Road transport is a major contributor to air pollution, necessitating sustainable solutions for urban logistics. This study presents a time-dependent vehicle routing problem (VRP) model aimed at minimizing fuel consumption and greenhouse gas emissions while addressing stochastic customer demands. By incorporating key environmental factors such as road gradients, vehicle load, temperature, wind direction, and asphalt type, the proposed model provides a comprehensive approach to reducing transportation-related pollutants. To solve the computationally complex problem, a hybrid algorithm combining the gray wolf optimizer (GWO) and the multilayer perceptron (MLP) neural network is introduced. The algorithm demonstrates superior performance, achieving an error rate of less than 2% for medium-scale problems and significantly reducing fuel and driver costs. Sensitivity analyses reveal the profound impact of environmental parameters, with wind speed and direction altering optimal routing in over 80% of cases for large-scale instances. This research advances green logistics by integrating dynamic environmental considerations into routing decisions, balancing economic objectives with sustainability. The proposed model and algorithm offer a scalable solution to real-world challenges, enabling policymakers and logistics planners to improve environmental outcomes while maintaining operational efficiency.

The rapid expansion of livestream e-commerce has propelled the adoption of AI-driven virtual streamers. Drawing upon the Stimulus-Organism-Response (S-O-R) framework and the Computers Are Social Actors (CASA) paradigm, this study investigates how the technical and social stimuli of AI streamers—namely persona, anthropomorphism, and interactivity—shape consumer purchase intentions. The research focuses on the context of an MNC’s Vietnam AI virtual streamer (Vinamilk) on TikTok platform, utilizing a valid sample of 163 predominantly Generation Z consumers. To capture both linear net effects and causal complexity, this study employs a dual-methodological approach integrating Partial Least Squares Structural Equation Modeling (PLS-SEM) and fuzzy-set Qualitative Comparative Analysis (fsQCA). The PLS-SEM results reveal that an AI streamer’s persona and interactivity significantly enhance both consumers’ parasocial relationships (PSR) and perceived credibility (PC). Interestingly, while anthropomorphism successfully fosters PSR, it fails to significantly improve PC. Furthermore, within the highly entertainment-driven environment of TikTok, emotional connection (PSR) acts as the primary driver of Brand TikTok Engagement (COBRAs), which subsequently leads to actual purchase intentions. Complementing these linear findings, the fsQCA uncovers the principle of equifinality, identifying three distinct configurational pathways that sufficiently lead to high purchase intentions, indicating that no single condition is absolutely necessary. The findings suggest that digital marketing strategies should prioritize interactive capabilities and emotional resonance over mere visual realism to optimize AI-mediated commerce.