There are a half dozen refrigerator-sized cabinets, Inside of a LAB of Stanford University’s Precourt Institute for Energy. They are to kill batteries as fast as they can. Each holds 100 lithium-ion cells secured in trays that can discharge and charge the batteries dozens of times per day. Usually, the batteries in electrochemical torture chambers are inside electric vehicles and gadgets. Nevertheless, when they are in those hulking machines, they are not powering anything at all. Instead, energy is going in and out of those cells as fast as possible. It is for generating performance data reams that will teach artificial intelligence how to build a better battery.
A team of researchers from the Toyota Research Institute, Stanford, and MIT, in 2019, used artificial intelligence on data generated from those machines. It was to predict lithium-ion batteries’ performance over the lifetime of the cells before their performance had started to slip. Usually, artificial intelligence needs data from after a battery had begun to degrade. It is for predicting how it would perform in the future. It can take months to cycle the battery enough times to get that data. Nevertheless, after only hours of data collection, artificial intelligence researchers could predict lifetime performance, while the battery was still at its peak.
William Chueh works at Stanford. He is a materials scientist there and one of the lead authors of the 2019 paper. He said that before their work, nobody thought that was possible. They did it again earlier this year. Chueh and his colleagues described an experiment in which artificial intelligence could discover the optimal method for 10-minute fast-charging a lithium-ion battery.
Many experts think that fast-charging batteries will be critical for electric vehicle adoption. Nevertheless, dumping enough energy to recharge a cell in the same amount of time it takes to fill up a gas tank can quickly kill its performance. To get fast-charging batteries out of the lab and into the real world, it means to find the sweet spot between battery lifetime and charge speed. Nevertheless, the problem is that there is effectively an infinite way to deliver charge to a battery. Chueh compared it to searching for the best way to pour water into a bucket. It is an arduous and slow task to sift through all those possibilities to find the best way. Nevertheless, it is where artificial intelligence can help.
Let us wait to see what Artificial Intelligence can do about the batteries. AI is very useful for batteries.