Knowledge based advisory system for flexible pavement routine maintenance

Hemaya, Paulette B. (1992) Knowledge based advisory system for flexible pavement routine maintenance. Masters thesis, Memorial University of Newfoundland.

[English] PDF (Migrated (PDF/A Conversion) from original format: (application/pdf)) - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (15MB) [English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. (Original Version)

Abstract

In recent years, pavement maintenance has become an area of major expenditure for many highway agencies. Making good maintenance decisions requires years of practical experience and judgement. Deciding on the best maintenance strategy is largely a subjective problem mainly because the benefits of a particular type of maintenance have not been quantified. Hence the problem does not lend itself to a traditional economic analysis where benefits and costs can be explicitly considered. As a result, the field seems perfectly suitable for an expert system application. -- Only four systems in flexible pavement and one system in rigid pavement were reported in the literature, namely, PARADIGM, PRESERVER, ROSE, ERASME, and Concrete Pavement Evaluation. These systems are limited in their application to particular jurisdictions and environmental conditions. The only three systems which consider routine maintenance are PRESERVER, ROSE and ERASME. PRESERVER runs only on a mainframe using the specially developed expert system programming environment OPS5. ROSE is a knowledge-based computer program intended for selecting and recommending routing and sealing of cracks only. ERASME is of limited scope and applicable only to pavements in France. Therefore, it is essential to develop an expert system program to facilitate pavement maintenance decision making in Newfoundland. -- This thesis describes the development of the Pavement Maintenance Advisory System (PMAS). It involves two areas of study. The first area is the process of knowledge acquisition. Knowledge was acquired for PMAS both from formal documents, documented case studies and from interviews with experts from the Newfoundland Department of Works, Services and Transportation. The interaction with the experts was invaluable and their expertise, which was not available in any literature, was encoded in the system. -- The second area of this study is the development of a computer program for selecting the appropriate maintenance strategies in cold/coastal regions. The system uses input including the type of distress, surface condition (severity and density), traffic volume, Riding Comfort Index, climate condition and outputs maintenance strategies with expected repair life in years as well as the associated total and unit cost of equipment, labors and materials. The cost comparison for the recommended maintenance strategies does not affect the internal decision making of the program. This feature simply allows the user to evaluate other maintenance options in case particular materials, laborers and equipment are available. The system is developed using two expert system shells; Exsys Professional on IBM-PC and compatibles and Instant Expert Plus on Apple Macintosh. PMAS utilizes backward chaining mechanisms to arrive at a conclusion. Confidence factors were introduced to indicate expert opinion on the suitability of the alternative maintenance strategies. The operational system was successfully tested by the experts through nine test case studies. PMAS can serve as a training tool for engineers who do not have much experience in selecting routine pavement maintenance strategies.

Actions (login required)

View Item