This paper presents the results of a study to evaluate the substantial changes to the street network in the `Village' of West Lafayette, IN, USA, made in May 1991, and possible alternatives to the revised street network. The changes were made in an attempt to improve pedestrian safety, and ease the considerable traffic congestion. Streets in The Village could not be widened without removing buildings; instead, several of them were converted from two-way to one-way operation. Although this led to some improvements, many retailers complained of a major reduction in sales. Before May 1991, extensive volume counts and a licence-plated survey had been conducted, and a simple link-node network model of The Village network model of The Village had been used to recommend the changes. When the street changes were re-evaluated, the model was recalibrated, to enable accurate tests of the possible alternatives, some of which were quite different from either the original or the new street configuration. In the course of seeking more precise calibration of a traffic assignment model, the authors learned several important lessons; for example, the well-known commercial travel demand software packages used did not have sufficient detail or flexibility for their needs.
Samenvatting