Development of network-level linear programming optimization for pavement maintenance programming

Infrastructure systems in the US are in an urgent need of maintenance and rehabilitation. According to the most recent factsheet published by the American Society of Civil Engineers, one of the top five infrastructure concerns of today in the US is transportation systems. The major challenge facing maintenance managers in state Departments of Transportation (DOTs) today is to preserve the road networks at an acceptable level of serviceability subject to the stringent yearly maintenance and rehabilitation (M&R) budgets. Maintenance managers must allocate such limited budgets among competing alternatives. Absence of decision-making tools exacerbates the matter. This paper presents the development and implementation of a network-level pavement maintenance optimization model, which is established with the application of the Linear Programming algorithm and is subject to budget constraints and the agencies’ pavement performance goals in terms of total lane-miles in each pavement condition state. A decision-making tool is developed using Frontline Systems’ Premium Solver add-in for Microsoft Office Excel. This decision-making tool can compute the optimal amount of investment for each pavement treatment type in a given funding period. Pavement condition data pertaining to the Salem District within the Virginia Department of Transportation (VDOT) is used to test the model presented herein. Within this context, nine treatment types along with their corresponding unit prices ($/Lane-Mile), five pavement condition states, pavement deterioration rates, VDOT’s pavement performance goals, and available annual maintenance budget for the next fifteen years are defined. The results show that the network-level optimization model developed herein yields more efficient pavement maintenance programming than that obtained from the manual budget allocation process.