Measuring Behavior 2005: Williams and Rakic

An observation and rating scheme for driving situations

M. Williams and M. Rakic

Institute of Ergonomics, Darmstadt University of Technology, Darmstadt, Germany DaimlerChryslerAG, Sindelfingen, Germany

The aim of the study is first of all to develop an observation scheme for driving situations which should both give the possibility of integrating different types of data and be practicable. A driving situation consists of one or more components; each component can be described with specific parameters. During the whole driving situation the components and their characteristic parameters do not change. In case of driving on motorway the following components were identified: free driving, driving behind another car, driving in front of another car, overtaking, being overtaken. The components could be connected by one of the following transitions: lane changing of the reference vehicle or lane change manoeuvres of other vehicles. For example, the component ‘driving behind another vehicle’ can be described by the following parameters: headway, time headway, relative velocity, variation of the speed of the vehicle driving ahead, type of vehicle driving ahead, and number of vehicles driving in the observed lane. The relevant areas in front, behind, on the left side, and on the right side (all related to the reference car) have been defined and divided in fields. Each of these fields is also a matrix cell. By filling in the matrix with the number of vehicles which are in each field at a certain moment we get a picture of that driving situation and of its components. In the next step parameters like speed (absolute and relative) and distances (e.g. headway) are associated to each vehicle. The parameters are obtained from the reference vehicle sensors, object recognition system, and frame by frame video analysis The practicability has been checked by using data from an experiment conducted on a twolanes German motorway. Subsequently, the identified driving situations have been rated in terms of information sources, information processing, decision-making and handle by using three keys (importance, accuracy, speediness). The aim was to identify possible situations clusters, i.e. situations which pose similar demands to the driver.

Paper presented at Measuring Behavior 2005 , 5^th International Conference on Methods and Techniques in Behavioral Research, 30 August - 2 September 2005, Wageningen, The Netherlands.