Up to date information about road surface conditions is a valuable asset that is globally relevant. An efficient system that can collect road surface data allows for targeted and timely use of resources in case of various road surface related problems. The most prevalent use cases are road maintenance planning, emergency repairs and weather effects including floods or snow. Using the data can also benefit road users through enhancing route planning. This study relates to the concept of utilizing a network of UAVs for road surface data provision. While UAVs are more than capable of collecting road surface data, developing a complete system, that can do so safely and efficiently is challenging. To aid the development of such a system, this study focuses on virtualization opportunities through urban environment modelling. Virtually representing road networks in a relevant environment enables more efficient development of the system by reducing cost, risk and greatly improving mission availability and execution performance. The presented methodology is capable of rapid model generation of arbitrary urban regions based on GIS data at a medium fidelity level. The various road surface features (potholes in this example) can be injected into this model and can be detected with UAVs using the Gazebo SITL simulation environment and a computer vision recognition system. The proposed approach demonstrates the viability of further system development within a virtual environment.