Application of Fuzzy Logic and Clustering Techniques for Pavement Maintenance

Abstract

When a large number of pavement stretches are to be maintained, the decision making becomes complicated and prioritization of individual stretches is not really useful in taking maintenance management decisions. In such situations, grouping or clustering the pavement stretches having similar quantified distress characteristics would be the practical and effective approach. Since a number of distresses are observed on pavements and usually they are being represented in different levels of severity and extent, it gets rather hard to cluster them. The problem becomes further aggravated as the weights of various distresses also play an important role and it is difficult to express them objectively. In this paper, a methodology has been proposed to quantify the pavement distresses and clustering the pavement stretches. The technique has been explained with the aid of a case study carried out in a few selected stretches in the state of Rajasthan, India.