A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Civil Engineering

The Maintenance & Planning Divisions of the Thailand Department of Highways (DOH) are both responsible for road maintenance which includes planning, budgeting and maintenance scheduling. DOH has implemented pavement management program called "Thailand Pavement Management System (TPMS)" in cooperation with a consulting firm from Canada in 1989. The data collection process for the existing TPMS concentrates more on the aggregate level of the pavement condition and other relevant data. TPMS recommends the pavement treatment for each 200-m subsection of DOH network at the network level. However, in the actual practice, the suggested treatments are overridden by the judgment of DOH practitioners due to several project-level constraints ranging from technical level to political issues. This study tries to improve the gap between the TPMS recommendation and the DOH actual practice. Thailand Pavement Maintenance Decision Support System (TPMDSS) is developed in this study to help select appropriate pavement treatments taking into account distress characteristics, roughness measurement and traffic volume. TPMDSS follows a logical progression, synthesizing various inputs in order to analyze for pavement problems and then recommend a feasible set of treatments for each TPMS subsection. Instead of traditional schedule-based treatment, TPMDSS employs a cause-based strategy. TPMDSS architecture provides an integrated framework between database management systems (DBMS) and expert systems (ES) in order to efficiently utilize available pavement information in the TPMS database. TPMDSS treatment decision model comprises four major modules which are distress evaluation, problem identification, problem evaluation and treatment determination. A TPMDSS logic-based approach is implemented with decision matrices, system rules and scoring systems. TPMDSS is validated from three analytical perspectives - overall system, improvement section and specific section. Three primary techniques are used in the validation process - model-generated data, field inspection, and expert interview sessions. Two-step treatment screening algorithm is proposed to recommend a final set of treatment for an improvement section. The validated results have revealed that TPMDSS analysis yields quite reasonable solutions in overall aspects.

Thesis (M.Eng.) - Asian Institute of Technology 2000