Answers are grounded in this paper and cite their sources.
The evolution of lithium battery performance involves complex coupled processes across multiple scales, ranging from electronic and atomic scales to mesoscopic structures and battery systems. Traditional simulation approaches face significant challenges in accurately predicting realistic battery electrochemical responses because of scale-bridging limitations, high computational costs, and insufficient dynamic feedback mechanisms. In recent years, advances in artificial intelligence (AI) have provided opportunities for cross-scale simulations by enabling knowledge transfer and intelligent prediction through learning physical representations from data. This Review provides a comprehensive overview of AI-driven cross-scale simulation for lithium batteries. The review first discusses AI-enabled simulation approaches spanning the micro- to mesoscale, including machine learning potentials (MLPs), AI-driven kinetic Monte Carlo (AI-kMC), AI-driven phase-field modeling (AI-PFM), and AI-driven coarse-grained molecular dynamics (AI-CGMD). It then summarizes meso- to macroscale methodologies, including surrogate models, reduced-order modeling, and physics-informed neural networks (PINNs). Furthermore, AI applications in multiscale feature fusion, inverse materials design, and battery design automation (BDA) are discussed together with computational paradigms toward digital twins and intelligent battery research and development. Despite substantial advances, several critical challenges, including limited model generalization and insufficient uncertainty quantification, remain. Integrating foundation models, generative AI, and multiscale physical modeling could enable more accurate and computationally efficient battery simulation frameworks. These advances may accelerate the discovery, optimization, and deployment of next-generation energy-storage technologies.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: