The electric automotive transition is crucial for achieving carbon neutrality, especially in emerging economies like China. However, the scarcity of critical materials in lithium-ion batteries (LIBs) challenges electric vehicle (EV) deployment targets. Our work delivers a comprehensive framework of EV battery recycling, considering resource compensation, environmental performance, geospatial optimization, and cost feasibility of closed-loop LIB recycling under China's carbon neutrality. Our findings show that meeting EV deployment targets will widen the supply-demand gap, with cobalt and manganese demand exceeding 2022 production levels by 54-fold and 116-fold, respectively. Battery recycling is crucial for mitigating material scarcity, necessitating a minimum 84% collection rate to stabilize supply by 2060. Battery recycling remains economically viable in most scenarios, generating a net profit of US$58 billion in the optimal scenario. Here, our work underscores inherent trade-offs among integrated metrics, informing battery recycling strategies to strengthen supply chain resilience and advance automotive electrification under decarbonization goals.

The identification of accident hot spots is a central task of road safety management. Bayesian count data models have emerged as the workhorse method for producing probabilistic rankings of hazardous sites in road networks. Typically, these methods assume simple linear link function specifications, which, however, limit the predictive power of a model. Furthermore, extensive specification searches are precluded by complex model structures arising from the need to account for unobserved heterogeneity and spatial correlations. Modern machine learning (ML) methods offer ways to automate the specification of the link function. However, these methods do not capture estimation uncertainty, and it is also difficult to incorporate spatial correlations. In light of these gaps in the literature, this paper proposes a new spatial negative binomial model which uses Bayesian additive regression trees to endogenously select the specification of the link function. Posterior inference in the proposed model is made feasible with the help of the Pólya-Gamma data augmentation technique. We test the performance of this new model on a crash count data set from a metropolitan highway network. The empirical results show that the proposed model performs at least as well as a baseline spatial count data model with random parameters in terms of goodness of fit and site ranking ability.