One of the biggest things holding back EVs today is battery technology. We all dream of a future where we can charge our EVs just as quickly as going to a gas station. Well, research published by a Stanford lead team could make that a reality. And it’s not just charging. We could soon see major advances in every aspect of battery technology, with the help of AI.
Eliminating roadblocks
The largest difficulty in developing new battery technology is testing. It’s often necessary to try out a multitude of different things that can have an effect on performance.
In battery testing, you have to try a massive number of things, because the performance you get will vary drastically.
Stefano Ermon, assistant professor of computer science
This testing can be a major roadblock when trying to develop fast charging technology. Engineers must test many different charging methods to find the optimal way to charge the battery without reducing lifespan. This study, lead by Stanford in collaboration with MIT and Toyota engineers could speed up this process by 98%.
Machine learning is trial-and-error, but in a smarter way.
Aditya Grover
While these trial-and-error tests can take years for humans, they’re a perfect problem for computers. Rather than testing every possible method equally, machine learning algorithms can efficiently determine which methods are most promising and quickly zero in on them.
Speeding things up
This approach doesn’t just help find new methods for fast charging. This technology can be applied to every step of the battery development process.
This is a new way of doing battery development
Having data that you can share among a large number of people in academia and industry, and that is automatically analyzed, enables much faster innovation.
Patrick Herring from the Toyota Research Institute
This team also previously demonstrated a faster method of testing the lifetime of batteries. Rather than repeatedly cycling batteries until they fail, they could accurately predict the performance of a battery after just 100 charges.
The researchers are making their machine learning and data collection system freely available to battery scientists. By making testing quicker and using machine learning to find promising approaches to the development of new battery technology could speed up greatly in the coming years.
Our take
It’s always important to realize the amount of time it takes for this kind of research to actually get use. It will take time to validate this technology. But I’m excited to see the speed of battery development once these techniques start getting used.
The potential of faster charging, longer range, and cheaper batteries can’t be overstated. The world’s lithium supply is limited, and increased demand for EVs makes better battery technology essential. We will mine the earth barren of lithium if we do not improve battery tech in the next few years.
I think the approach of this study is genius. They aren’t just developing a slightly better battery than we currently have. This team is working to drastically change how batteries are developed forever. This is really what’s needed to solve the problems EVs are set to face in the future.
This research could completely revolutionize battery development. I can’t wait to see the effect this has on EVs of the future.
We’re always watching for new developments in EV technology, so follow our social media up top to keep up to date!
