Infrastructure condition forecasting using neural networks : Schwartz, C.W. Infrastructure Planning and Management Conference Proceedings, Denver, Colorado (United States), 21–23 Jun. 1993. pp. 282–284. Edited by J.L. Gifford, D.R. Vzarski and S. McNeil. American Society of Civil Engineers (1993). ISBN 0-87262-917-1

Abstract

Condition forecasting models are critical components of any infrastructure management system. infrastructure condition assessed through periodic inspections which assign ratings to various functional or distress categories based on the severity and extent of deterioration. These detailed distress ratings are often combined into a single composite condition index, e.g., the Pavement Condition Index, or PCI, used in some pavement management systems (Shahin, 1982). In order to determine the optimal timing and methods for planned maintenance and rehabilitation actions, the composite condition index--or, ideally, the individual distress ratings--must be projected over some multilayer planning horizon.