Traffic accidents are most likely consequences of normal driving behavior, which has not been adapted to a respective situational setting and thus facilitated an accident. Within the project reported here, a taxonomy for categorizing drivers' allegedly erroneous behaviour in such circumstances has been developed. It integrates aspects of human information processing as well as the three error types postulated by Rasmussen (1983). Decision nodes with linked questions support classification decisions with reference to seven error types. In summary, the taxonomy is widely applicable for classifying driver errors and error free behavior in manoeuvres, critical incidents, near accidents, and accidents; as such it forms a base for a structured (video) analysis, e.g. for harmonizing the analysis of FOT- and NDS-data or for in-depth accident analyses. The taxonomy is complemented by a scheme integrating error antecedent conditions, which in line with Hollnagel's (1998) genotypes build - in their respective manifestation - triggering or facilitating conditions for errors, near accidents or accidents. The scheme is an expandable, structured document, which is open for amendments in accordance with (new) research results. Together with the taxonomy the scheme builds a base for deriving driver assistance demand and other supportive measures, for generating research questions, and for collecting research results in a structured way. The present report addresses the FOT- and NDS-community as well as in general all readers interested in transportation psychology and research. The working steps and results of the taxonomy development are reported in eight chapters. (Author/publisher)
Abstract