Developing the potential of micro-simulation modelling for traffic safety assessment

The idea of using micro-simulation modelling for road traffic safety assessment is still a controversial issue in the traffic research community. A recent investigation sponsored by the European Union’s 4th Framework Program revealed that there is an implicit and substantiated need for the development of micro-simulation systems designed for the assessment of traffic safety. Despite this recognised need, there appears to be a number of considerable practical and conceptual problems that have prolonged the introduction of such systems despite advances in computer power and capacity, modelling techniques, and data-collection methods. In particular, the typical simplified “normative ” model of driver behaviour that is sufficient for the study of traffic system efficiency, must be revised to allow representative levels of variance in road-user communication and interaction. It is conceivable that behavioural variance can be introduced through a naturalistic modelling of perception, decision-making and action-taking processes, thereby allowing “less-than-perfect ” behaviour and the possibility for various types of typical driver errors to occur. A second main issue concerns the identification of safety indicators that can be used to create a comparable and situation specific safety assessment profile. The safety profile emanating from a real-world situation must be reproduced from the microscopic behaviour of road-users in the logical simulator environment given the bounds of normal statistical error variance. This paper describes current work on the SINDI-project at the Centre for Traffic Simulation Research, Stockholm, which aims to break new ground in the use of micro-simulation for traffic safety assessment through the development of a detailed “nanoscopic ” model of driver behaviour, and the establishment of a safety profile based on identified safety indicators. The project focuses primarily on safety related to the interactions of different types of road-users at urban intersections, and the possibilities to improve traffic safety through the use of intelligent transport systems