Development of an Enhanced Decision-Making Tool for Pavement Management Using a Neural Network Pattern-Recognition Algorithm

Abstract

Network-level pavement maintenance and rehabilitation (M&R) decisions are typically based on pavement surface condition thresholds, which may lead to inadequate treatment selection and loss of funds. The objective of this study is to develop a one-step enhanced decision-making tool that considers both structural and functional pavement conditions in treatment selection. To achieve this objective, structural number (SN) based on rolling wheel deflectometer (RWD) measurements was used to calculate a pavement structural health indicator known as the structural condition index (SCI). Two enhanced decision flowcharts were developed, for the functional classes of arterials and collectors, using the SCI. An artificial neural network (ANN)–based pattern recognition system was then trained and validated using pavement condition data and RWD measurements–based SN to arrive at the most optimum M&R decisions. Based on the results of the analysis, it is concluded that the proposed ANN-based pattern recognition systems can be used successfully as a pavement management system (PMS) decision-making tool at the network level. The developed model showed an acceptable overall M&R decision prediction accuracy of 96.9% and a precision range from 93.2 to 1%. Furthermore, the developed tool is time efficient because it allows PMS engineers to define the final enhanced M&R decisions in only one step.