Evaluating the effects of variable message signs in an urban network

Richards, Andy (2004) Evaluating the effects of variable message signs in an urban network. University of Southampton, Doctoral Thesis.

Abstract

Variable Message Signs (VMS) provide real-time traffic information to drivers, with the aim of improving network efficiency, especially in incident situations. The effects of VMS are entirely dependent upon behavioural response, both in terms of interpreting the message and in reacting to the information displayed. An evaluation of the signs will aid the understanding of these effects, enabling a more successful long-term development of VMS strategies. Although previous research is well-documented regarding VMS located in inter-urban areas is not widespread. This study was focused on an urban network in the city of Southampton, UK. This site was chosen both because of the innovative large-scale application of an urban VMS system and the availability of traffic data derived from the SCOOT UTC detectors.

In this research, three different evaluation methodologies were employed. Questionnaires were initially distributed to some 20,000 households to recruit commuters to Southampton to partake in travel diary surveys over a one-week period. In addition, attitudinal data was obtained regarding perceived use and effectiveness of VMS. Later case studies used quantitative loop detector data collected during incident scenarios to assess driver response to VMS messages. Finally, a simulation model was used to quantify the network effects of the VMS.

The main findings from the questionnaires concluded that the VMS messages were well understood and legible, although only a small number of travel diary respondents actually diverted as a result of VMS information. In addition, the perceived usefulness of the VMS was inconclusive with no evidence of an increased usefulness during incident situations. Traffic data was collected and analysed for seven different incident locations and positive driver responses were measured for all, with diversions ranging from 2% - 30%, although not all were solely attributable to VMS information. No relationships between severity of VMS message and driver response could be identified. Simulation found that if the diversionary route had sufficient spare capacity, substantial VMS benefits were potentially attainable, although these benefits were not distributed equally.

This record has no associated files available for download.

Contributors

Download statistics