Development of a decision-making tool for bridge preservation and maintenance

Civil infrastructure is critical to the operation of transportation networks and many countries have mature asset portfolios requiring increasing amounts of investments to provide adequate capability and capacity for forecasted requirements. There are more than 617,000 bridges across the United States. Currently, 7.5% of the nation’s bridges are considered structurally deficient, meaning they are in “poor” condition. Federal and State agencies have been trying several ways to better manage and operate the bridges. A recent estimate for the nation’s backlog of bridge repair needs is $125 billion. There is a need to increase spending on bridge rehabilitation annually to improve the condition. Therefore, the nation needs a systematic program for bridge preservation like that embraced by many states, whereby existing deterioration is prioritized, and the focus is on preventive maintenance. The goal for this study is to develop a tool to determine deterioration models using Normal and Weibull Distributions as well as Machine Learning Applications for different types of bridges using the inspection data from more than 5000 Maryland bridges, from which project needs will be determined. These models are used in the development of a decision-making tool to allow users to compare different maintenance and repair scenarios of bridges and select the best plan to minimize the cost and maintain an acceptable bridge condition rating.