The field of urban science and transportation research is moving towards leveraging large language models and multimodal approaches to extract insights from unstructured data, such as street view images and incident reports. This allows for more interpretable and actionable hypotheses to be generated, enabling the discovery of previously overlooked correlations between urban design and safety outcomes. The integration of AI-assisted perception with urban morphological analysis is also becoming increasingly important, as it can capture non-linear and context-sensitive drivers of commercial success and community vitality. Noteworthy papers include: Parking, Perception, and Retail: Street-Level Determinants of Community Vitality in Harbin, which proposes an interpretable, image-based framework to examine the impact of street-level features on retail performance and user satisfaction. Interpretable Multimodal Framework for Human-Centered Street Assessment: Integrating Visual-Language Models for Perceptual Urban Diagnostics, which introduces a novel framework that fuses a vision transformer with a large language model, enabling interpretable dual-output assessment of streetscapes.