Scientists want to improve batteries using neural networks: how it will work

Today, artificial intelligence is actively used in various areas of our life, including scientists who actively use it for their developments. Among them are researchers at the University of Toronto who, using neural networks, want to improve the properties of batteries, writes Techexplore.

The fact is that scientists are now trying to find new combinations of materials that will replace the usual lithium in batteries. Few people know, but today there are already 140 thousand known materials, but the study of their combinations with each other, taking into account the required characteristics of the final product, takes not just years, but decades.

With the help of a neural network and their own analysis algorithm, the Toronto researchers plan to reduce the time for their research. They want to find the best combinations for solar panels and batteries.

Of course, ideally, it is best to use a quantum computer for such research, but they are now in a state of refinement and active testing, so there are only supercomputers (which is not enough) and neural networks at hand.

In their work, Canadian scientists actively use the power of the Niagara supercomputer, located at the SciNet center. They hope that in the near future they will be able to present a material that, in terms of its characteristics, will store energy much better than lithium-ion batteries.

Focus previously wrote that a mega-energy storage facility was built in California from old electric vehicle batteries .