Fig. 1

"Platform Research" model with a "Four Beams N Pillars" architecture[3]"

Fig. 2

The process of battery R&D under the BDA platform is "Read, Design, Make, Test, and Analysis""

Fig. 3

Principle of Uni-SMART model[27]"

Table 1

The prediction effect of Uni-SMART model under different tasks in battery[27]"

Fig. 4

Application of cross-scale simulation algorithm in battery R&D: (a) multi-scale algorithm; (b) DeePMD breaks through the bottleneck of traditional modeling methods to develop a new anode system: reproduces the voltage plateau gap between crystalline and amorphous, and the latent heat of phase transition between a/c and Li15Si4 in the initial state of lithiation is 0.26 V, corresponding to the voltage lag affected by crystallization[45]; (c) DeePMD is used to reveal the diffusion mechanism of lithium ions in the main components of lithium-ion battery SEI membranes[46]; (d) AI-assisted first-principles molecular dynamics studies of lithium-ion transport processes in SEI[47]"

Fig. 5

Design and development of data-driven batteries for pre-trained models: (a)—(e) first-principles precision solid-state electrolyte pre-trained models study microscopic laws in depth to accurately predict the conductivity and migration barriers of Li examples[58]; (f)—(h) Prediction of melting point, boiling point and dielectric constant based on Uni-Mol model[55]; (i)—(k) Coulombic efficiency and conductivity of electrolyte formulations were predicted based on the Uni-ELF model[55]"

Fig. 6

Application of AI4S product tools in electrolyte and electrochemical simulation scenarios[64]: (a) electrolyte high-throughput screening by an APP for electrolyte design; (b) Electrochemical Simulation and prediction APP accelerates cell test analysis"

Fig. 7

AI accelerates material-to-cell fabrication process optimization"

Fig. 8

The particles were identified by AI4S characterization tool"

Fig. 9

Battery data analysis and management process"

Fig. 10

The platform-based intelligent R&D of batteries under the paradigm of AI for Science realizes a new R&D paradigm of "integration of software and hardware, combination of experiments and computational simulations", creates a smart large device and super laboratory for the whole life cycle of batteries, and comprehensively empowers the upgrading of the battery industry"