In recent years the high cost of constructing and maintaining overhead guide signs has prompted the highway community to explore less expensive signing methods. In particular, California suggested the possibility of modifying or relaxing the requirement in the Manual on Uniform Traffic Control Devices (Section 2F-13) that "the background of all overhead signs that are not independently illuminated shall be reflectorized." A review of the literature (FHWA/RD-84/087), covering over 100 papers dealing with sign visibility, found the following three principal issues not resolved: • The comparative conspicuity between nonilluminated signs with opaque backgrounds and fully reflectorized or illuminated signs. • The speed and accuracy of motorist responses. • The inherent value of a color coding scheme. This study was undertaken to review and investigate these specific problem areas. The results indicate that color (i.e., green background of the illuminated or reftectorized sign versus the black appearance of a nonilluminated, opaque background sign) enhances sign detectability, while somewhat degrading internal contrast and reducing legibility about 5 percent. The general consensus is that color coding appears to help traffic safety, especially where signs must compete with advertising signs and other lights in the visual field. The first laboratory study was an experiment using a static 35-mm slide presentation of overhead guide signs in typical freeway environments. The subject was required to read and follow directional information given on typical guide signs by moving a control lever. The purpose of this study was to determine situations where the driver had problems detecting overhead signs and to determine whether color or additional brightness aided the driver in finding the sign. The data indicated that, where a sign was fully visible, there was little difference in detection rates. However, when the signs were partially obscured by physical obstructions (bridges) or other traffic, color, not brightness, was the key factor in accounting for high detection rates. Following the static experiments, an interactive driving simulator was used to study the driver's ability to respond in a dynamic, driving situation. Figure 1 shows a cumulative distribution of the distances at which drivers responded to the exit information for each of the sign background color/luminance conditions. In the simulator, the signs first become visible at 750 ft prior to the sign. The results were similar to those obtained in the static experiments with green signs performing better than black backgrounds and brighter signs performing better than dim ones. These differences, while statistically significant, were very small and there appears to be little practical difference. There was almost no difference in the results unless the signs were partially obscured; thus these results really apply only to conditions with significant amounts of obscuration present. However, these experiments did not require the driver to interact with other vehicles, as would be normal in freeway driving. This interaction with other traffic would take time away from the sign reading task, making it more difficult. The results, therefore, should be viewed as an optimistic representation of the true driver response. In a real-world traffic situation, it is very likely that the differences in sign detection and driver response would have been much greater. (Author/publisher)
Samenvatting