ABSTRACT
Civil engineers face numerous challenges in monitoring roadway deterioration and in assuring pavement maintenance to the preferred level of serviceability without disrupting traffic flows. Presented is a study on data collection and analysis by use of sensors from smartphones utilizing On-Board Diagnostic (OBD-II) devices while vehicles are in movement, for the detection and classification of roadway pavement surface anomalies. Robust regression analysis and bagged trees classification model are used to compliment smartphone-based data collection. The recommended system is based on readily available, low-cost and sufficiently accurate technology, and can be used in crowd-sourced applications for pavement monitoring. Further, the proposed methodology has been field-tested (detection and classification of two types of pavement surface anomalies, exhibiting accuracy levels higher than 90%) and at this time it is expanded to include larger datasets and a bigger number of pavement defect types.
