The need is noted for improved pavement maintenance and rehabilitation systems that can function with incomplete and/or poor qualitydata. Expert systems can deal with knowledge manipulation, and is divided into a generalized solution strategy component (inference engine) and the knowledge base. Other subsystems usually included in such systems are the knowledge acquisiton subsystem, the explanation subsystem, and the user interface. Expert systems applications in pavement maintenance management are listed and described. The details are described of pamex, an expert system for maintenance management of flexible pavements. The system builds on the pavement condition indicators devised by jugo and the decision tree knowledge formulationwhich he created. The indicators combined information obtained fromride quality, structural condition, and functional performance surveys. These were then used to develop and refine a sequence of decision trees for evaluating pavement performance, identifying pavement problems and their probable causes, and recommending appropriate corrective measures. Validation and system enhancements are also described.
Abstract