The EV industry is currently obsessed with charging speed. Chinese automakers like BYD and Geely are already pushing megawatt-level charging systems capable of adding hundreds of miles of range in just a few minutes, essentially turning EV charging into something that starts to resemble a gas station refill.
That speed, however, comes with consequences. Frequent ultra-fast charging places enormous stress on lithium-ion battery packs, generating additional heat and accelerating long-term battery degradation over time.
Now, researchers at Chalmers University of Technology believe artificial intelligence could help solve part of that problem. In a newly published IEEE study, researchers developed an AI-controlled charging system designed to adapt charging behavior based on the battery’s age and condition.
According to the research, the system could extend usable battery life by nearly 23 percent without meaningfully increasing charging times. If those claims hold up outside simulations, the implications for EV ownership could be massive.
The AI Learns How Healthy The Battery Is
The charging strategy relies on reinforcement learning, a type of machine learning where the AI continuously improves decision-making based on feedback and outcomes. Instead of applying the same charging behavior to every battery regardless of condition, the system actively adapts charging voltage and current depending on the battery’s health.
That’s important because aging batteries behave very differently from new ones. Older battery cells become increasingly vulnerable to harmful chemical reactions during aggressive charging, particularly something known as lithium plating.
Lithium plating occurs when metallic lithium builds up on the battery’s electrodes instead of properly storing energy inside the cell structure. Over time, this can reduce battery capacity, hurt performance, and potentially create safety risks.
Current charging systems generally treat batteries the same, whether they are brand-new or several years old. The Chalmers system instead continuously monitors the battery’s electrochemical condition and modifies charging behavior in real time to reduce wear without sacrificing speed.
Faster Charging Usually Means More Battery Stress
In the simulations, batteries using the AI-based charging strategy reportedly survived around 703 charge-discharge cycles before dropping below 80 percent capacity. Batteries using conventional charging methods lasted roughly 572 cycles under the same testing conditions.
What makes the results especially interesting is that charging times barely changed. Researchers reported average charging sessions of roughly 24 minutes using either method, meaning the longevity gains did not come at the expense of convenience.
The software essentially learns when to ease off charging aggression during vulnerable moments while still maximizing overall efficiency. Researchers claim the system can even function using existing battery-management hardware without requiring major new sensors or expensive physical modifications.
That could make implementation relatively straightforward for automakers. In theory, some vehicles might eventually gain similar functionality through software updates alone.
The Industry Is Already Moving In This Direction
Even though the Chalmers project remains in the testing phase, the automotive industry is already exploring similar ideas. Battery-management startups and charging software companies are increasingly focusing on adaptive charging systems capable of balancing charging speed against long-term battery health.
Volvo Cars has already partnered with charging-software company Breathe for future EV programs, including the upcoming EX60. Breathe claims its software can improve charging speeds by as much as 15 to 30 percent while simultaneously reducing battery degradation.
The growing importance of this technology becomes obvious once you look at modern charging trends. Some next-generation EVs are now charging at over 1,000 kW, with manufacturers promising charging sessions measured in minutes rather than hours.
That may sound incredible from a convenience standpoint, but repeatedly hammering batteries with extreme charging power creates long-term durability concerns. If AI can genuinely reduce degradation while maintaining fast charging performance, it could become one of the most important hidden technologies shaping the future of EV ownership.
For consumers, longer-lasting batteries mean lower replacement risk, stronger resale values, and potentially reduced warranty concerns. For automakers, it could also reduce warranty costs while helping maximize the lifespan of expensive raw materials used inside EV battery packs. The biggest question now is whether the real-world results will match the simulations once physical testing begins.