Driving in Roundabouts: Why a Different Theory of Expert Cognition in Social Driving is Needed for Self-driving Cars
Raúl Sánchez-García 1and Duarte Araújo 2
This paper considers the study of driving skill and performance of human drivers as a key domain to improve the design and the “expertise” of self-driving cars. Concerted driving with others in roundabouts is a special kind of social expertise —understood as competence in ordinary, mundane activities— deployed by members of society. We illustrate our theoretical arguments about how human cognition works in driving contexts with the analysis of three paradigmatic examples through the lenses of the ecological framework. Based on these analyses, we discuss the need for a more robust and realist theory for developing self-driving cars as potential complications arising from interactions between human drivers and self-driving cars cannot be solved with current socio-cognitive models for decision-making and social coordination. The ecological framework considers that drivers’ exploratory activities rely on the utilization of affordances, rather than on the internal processing of information, which is currently the default assumption guiding self-driving car design. The ecological approach assumes that drivers are embedded agents that act within increasingly complex technological envelopes. Such a framework could be used to investigate how digital driving landscapes may be closely tailored to the drivers’ activities. Finally, future research on driving design should investigate how affordances can be matched with emerging digital technologies for reducing accidents and improving traffic flow.