Bridge strike reduction: the design and evaluation of visual warnings

The aim of this investigation was to consider the problem of road vehicles that strike rail-over-road bridges and how such incidents can be reduced. In particular, it examined the design of both the warning markings placed on bridges and the road signs situated some distance in front of the bridge, each warning of reduced clearances ahead. Initially, a literature review was conducted to reveal the nature of the problem, including the history of bridge strikes, previous attempts to quantify why bridges are hit, the cost of strikes, countermeasures to prevent them and the legal restrictions relevant to the area. It was concluded that no single countermeasure had been found to be effective when the cost and the legal restrictions were taken into account.A field assessment of possible causal factors was performed in order to compare a group of bridges that had been frequently struck, against a group of control bridges. It was found that frequently struck bridges generally were in busier environments (as might be expected), and in more visually complex environments where there were, on average, more advertisements nearby - thus more potential distractions were present at these sites. The research then considered what drivers look at when driving towards low bridges,specifically focusing on the amount of visual attention given to warning signs prior to a low bridge, and on the specific areas drivers look at in the final few seconds before reaching the bridge. It was found that the bridge warning signs and bridge markings performed badly on measures of visual attention. In addition, if an advertisement was placed on the top section of a bridge, this was looked at for a large proportion of the time - thus reducing the proportion of time which the drivers gave to other features of the environment. The development and evaluation of alternative bridge warning signs was then considered. Newly created and existing signs were evaluated on tests of comprehension and hazard perception. The results demonstrated that text-based versions of the warning sign with a yellow border performed best The development and evaluation of markings for low bridges were then examined by evaluating newly created and existing markings. The research focussed on their capacity to make a bridge appear lower than it really was - so influencing drivers' judgement of height when they approach such a bridge. The current low bridge marking standard achieved inferior scores on the experimental measures employed when compared to several of the alternative bridge marking designs that were developed. Finally, the investigation examined driver responses to both the bridge signs and markings. Using a virtual reality road scene, an experiment was performed which assessed if the existing and modified designs of the signs and markings identified earlier had any behavioural effects upon drivers as they approached the 'virtual' bridges. The addition of warning signs before the bridge was found to have no significant influence on subjects' decisions regarding stopping before the bridge. However, the type of markings displayed on the bridge did significantly affect their responses.