The selection of highway safety improvement projects is indeed a complex task. Usually, numerous hazardous locations are candidates for improvement, with several acceptable projects available at each location. Funds are seldom sufficient to implement projects at each hazardous location. Therefore, it is necessary that the project selection process function under the restriction of a limited budget. The classical economic analysis techniques such as rate of return and benefit cost cannot be relied upon to determine the optimum project combination with this restriction. The technique of dynamic programming is ideal for application to this problem. An example consisting of three hazardous locations illustrates the pitfalls associated with previously recommended project selection processes and demonstrates the manner in which dynamic programming leads directly to the optimum safety project combination.
Abstract