An Efficient Model for Planning Bus Routes in Communities with Populations Between 20,000 and 250,000

Abstract

The paper discusses a flexible, inexpensive, interactive computer model specifically designed to act as an aid for planning routes for conventional bus systems in communities with populations between 20,000 and 250,000. The characteristics of these communities, the desires of decision makers, the cost and availability of data, and the special problems inherent in route planning are examined and a useful routing model is designed. A computer implementation of the structure of its major component was developed. Usage costs are examined. (An implementation of the entire model has not yet been completed.) The completed model will not replace community leaders or transit managers; instead, it will enable a decision maker to quickly and inexpensively test a potential route or operating decision without actually implementing it. To test the route, the decision maker need only input the operating decision or route choice. The routing model automatically traces the route through the community and estimates whom the route serves, how well, and at what cost. It predicts how many people will use the route, their geographic location, possibly their income distribution, the travel time, wait time, the walking distance they experience, and the operating cost of the route. Based on these estimates, the decision maker can alter any decision and immediately test the new route. In the course of model design, two new demand models are formulated: an alternative logit model for probability estimation or modal split calculations, and an extension of the intervening opportunities model for desire prediction or trip generation.