, to elaborate new drugs, to discover exoplanets, or even to filter the data that LHC detectors collect (Large Hadron Collider). These are just some of the applications that are already underway, and we can be sure that many more will arrive in the future. In any case, the authentic protagonist of this article is a discipline in which AI already has much to say: the science of materials. And this technology is already being used to Design new materials with optimal properties for biomedicine, energy, aeronautical or electronic applications. The era of supermaterials is in the hands of artificial intelligence Domenico Vicinanza, who is an associate professor of smart systems and data science at the Anglia Ruskin University in Cambridge (United Kingdom), explains in the very interesting article he has published in The conversation that the complexity of many materials at the atomic and molecular level forces scientists to invest a lot of time and effort in their search. Until the arrival of AI, the design of a new material required to resort to specialized equipment and apply the test and error method once after another. Fortunately, AI is changing everything. And is doing it because Automatic learning In particular, it puts in the hands of scientists and engineers the possibility of developing more efficient and better directed strategies. In fact, current AI models are already capable of Generate new materials in a direct wayand, therefore, without the need to resort to the essay and error in the traditional way, from the set of requirements and properties specified by the researchers. Somehow we are facing a technology that allows us to obtain new “to the letter” materials. Automatic learning puts in the hands of scientists and engineers the possibility of developing more efficient and better directed strategies On January 16, a group of Microsoft researchers published an article in the scientific journal Nature in which Mattergen and Mattersim announced. They are two AI tools designed to elaborate inorganic materials, and, therefore, not based on carbon. The first one is able to generate new candidate materials, while Mattersim carries out the filtering of the candidates and their validation with the purpose of ensuring that it is possible to manufacture those materials with the capacities of the real world. The most surprising thing is that Mattergen can generate thousands of materials with specific properties in much less time than it would be necessary to invest using conventional material science techniques. In this way, researchers can explore a much broader fan of possible new materials and then carry out an exhaustive analysis only of the most promising candidates. The implications of this technology are huge. And it is that in the short term in the short term it will have a very deep impact on the areas of battery design, renewable energies, the manufacture of medical devices, implant setting, obtaining new drugs, carbon capture or The administration of waste, among many other possible applications. Image | Oak Ridge National Laboratory More information | The conversation | Nature In Xataka | Solid state batteries will only be a success if they are safe. Superionic materials have reached the rescue