Artificial intelligence (AI) could help speed up the development of electric-car batteries, according to a recent report from The Wall Street Journal.
There is certainly no shortage of areas where engineers seek to improve on current lithium-ion batteries. Increased range, lower cost, and quicker charging times could speed up EV adoption, and they all depend on better batteries.
Much of battery development involves testing different combinations of materials, and AI could allow that to happen in less time, according to the report.
That’s because computers can sift through large amounts of data, and identity relevant information, much quicker than individual human researchers.
One example cited was an IBM project to develop a faster-charging battery free of nickel and cobalt. This involved evaluating a set of 20,000 potential compounds for the battery’s electrolytes. This would normally take five years, but was accomplished in nine days by AI, according to the report.
GM Ultium battery – cell stacking
AI can also accelerate testing of experimental batteries, the report said. The conventional method involves a large number of time-consuming charging and discharging cycles, but algorithms can be trained to make predictions about battery performance based on a smaller amount of data.
As a result, what previously required about three years of testing can now be done in as little as six months, Kensuke Nakura, general manager of Panasonic’s Energy Technology Center, told The Wall Street Journal.
AI is a popular tech-industry buzzword, but various companies have emerged in the past year or two claiming to use AI to speed development. This suggests it could be more than marketing gloss.
Volkswagen has said that AI and quantum computing can help lay out the roadmap for new materials—including within batteries—while Dyson has claimed to be using AI tools to help bring its solid-state battery tech to market.