Operating Performance of Pole Mounted Curb-Side Automated Pedestrian Detectors at Signalized Intersections

摘要

The equitable accommodation of pedestrians is a critical issue in transportation engineering and planning. With Canada's population aging steadily and the associated disability rate expected to rise significantly over the coming decade, accessible transportation and travel resources will be in increasing demand. The need or requirement for accessibility in transportation is often semi-independent of the percentage of the population with disabilities. In Canada, the United States, and in many other countries, the requirement for accessibility in transportation comes from the statutes that these societies have enacted requiring equitable and reasonable accommodation for all. However, this is sometimes not enough to equitably accommodate all users of the system. In June of 2008, human rights complaints against a North American jurisdiction pertaining to the equitable accommodation of pedestrians at signalized intersections were settled. The complaints illustrated the lack of equitable access provided to pedestrians at signalized intersections as they are often required to press a pushbutton to activate the pedestrian signal to cross the intersection. This is not the case for motorized vehicles, which are either automatically detected or accommodated by signal timing plans. This paper presents the preliminary results from a year-long field performance analysis of three commercially available curb-side Automated Pedestrian Detectors (APDs). Pole mounted microwave, infrared, and combination video and infrared curb-side detectors were installed at two signalized intersections in downtown Winnipeg, Manitoba. The operating performances of the selected detectors are presented in sensitivity and selectivity rates corresponding to observed temperature, weather, and day and night operation for winter months. A sampled 84 hours of video and recorded APD output from the five month winter period at one of the sites provided 6,310 detection events, including 811 events where crossing pedestrians required detection. Overall, the APDs analyzed are generally sensitive for sending calls for valid pedestrians, but selectivity rates tend to remain below 50 percent. Of all the conditions, and throughout the ranges observed, the infrared detector has the highest sensitivity and lowest selectivity of the three detectors analyzed. The combined infrared/video detector has the second highest sensitivity and highest selectivity. The microwave detector has the lowest sensitivity and second highest selectivity.