This paper presents some results of a local collaborative study, whose researchers examined the effect of typical urban lighting levels on visual perception and reaction times. Its findings challenge some of the conventional assumptions included in traffic legislation and advisory codes. The paper's first part discusses some statistics on night driving. The number of accidents per mile driven and their severity are both increased at night. There is compelling evidence that the main reason for this is low illumination at night. It seems likely that drivers' perceptual errors are even more common at low light levels. Thus we need to learn more about the nature of perceptual errors that lead to accidents. At present, no vision test can identify drivers prone to make the types of perceptual errors leading to accidents. The paper's second part discusses the physiology of the eye at low light levels, and the third part describes experiments based on the relatively poor form of vision mediated by rod cells. There is a critical illuminance level, where rods dominate perception, at the cost of detecting movement and gaps. Data on reaction times shows that reaction to small low-contrast objects is much slower than to large high-contrast objects; this corresponds to quite large increases in stopping distances.
Abstract