Choosing a maintenance strategy is a major challenge for road operators because of its implications on the required budget and the quality of the road network. However, a major problem has to be solved before we can compare the performance of different strategies: on many real networks, essential data for simulating the evolution of the distress of the pavement for the coming years and hence simulating the strategies in order to evaluate them, are missing or highly uncertain. In this paper, we introduce a decision support tool for evaluating maintenance strategies on roads networks with incomplete data. First, this tool is used to build a Virtual Road Network (VRN) that represents the real network, and on which plausible values are allocated to missing or uncertain data. Then, it is used to simulate alternative strategies. To this end, the distress state of the pavement is evolved using an appropriate evolution model and each strategy is applied in order to determine the associated maintenance interventions. The different strategies are thus compared (on the network scale) on the criteria of cost and resulting distress state of the pavement in order to choose the best solution.
