The current research landscape in urban sensing and smart city applications is witnessing significant advancements, particularly in the areas of data fusion, real-time processing, and multi-modal integration. Innovations are being driven by the need to manage and interpret vast amounts of heterogeneous data from diverse sources, such as GPS, RFID, and image processing, to enhance urban operations and services. A notable trend is the development of unified models capable of handling both individual-level mobility data (trajectories) and population-level mobility data (traffic states), addressing the limitations of traditional single-modality approaches. These models are designed to improve the accuracy and scalability of urban applications, such as traffic management and pedestrian safety, by leveraging advanced machine learning techniques and spatiotemporal data analysis. Additionally, there is a growing emphasis on creating synthetic datasets and sandbox environments to validate and calibrate processing algorithms, ensuring robustness and reliability in data-driven urban applications. The integration of V2X technologies is also gaining traction, with frameworks that focus on spatiotemporal fusion for multi-agent perception and prediction, enhancing the capabilities of autonomous systems in complex urban environments. Overall, the field is moving towards more integrated, scalable, and intelligent solutions that bridge the gap between sensing infrastructure and actionable urban intelligence.