Mining and Analysis of Traffic Safety and Roadway Condition Data

Decision makers working to improve transportation systems must constantly balance the need to improve roadway safety through infrastructure investment with constraints of available resources. We have explored the additional information available to transportation decision makers by integrating two independent datasets, traffic accident and roadway pavement condition data. The roadway condition data are collected by the Alabama Department of Transportation. The traffic accident data are available in the University of Alabama’s Critical Analysis Reporting Environment (CARE) system. The Algorithm Development and Mining (ADaM) [Rushing] system, a data mining toolkit developed by the University of Alabama in Huntsville (UAH) Information Technology and Systems Center (ITSC) was used to examine the possible relationships between roadway data and traffic safety data. The preliminary study showed that discernable and predictable associations can be identified through analysis of the roadway condition and reported traffic accident events. The analysis results indicate that data mining is a successful method to perform advanced analysis to improve infrastructure investment decisions.