There are compelling research findings that using landmarks such as traffic lights as navigation cues can enhance the acceptability, usability and safety of vehicle navigation systems. In this study, 30 participants undertook an on-road trial using simulated navigation instructions that omitted any reference to distance-to-turn, and instead incorporated information requiring drivers to count traffic lights and pedestrian lights. The study results confirmed that traffic lights can be used as navigation cues even when the density of traffic lights along a stretch of road is high and that the term "set of traffic lights" results in slightly better performance than "traffic lights" alone. Factors other than density were found to have an effect on navigation performance. More errors were made if the target turn was at a pedestrian light and if the turn was visually complex or concealed. In the absence of a navigation task, most drivers did not differentiate between traffic and pedestrian lights when counting the occurrences and only experienced difficulty in counting when the number exceeded 4. Implications for the design of future navigation systems are highlighted. (A) "Reprinted with permission from Elsevier".
Abstract