Modelling and evaluation of the effects of traffic safety measures Comparative analysis of driving assistance systems and road infrastructure

University dissertation from Lunds Tekniska Högskola Institutionen för Teknik och Samhälle Trafik och Väg

Abstract: Improvement of road traffic safety is an important policy issue. Various categories of meas-ures are available to effectuate this. The thesis focuses on two categories: road infrastructure and driv-ing assistance systems. A relevant policy question is if certain driving assistance systems may be de-sirable substitutes for infrastructure measures, in terms of traffic safety effects and other relevant as-pects. The objective of this thesis is to develop an analytical method for comparative analysis of the traffic safety effects of these measures with dissimilar nature, as well as to identify an adequate method for policy evaluation, and to design, based on this method, an approach for strategic evaluation of alternative implementation schemes of different measures, which can take into account the results from the comparative analysis method and/or other method(s) concerning traffic safety effects, and other relevant impacts and constraints. For this comparative analysis and evaluation the following re-search topics are formulated: (1) identification of the functional substitutability relationships between driving assistance systems and infrastructure measures; (2) modelling of the fundamental schema of the influence of a measure on road traffic safety; (3) design of a method for comparative analysis of measures of different character; and (4) elaboration of a systematic analytical framework for linking decision making at the aggregated global level concerning overall effects of alternatives in a network, and decision making focusing at the local level concerning traffic safety effects of one or several measures at an intersection or road section. Through identification of a benchmark, i.e. a comprehen-sive set of traffic safety principles, and related operational traffic safety requirements, the functional substitutability relationships between these two categories of measures are qualified. In addition, a quantitative causal chain (QCC) model is developed based on a breakdown of the causal chain be-tween measures and traffic safety effects. Based on the QCC model, a method for comparative analy-sis is derived for assessing safety performance at a local level. Furthermore, the thesis elaborates a sys-tematic analytical framework for policy evaluation, and proposes a normalisation based evaluation method, named grey relational analysis (GRA), to aggregate data at the global level. Application of GRA provides a clear-cut ranking of strategic alternatives for improving traffic safety.