The primary objective of bridge management is to provide tools for the optimal allocation of economic resources for transportation agencies and highway operators, in order to maintain an appropriate level of safety and serviceability of their bridge stock. Traditionally most bridge management is based on the qualitative evaluation of the condition through visual inspection. To overcome the limits of this approach, we carry out research aimed to incorporate the concepts of risk and reliability into prioritization criteria, and to optimize the inspection, maintenance and repair strategy. This activity includes, on one side, search for of prioritization and management algorithms and, on the other, development of tools such as software and procedures, to make these concepts easily available to road operators.

The improvements in computational performance occurred in the past few years enhanced the capabilities of decision support systems. Each decision support system is a comprehensive model, which incorporates monitoring data, structural models, heuristics, inspection outcomes, direct costs, indirect costs and risk profiles in order to maximize the utility of the involved resources. By combining pattern recognition, statistical decision theory and Bayesian networks, we develop interactive programs that help in managing major infrastructures. Decision support systems provide a rank of the optimal decisions selected by minimizing the risk. In this way, decision-makers do not need to perform frequent manual time-consuming data analysis because they are notified only if an anomaly in the structural condition is detected.