Sadeghi, L., Zhang, Y., Balmos, A., Krogmeier, J. V., & Haddock, J. E. (2016). Algorithm and software for proactive pothole repair (Joint Transportation Research Program Publication No. FHWA/IN/JTRP-2016/14). West Lafayette, IN: Purdue University. http://dx.doi.org/10.5703/1288284316337
Potholes are a common pavement distress, particularly appearing during the spring freeze-thaw period in northern climates. Potholes reduce ride quality, and if left unrepaired can lead to rapid pavement deterioration. Typically, when a pothole appears a repair crew is dispatched to place patch mixture in the hole with the hope that the patch will last until such time as a more permanent repair can be made. This reactive approach to potholes can often be too late to prevent further pavement damage and also makes it difficult for repairs crews to be scheduled in the most cost effective manner.
In this study, the relation between traffic loads combined with weather records, such as temperature, freeze-thaw cycles and the numbers of potholes requiring patching was investigated in an attempt to develop a model to predict pothole formation and distinguish the routes which are prone to pothole formation before the potholes begin to form. If pothole prediction were possible, this proactive approach would enable agencies to plan and schedule maintenance activities more cost and time effectively thus increasing ride safety and mobility.
To achieve the objective, four years of maintenance data from Indiana routes were collected and statistically analyzed to develop a model to estimate the probability of occurrence of a pothole due to annual average daily traffic and climate. The model indicates how significant traffic loads combined with weather condition influence the pothole. Also, although traffic loads and weather conditions are the essentials for potholes to form, the effect of pavement condition on the initiation of new potholes cannot be disregarded.
Additionally, this study began the development of a basic roadway distress evolution model by employing several standard statistical tools, such as, the empirical cumulative distribution functions (CDF) and the Kolmogorov-Smirnov (KS), to a pavement condition dataset. The goal of the model was to predict and rank areas of probable future concern by likelihood and severity. The resulting analysis showed promise but the data resolution was too low to achieve predictions on the desired fine scale.
pothole, predictive models, pavement, AADT, weather, cumulative distribution function
Joint Transportation Research Program
Indiana Department of Transportation
West Lafayette, Indiana
Date of this Version