The authors of another paper, Lee and Ran, defined that "winter maintenance in the area of transportation encompasses those activities that attempt to keep roadways in proper condition to travel during winter weather events." As technologies develop, winter maintenance becomes more complicated and requires more strategic planning for better performance. Recently, many state agencies in the U.S. have made rigorous efforts to log and archive their winter maintenance activities. It is expected that careful examination of the archived data will reveal much valuable information, which in turn, will enable them to quantify and improve winter maintenance. This study analyzed major contributing variables to "speed recovery duration," which is the time interval from when a winter weather event starts to when the roadway becomes wet condition using a regression tree algorithm "GUIDE" (Generalized, Unbiased, Interaction, Detection and Estimation). A total of 12,526 snow storm reports in Wisconsin since 1998 was analyzed. To represent the performance of winter maintenance, a variable, "PassRatio," was developed. It is defined as the time interval between when a "winter weather event starts" and when the "roadway becomes passable" over the duration of the winter weather event. According to the analysis with regression tree algorithm, GUIDE, it seems that "CrewDelayed" is a significant independent variable that has a significant association with "PassRatio."
Abstract