Currently available bicycle tail-lights and head lights were evaluated in a field experiment in which the distances at which stationary cyclists could be detected was measured. Observers in the experiment drove around a circuit of suburban streets at night, with cyclists located at selected points on the route. It was possible to vary the background against which particular devices were viewed to a limited extent. In general, tail-lights performed well, especially a flashing tail-light which was detected at long distances against a complex background. Head lights in general performed poorly. This is particularly worrying as an analysis of bicycle crashes in the State of Victoria indicated that a much higher proportion of night time bicycle crashes related to frontal rather than rear conspicuity. There is little objective information about the effectiveness of conspicuity aids for cyclists, despite a growing awareness of the very large numbers of bicycle accidents and associated costs. On behalf of the State Bicycle Committee, VicRoads commissioned the Australian Road Research Board (now ARRB Transport Research) to carry out a study of bicycle conspicuity aids which reviewed existing literature, tested how well a range of modern conspicuity aids could be seen under realistic conditions and related the conspicuity of the devices to physical measures of their reflectivity or luminance.
Abstract