The recent advancements in traffic management and safety have seen a shift towards more sophisticated models that integrate both spatial and temporal dependencies. Researchers are increasingly leveraging graph neural networks (GNNs) to capture complex interactions within road networks, leading to more accurate predictions of traffic flow and incident likelihood. Notably, the incorporation of dynamic factors such as road traffic measures and environmental knowledge has significantly improved the precision of cellular traffic predictions. Additionally, the field is witnessing innovative approaches to systemic safety analysis, which aim to provide more comprehensive risk evaluations and countermeasure selections. These developments underscore a trend towards more integrated and data-driven solutions in traffic management, promising enhanced efficiency and safety in urban and road networks.