OECD Workshop on knowledge-based expert systems in transportation. Vol. 2

This study was undertaken by a Scientific Expert Group from nine OECD Member countries, with the aim to review the state of the art for planning, developing, implementing and maintaining knowl edge based expert systems.The intent of this report is to provide guidance to roadway decision makers who may be required to address such questions such as, "Does this technology have application in roadway engineering and operations?", "What types of problems can be addressed?", "How are these computer programs different from other computer programs?"The report is organized in four chapters.An introduction is provided in Chapter I; the problems amenable to solution using knowledge based expert systems are given in Chapter II; the architecture and control strategies of knowledge based expert systems are discussed in Chapter III; and guidance for planning, developing, testing, implementing and maintaining knowledge based expert systems is provided in Chapter IV.Knowledge based expert systems are computer programs that attempt to embody the heuristics (private knowledge and rules of thumb gained through experienCe) of experts. These programs offer a means to capture the knowledge and experience of current professionals and organize, save, and apply this information to be used as assistants, decision aids or training aids by other less experienced professionals.This provides a valuable tool for the transportation specialist. These issues and the conditions necessary for the development of successful knowledge based expert systems are outlined in Chapter I, Introduction.The architecture and components of knowledge based expert systems are described in Chapter II, Definition of Knowledge Based Expert Systems.Knowledge-based expert systems differ from conventional programs in several ways.The most significant is that knowledge based expert systems include concepts and a simulation of an expert's reasoning processes that cannot be expressed in equations as knowledge is represented in conventional or algorithmic programs. This and the separation of knowledge from the control strategies are the key features that allow knowledgebased expert systems to function with incomplete data, in contrast to algorithmic programs which will not.Knowledge based expert systems also have the capability to explain their reasoning and logic which make them much easier to understand and more valuable as training tools. In Chapter III, Components of Knowledge Based Expert Systems, the components are described in detail and the knowledge representation schemes are outlined. Chapter IV, Developing Knowledge Based Expert Systems, constitutes a practical guideline for planning, developing, testing, distributing, and maintaining knowledge based expert systems.It is critical in the planning stages that the necessary resources are secured and that the proper development team is assembled.Both the successes and the failures in the development of knowledge based expert systems can usually be traced back to the planning phase of development.