Diss. Leidschendam : Stichting Wetenschappelijk Onderzoek Verkeersveiligheid, SWOV, 2008

Most of the research on traffic behaviour has been focused on the individual road user, despite the fact that road users rarely encounter traffic situations in which they are not confronted with other road users. An associated finding is that most models of the driving task represent the driving task from the perspective of one individual road user, rather than incorporating the interactive nature of traffic behaviour. To achieve an understanding of the way road users are able to gear their behaviour to one another this thesis included the following research questions (Chapter 1): - What is the role of expectancy in interaction behaviour at intersections? - What is the influence of expectancy on traffic safety at intersections? - What are the implications of human interaction behaviour for ADAS? Chapter 2 provided a theoretical framework for the research discussed in subsequent chapters and centred on the concept of expectancy. Chapter 3 discussed the range of research methods applied throughout the research conducted for this thesis. Chapter 4 dealt with the research of a more explorative nature and was particularly aimed at exploring the concept of expectancies in interaction situations (at intersections). Chapter 5 discussed the first of two driving simulator experiments. To study interaction behaviour, two driving simulators were linked to allow the participant in the first simulator to interact with the experimenter driving in the second simulator. The research for Chapter 6 again employed two linked driving simulators since the variation between similar situations (which is also found in real life traffic situations) seemed to be more realistically reproduced using the linked simulators. The results of the previous experiment suggested that the time available to both interaction partners to successfully cross the intersection (i.e. interaction space) could help to explain the effects of uncertainty of expectancies on traffic safety. This second experiment was aimed at further studying the concept of interaction space and the effects on both interaction safety and efficiency. Chapter 7 discussed the main findings of the research conducted throughout the thesis and considered the implications of these findings in several contexts. Based on the research discussed above, the model postulated in Chapter 2 was revised and presented in Chapter 7 Chapter 7 also discussed the merits of using linked driving simulators when studying interaction behaviour, allowing for more natural interaction situations compared to when using a traditional (single) driving simulator where behaviour of the interaction partners is pre-programmed. Chapter 7 ends with the overall conclusion that the research presented in this thesis has shown that road users are able to cope rather well with situations where other road users' behaviour conflicted with the right of way regulations, illustrating that road users are indeed quite capable of compensating for deviating behaviour of other road users. Also, the use of linked simulators proved to be a rather promising method to study interaction behaviour systematically but also intensively. As ADAS will also influence interaction behaviour it is stressed that a greater understanding of all aspects of interactive behaviour in traffic is required to be able to exploit the potential of ADAS to the fullest.