Driving with intelligent vehicles: driving behaviour with Adaptive Cruise Control and the acceptance by individual drivers

This thesis focuses on the following research questions: What are the effects of driver support systems on driving behaviour? To what extent will driver support systems be accepted by individual drivers? To what extent will driving behaviour and acceptance be determined by individual differences? Based on a literature study an integrated hierarchical framework of driver behaviour is proposed that is structured along the three different driver behaviour levels and carries aspects of most of the theoretical models that were discussed. The research described in this thesis uses the hierarchical framework of driver behaviour as a tool to make predictions about needs, driving styles and dntrol system. The results of a questionnaire survey show that the respondents see both positive and negative aspects to the four driver support systems. About 50% of the respondents indicates that traffic safety will increase when driver support systems are introduced. Respondents were negative about having taken over control of driving by a system. As was predicted from hierarchical framework of driver behaviour, the results show a relationship between needs and some of the self reported driving styles. Needs and driving styles can be seen as stable traits that can be different between drivers but not very much between trips. That is, drivers don't change their needs and accompanying driving styles when going on different trips like commuter or holiday trips. This conclusion is used in the experimental studies described in the following chapters. An important selection criterion for the subjects participating in the experiments was their self-reported driving style. This gives the opportunity to predict their behaviour and reactions to driving with driver support systems, because we can expect this to be in line with their higher level needs. The results of two driving simulator experiments show that the driving style Speed was a good predictor of actual driving behaviour, whereas the driving style Focus did not show any effect on actual driving behaviour as measured in the simulator. In general all drivers adapted their behaviour according to expectation, irrespective of their predetermined driving style. In the light of the higher speeds, smaller time headways and more efficient merging manoeuvres that were found when participants drove with ACC, the system seems a promising development in terms of traffic efficiency. However, the higher speeds and smaller headways that were found when driving with an ACC, might point to possible negative effects on traffic safety. From the simulator experiment on rural roads the conclusion is that we should be careful with the introduction of ACC's to road types other than the motorway. Drivers will probably use their ACC equally often on both rural roads and on motorways, because of the high acceptance, but behavioural adaptation may seriously deteriorate traffic safety, particularly on the rural road. ---- Conclusions: This thesis shows that Adaptive Cruise Control seems to be a promising new technology in traffic. That is, under certain conditions and limitations. In the first place, the experiments clearly demonstrated that driving behaviour with ACC leads to positive effects in terms of traffic efficiency. Driving with ACC reduces speed variability and initial individual differences in driving behaviour on motorways, which harmonises traffic. On top of this, acceptance results indicate that the headway adopted by the ACC does not influence the preferences of drivers. Even very short headways (0.6-1.0 s), which are shown to increase motorway capacity by Minderhoud (1999), are accepted very well. Sed very well. Sey Minderhoud (1999), are accepted very well. Seondly, a more harmonised traffic pattern can also reduce the number of accidents and thereby increase traffic safety. In addition to increasing traffic efficiency ACC's could therefore also increase traffic safety