Every time vehicles leave the the road they are likely to collide with fixed roadside objects, such as lighting columns, signposts, etc. Traffic and highway engineers have always been faced with the difficult task of making important decisions regarding the selection and implemention of roadside safety improvements. To facilitate their decisions, a knowledge-based system called ROSDIS (Roadside Safety Design and Improvement System) has been developed. This paper describes the knowledge-based system which constitutes a significant extension of an existing model and uses system safety methodology. An analysis of single vehicle roadside accident data has been carried out with the objective to extract information from the available data, so it can be utilised by the system. The system enables the user to experiment with different configurations of fixed roadside objects and select the configuration which minimizes the probability of collision. The explanation facility of the system allows the user to see why a question is asked by the system and how the probability of collision was reached. In addition to presenting ROSDIS, the paper will discuss the differences between the present study in the Tyne and Wear area (United Kingdom) and previous studies that have been conducted in the United States. This work was also reported in the OECD workshop on knowledge-based expert systems in transportation, Vol 1, Espoo, June 1990 (IRRD 841231).
Abstract