Forest road planning: A knowledge-based approach

A methodology for automatically determining preliminary forest road corridor locations using artificial intelligence techniques is developed in this research. The main features of the methodology are an object-oriented data representation, a rule-based component that incorporates the expertise of transportation engineers, and a network analysis algorithm to aid in laying out forest roads. A computer program incorporating the methodology is described that uses the locations of existing roads, privately-owned land, areas which have been identified as needing road access, and areas that must be avoided by roads; and terrain information in the form of a digital terrain model. The methodology meets all of the original project goals, including the incorporation of heuristic road planning knowledge, the development of a structure for accessing and using matrix data with rules, the integration of algorithmic knowledge, and the ability for the user to challenge the results of the automatic planning process in order to determine the effects of specific changes in the road network. In addition, a graphics interface is used to interactively alter underlying data structures