Yongzhi's research develops computational methods that infer a battery's hidden internal states — concentration gradients, degradation modes, and safety-critical thresholds — from routinely available operational measurements. His work spans three interconnected directions:

• Battery degradation diagnostics and remaining useful life prediction. Physics-informed methods that identify which degradation mechanisms are active and predict how battery health will evolve.
• Differentiable digital twins, Bayesian inference, and identifiability analysis. Coupling electrochemical physics with statistical inference to determine what can and cannot be known about a battery's interior from external data.
• Battery thermal safety. Linking degradation diagnostics with thermal failure models to predict how a battery's susceptibility to thermal runaway evolves as it ages.