Improving the Design of Information Provision for Advanced Driver Assistance Systems in Multi-warning Situations.

Abstract

Advanced Driver Assistance Systems (ADAS) are applicable to support drivers in making better decisions through warnings. Despite more types of ADAS becoming available within vehicles due to the progress in modern sensor technology such as connected vehicles and vehicular networks, it is still possible for the driver to experience the scenarios where multiple warnings are issued simultaneously. Also, such scenario can distract drivers if the warning is poorly designed. In order to have a better understanding of ADAS-driving scene, an ontology-based approach is adopted in this thesis. The created ontologies help to determine potential ADAS warnings for the design of multi-warning scenarios that may arise in the future. For a better understanding of how warnings are designed, initial experiments were performed to explore the effects of driving experience on the response made to multi- warning scenarios, with attention paid to the differences between experienced and inexperienced drivers in the way of responding to multi-warning modes. Compared to experienced drivers, inexperienced drivers showed more potential negative effects in their response to multiple warning modes. Specifically, there were more inappropriate driving behaviours performed by inexperienced drivers in the multi-warning mode, such as directly changing lanes when there is a need to slow down and give way. It is implied that the current warning design is possibly unsuited to inexperienced drivers in multi-warning mode. Therefore, the warning design is improved by considering two design philosophies. One focuses on the warnings with a higher priority (using a colour or flashing cue), and the other issues a main summary warning to direct the driver. According to the results, the main warning mode reduced the frequency of gaze change. However, it takes time to read the text, which prolongs the reaction, which may be the reason why this warning mode was found unfit for complex and/or urgent situations. Based on this measure, a recommendation is made for future ADAS design.