An AI called Dragonfly helps design faster-charging batteries

An AI called Dragonfly helps design faster-charging batteries

Batteries are more crucial than ever as they power cars, power our countless devices, and even allow some experimental aircraft to fly. But battery technology still has a long way to go before we see more widespread adoption of electric vehicles, laptop battery lifespans of months, and longer flights on electric planes. That’s why engineers and researchers around the world are constantly searching for the next big battery innovation.

According to an article recently published in Nature Communication, Carnegie Mellon researchers have used a combined system of robotics and artificial intelligence to design better electrolytes for lithium-ion batteries. In particular, the team was looking for electrolytes that would allow batteries to recharge faster, which is one of the biggest problems in battery technology today and a major obstacle to the widespread adoption of electric vehicles.

Lithium-ion batteries have a cathode and an anode surrounded by an electrolyte. When charged, the ions migrate through the electrolyte from the cathode to the anode (and vice versa when discharging). The exact composition of the electrolyte determines the rate at which a battery charges, discharges, and otherwise functions. Optimizing the electrolyte solution is therefore one of the main challenges for battery designers.

The research team used an automated arrangement of pumps, valves, vessels and other lab equipment they dubbed “Clio” to mix different ratios of three potential solvents and a salt. As the document points out, “battery innovations can take years to deliver,” in part because there are so many potential chemicals that can be used in various ratios that optimizing them is “long and laborious.” , at least for people. But with its various automated parts, Clio was able to conduct experiments much faster.

[Related: Why Dyson is going all-in on solid-state batteries]

To further eliminate the human element, Clio’s results were fed into a machine learning system called “Dragonfly” which analyzed the data to look for patterns and come up with alternative ratios that might perform better. Clio then automatically ran these proposed new experiments, allowing Dragonfly to further optimize chemical recipes.

In total, working with a single salt and three solvents, Clio and Dragonfly were able to conduct 42 experiments over two days and come up with six solutions that outperformed an existing electrolyte solution made from the same four chemicals. The best test cell containing one of the electrolytes developed by robot-IA showed a 13% improvement in performance over the best performing test cell using the commercially available electrolyte.

In an interview with MIT Technology ReviewVenkat Viswanathan, associate professor at Carnegie Mellon and one of the co-authors of the Nature Communication paper, explained that the problem with working with electrolyte ingredients is that you can combine them “in a billion ways.” Previously, most research was based on guesswork, intuition, and trial and error. By being both bias-free and able to quickly navigate experimental conditions, Clio and Dragonfly can test many more options than human researchers – whether minor refinements or lunar solutions – and are not paralyzed by their preconceptions. They can then take what they learn from each experience and tweak things to find the optimal electrolytes for whatever the research team needs.

In this case, Clio and Dragonfly were optimizing recharge speed, but similar “closed-loop” experiments could optimize capacity, discharge time, voltage, and all the other factors that matter to commercial battery performance. In fact, the team believe their work will be “useful beyond the battery community”, saying their “custom-designed robotics platform, experiment planning and integration with device testing will useful for optimizing other autonomous discovery platforms for energy and materials applications”. science in general”.

The Carnegie Mellon team isn’t alone in exploring how machine learning can optimize the many design considerations and complex variables that go into making, maintaining and charging batteries. Late last month, a team of government researchers from the Idaho National Laboratory, run by the Department of Energy, announced that they had found a way to safely and reliably charge vehicles. electricity up to 90% in just 10 minutes. They used a machine learning algorithm to analyze between 20,000 and 30,000 data points from different types of lithium-ion batteries to find the most efficient and safest charging method. They were then able to confirm their results by testing the newly developed charging protocols on real batteries.

And while liquid electrolytes are one frontier for battery research, another is exploring ways to replace that liquid with a solid instead.

#called #Dragonfly #helps #design #fastercharging #batteries

Leave a Comment

Your email address will not be published. Required fields are marked *

Prendre rendez-vous en ligne