View article

Signalized intersections are one of the most critical elements of the transportation network. Although in many cases there are some fairly simple ways of improving intersection performance, identifying which type of improvement is right for a given intersection requires in-depth knowledge of its operation that frequently can only be obtained from field observation. However, existing field methods are personnel intensive and often inaccurate in cases where queue lengths exceed 30 vehicles per lane. The University of Arizona is working on a new method that overcomes the personnel and queue length limitations of other field methods by using aerial platforms equipped with imaging devices. In the spring of 2002 three preliminary experiments were performed in Tucson to test the new method. The results of the experiments, as well as a comparison with results from the HCM methodology and an assessment of the capabilities and limitations of the technique are presented. One of the advantages of this method is that it derives control delay from field measurements instead of using estimates of traffic variables. The method also allows the observation of other important operational parameters such as queue length, saturation flow rate and arrival type. A tentative procedure for automating data reduction, as well as a summary of the outstanding issues in this area of research are also introduced.