This project aimed to identify the most effective, low-cost speed-reducing measures for a selection of urban and rural environments. This was achieved by furthering our understanding of how drivers choose their speed (consciously or not) and what sensory cues we might use to alter this. The overall project approach consisted of three sequential steps: • Stage 1 – a review of previous experience with speed-reduction treatments; • Stage 2 – applying expert judgement to the information gathered in Stage 1 to design a range of treatments for each of the problem areas and road types; and • Stage 3 – simulator experiments to identify the most promising treatments. The review guided the experimental work in Stage 2 of the project by considering the relationship between speed indices and accident rates at both the individual driver level and the collective (road) level. In addition, the review outlined the stateof-the-art speed-reducing measures, concentrating on engineering design. This involved consultation of the current research literature, analysis of existing data and a survey of local authorities in the UK. One of the main conclusions here was that the methods that local authorities use to evaluate and report the success of interventions should be standardised and the results made easily accessible by other authorities. Potential speed-reducing treatments were then designed for both rural and urban environments. A variety of road layouts were assessed, including curves, junctions and straight sections of road. Forty-four different speed treatments were compared across the different road layouts, allowing us to evaluate which were more suited to particular road environments. It was found that treatments generally achieved better results if there was a reason for drivers to slow down (e.g. a sharp bend or a junction) or clearly defined change of speed limit (e.g. a village entry). Physical treatments, including those which featured haptic feedback, achieved the lowest spot speed on urban roads, while rural roads were found to be amenable to peripheral hatching and vehicle-activated signs. Hazard marker posts at rural bends were also found to be successful in lowering speeds. Peripheral hatching achieved better results than central hatching, presumably due to the fact that drivers were guided towards the centre of the carriageway thus affecting drivers’ safety margin. Treatments with high contrast in relation to their environmental setting also played a role in capturing drivers’ attention and hence facilitating speed perception. The best performing treatments for each road layout were then evaluated for their persistence. By repeating each treatment at three or four locations, we evaluated their durability. Persistence of the treatments tested in Phase 2 of the study generally demonstrated satisfactory results. The exceptions were pedestrian refuges on urban straight roads and warning signs with an advisory speed on rural lanes which showed less prominent persistence. To conclude, we suggest that there is certainly further scope to use driving simulator studies for this type of research as treatments can be compared directly. However, further methodological work should be undertaken to establish the most appropriate way of comparing schemes that differ vastly in their length and mode of operation. Finally, pilot sites which implement the best-performing treatments in this experiment should be established in order to undertake real-world evaluations with large numbers of vehicles. (Author/publisher)
Abstract