Extending the Technique of License Plate Number Matching: Central Edinburgh Experience

T here is a growing interest in license plate number matching in the United Kingdom. This is partly because the planning emphasis has shifted to solving current traffic problems and partly because of the increasing popularity of traffic management models (e.g., CONTRAM and TRAFFICQ). The MVA Consultancy conducted one such survey recently in central Edinburgh, U.K. Edinburgh, the historic capital of Scotland, has a mixture of commercial, business, and residential activity that is centered predominantly around the late-Victorian new city center. Scope for improving traffic management in this part of central Edinburgh exists; however, the main concern up to now has been to disentangle the complex pattern of traffic demand. In particular, the origin-destination (O-D) pattern between the various feeder roads entering central Edinburgh conflicts with the activity to and within central Edinburgh itself. The O-D pattern is complicated by O-D changes during the course of the morning peak period and changes in vehicle routing. This situation, which is found to a greater or lesser extent in many U.K. towns and cities, is often alleviated by piecemeal infrastructure improvements. In the case of central Edinburgh, however, Lothian Regional Council has adopted a systematic approach by developing a central area traffic management model based on CONTRAM.’ The model requires a detailed description of the highway network, as well as a set of O-D matrices that reflect the buildup of traffic during the modeled period—in this case the morning and evening peak periods. The license plate number survey method was used to obtain the O-D matrices and to provide information on vehicle routings and travel times for model validation. The central Edinburgh study area (Figure 1) is larger than usually envisaged for license plate number surveys; however, a number of other features in the study show that the technique can successfully be extended to quite complex situations. The traffic movements themselves were complicated by a mix of on-street and metered parking, together with two large car parks, a bus station, and the railway terminal. Central Edinburgh was divided into a fine system of 52 zones and 26 cordon crossing points (Figure 1). To effectively model the delay at junctions, the O-D trip pattern had to be estimated for each tenminute time interval covering the twohour peak period. As conditions in the morning peak were different from those in the evening, a set of matrices was required for each peak period. The number of O-D movements to be estimated was therefore very high and the zone system very fine. The U.K. license plate number consists of 1. Up to three letters. Because these are allocated on a regional basis they were not recorded in-this survey.