Report numberRA-MOW-2011-003
TitleDecision criteria for the reconstruction of hot spots
SubtitleAn empirical analysis based on decision tree analysis
AuthorsTim De Ceunynck
Niels Janssen
Stijn Daniels
Tom Brijs
Published byPolicy Research Centre for Mobility and Public Works, track Traffic Safety 2007-2011
Number of pages85
Document languageDutch
Partner(s)Universiteit Hasselt
Work packageOther: Infrastructure and space

Based on accident registration, Flemish roads locations with a high accident risk have been identified, the so-called hot spots. The Flemish government has committed itself to investigate and adjust 800 of these locations to improve road safety. To facilitate the process of analysis for these locations, the Agentschap Wegen en Verkeer (AWV) of the Flemish Ministry of Mobility and Public Works formulated a manual to provide the government with a framework of methods and principles, which resulted in a decision tree. The intention of this report is to discover the actual choices decision makers have made for the adjustment of the hot spots, and compare these with the decision tree that is proposed by the AWV manual.


The decision tree proposed by the AWV manual consists of three steps. The procedure starts with a traffic safety analysis based on the AVOC-method (Approach for dealing with Accident Concentration Locations). In the second step, the transportation planning and spatial context are involved in the process. This is about the road classification, and whether the hot spot is located within a built-up area or not. Based on this, a number of configuration types are proposed. In the third step, these configuration types are checked with the traffic flows and spatial preconditions of the location.


Decision tree analysis is a data mining technique that indicates how the value of a dependent variable can be predicted based on the values of a set of independent variables. It visualizes these relations in a tree structure. The available dataset is repeatedly split based on the independent variable that provides the best (most homogeneous) split in two subgroups.


The analyses show that the traffic volume is a crucial variable for determining the configuration of the location. In case of a high traffic volume, a conflict-free traffic light regulation or a configuration that avoids crossing traffic flows is often chosen. In case of a lower traffic volume, traffic signals or a priority-rule intersection are often preferred. However, the AWV manual advises to start from the road classification. So, in practice, decision-makers often start from the current situation, rather than from the transportation planning ideal. A second important finding is that the existing configuration type is often maintained during the adjustment process. So, the adjustment of these hot spots often involves modifications to the existing layout of the location. Despite the fact that the decision making process in reality does not correspond with the structure that is proposed by the manual, the chosen intersection type after the reconstruction very often corresponds with the table of prototype solutions.

DownloadPDF icon RA-MOW-2011-003.pdf


The Policy Research Centre for Traffic Safety carries out policy relevant scientific research under the authority of the Flemish Government. The Centre is the result of a

cooperation between Hasselt University, KU Leuven and VITO, the Flemish Institute for Technological Research.


Leuven vito