Drivers’ gap acceptance during parking maneuvers as a basis for initiating driving actions in automated vehicles

Abstract

As a social system, traffic requires the coordination between different involved road users, particularly in shared spaces (e.g., parking areas) with fewer statutory regulations and hence more ambiguous encounters. By anticipating the development of driving scenes and adapting own driving actions, drivers maintain individually varying safety margins, such as accepted time gaps, to surrounding traffic participants. However, drivers’ gap acceptance (GA) was shown to be influenced by situational factors and driver characteristics. In order to apply expectable interactions in mixed traffic, involving manual and automated vehicles (AVs), intuitive gaps in traffic flow need to be selected by AVs when initiating parking actions. Derived time gaps of manual drivers could provide a basic orientation for intuitive driving functions in AVs. Thus, the current experimental study assessed manual drivers’ GA in a parking area scenario. Real-world video material was used to investigate the effects of interaction partners, their encountering speeds and drivers’ age and sensation seeking on GA. In the study, 42 participants (including two age groups: 18–30 years vs ≥ 45 years) indicated their GA when initiating a left-turn parking maneuver in front of approaching vehicles. The study examined three interaction partners (motorcycle, passenger car, truck), each approaching at six different speed levels (10–35 km/h). The results revealed that selected time gaps increased as vehicle size increased, with the largest GA for the truck and the smallest gaps for the motorcycle. Moreover, time gaps decreased at higher approaching speeds of the interaction partners. In addition, older participants selected more conservative gaps and participants with higher sensation seeking scores accepted lower time gaps to initiate a parking maneuver. The results of the current study revealed that there is no one single time gap that is always considered as appropriate to initiate parking actions. For intuitive interaction capabilities in AVs and to support smooth encounters in mixed traffic, the speed of the approaching interaction partners should be considered as an essential situational factor when initiating parking maneuvers. To consider individual driving style preferences (e.g., due to driver age) and support user acceptance, selectable driving style profiles should be available in AVs as a form of customization.