The Computing Continuum drives holistic system integration, especially in the road sector, where there's a growing need for comprehensive Decision Support Systems in operation and maintenance strategies based on sensor data. These systems offer dynamic assessments, critical for managing road assets vulnerable to climate change and extreme weather events. The resulting data uncertainties, including variations in data veracity and variability require timely decision-making for coordination and adaptiveness of the system as a whole. We introduce an urgent computing architecture for a dynamic risk management framework, discussing its related challenges and software design patterns.