This study looks at two machine learning methods—Fourier Neural Operators (FNO) and Multi-Layer Perceptrons (MLP)—to help speed up simulations of lithium-ion batteries. These simulations are based on a complex physics-based model called the P2D model. MLPs are commonly used because they’re simple and work well with regular data, but they struggle to handle data that changes across space and time. FNOs, on the other hand, are better at learning from this kind of data and need fewer parameters. We trained both models using data from simulations run under different current conditions, and then compared how well each model predicted key battery behaviors, like concentration and voltage, over time and space. Using error measurements to judge accuracy, we found that the FNO was more accurate and better at capturing the complex patterns in the battery data. This makes FNOs a promising tool for fast and accurate battery simulations.