Performance measurement for highway winter maintenance operations

Many highway maintenance agencies are facing an increased pressure to utilize their limited resources while still achieving the optimum winter highway maintenance outcome. Also there is a tendency to privatize maintenance operation, in order to improve the road user's satisfaction by bringing more competition to winter maintenance operations. Given this context the purpose of this research is to develop an effective performance measurement system that can evaluate how well agencies have conducted winter maintenance activities to meet the road user's expectations of safety and mobility.

Though there have been performance measurement studies conducted in the winter maintenance area, few of them are comprehensive enough to evaluate winter maintenance outcomes, while at the same time taking storm severity, road system characteristics, and maintenance effort together into consideration. To address this deficiency, several particular challenges must be considered: first, how to evaluate the storm severity for individual storms; second, how to evaluate maintenance outcomes using a series of quantitative measures; and third, what are the appropriate targets that maintenance outcomes can be compared with, considering outcomes are sensitive to maintenance input, weather severity, road classifications, and traffic specifications. To address these questions: A storm severity index is developed; studies on effects of weather were quantitatively synthesized by meta-analysis; effects of weather and maintenance on road surface conditions are estimated by MLR; SEM (Structural Equation Modeling) is applied to estimate the direct and indirect effects of maintenance on mobility and Multiple Classification Analysis (MCA) was applied to estimate the contribution of winter maintenance to safety.

The final result of this research is an applicable winter maintenance performance measurement system. It informs maintenance agencies where they excel at and where improvements are needed for the specified goals. Further, the developed road surface condition prediction model can be used as a predictive tool to allow agencies to conduct "what if" experiments that will lead to optimization of maintenance practice over time.

The relative magnitudes of the effects of different maintenance methods on mobility and safety that is estimated by the models will enable agencies to assign priorities, and to compare maintenance outcomes based on the input resources.