WIT Press

Ranking And Selecting Dangerous Accident Locations: Case Study


Free (open access)








420 kb

Paper DOI



WIT Press


K. Geurts, G. Wets, T. Brijs & K. Vanhoof


In Flanders (Belgium), approximately 1014 accident locations are currently considered as 'dangerous'. These 'dangerous' accident sites are selected by means of historic accident records for the period 1997-1999. More specifically, a combination of weighting values, respectively 1 for each light injury, 3 for each serious injury and 5 for each deadly injury, is used to calculate the priority score for each accident location. In this paper, a sensitivity analysis is performed to investigate how big the impact is on the current ranking of accident sites when alternative ranking criteria are used. More specifically, we only take into account the most serious injury per accident and use a valuation of casualties based on direct costs, indirect costs and validation for human suffering to give weight to the accidents. This valuation results in the weighting values 1_7_33 when the most severe injury respectively concerns a light, serious or deadly injury. Additionally, we generate probability plots, based on estimates from a hierarchical Bayes model, in order to visualize the estimated probability that a location will be ranked as dangerous. Results showed that combining these ranking criteria will have a big impact on the selection and ranking of dangerous accident locations. In particular, when selecting the 800 most dangerous accident sites of all accident locations, 82% of these locations will differ from the current selection. Considering this impact quantity, we want to sensitise government to carefully choose the criteria for ranking and selecting accident locations without stating that the criterion used in this paper should be preferred to the currently used ranking method. Keywords: ranking dangerous locations, injury weighting values, hierarchical Bayes, probability plots.


ranking dangerous locations, injury weighting values, hierarchical Bayes, probability plots.