Daytime Photometric Requirements for Pedestrian Signals

Abstract

Forty-eight senior citizens participated in a field study of the visibility of letters and symbols in pedestrian traffic signals. Subjects were asked to identify signal messages from distances of 18.3 m and 29.3 m, with signal voltage set at 100 percent, 75 percent, and 50 percent of full power. Incandescent, fiber-optic, and light-emitting diode commercially available pedestrian signals were tested, including 22.9-cm and 30.5-cm rectangular signal housings and two round red-amber-green signals with symbol masks. Each subject was asked to identify the signal’s location in the test stimuli array, to name the signal’s display configuration (Walk, Don’t Walk, walking person, or hand), and to assess the signal’s brightness on a five-point scale. Analyses also were conducted on the percentage of responses about “too bright” signals and subject uncertainty about the signal message. Testing was conducted only on bright sunny days but did not include the worst-case condition of direct sunlight on the signal face. The analysis of recognition, uncertainty, and “too bright” responses suggested that a signal intensity of 25 cd minimizes the frequency of both “too bright” and uncertain responses regardless of size, distance, or technology, or whether the message is symbol or text. The data further suggest that 22.9-cm incandescent signals provide sufficient visibility with less phantom effect than 30.5-cm signals.