Developing Socially Acceptable Autonomous Vehicles

Abstract

Recognizing that the movement of cars on the road involves inherently social action, Nissan hired a team of social scientists to lead research for the development of autonomous vehicles (AVs) that engage with pedestrians, bicyclists, and other cars in a socially acceptable manner. We are expected to provide results that can be implemented into algorithms, resulting in a challenge to our social science perspective: How do we translate what are observably social practices into implementable algorithms when road use practices are so often contingent on the particulars of a situation, and these situations defy easy categorization and generalization? This case study explores how our cross‐disciplinary engagements have proceeded. A particular challenge for our efforts is the limitations of the technology in making observational distinctions that socially acceptable driving necessitates. We also illustrate some of the significant successes we've already achieved, including the identification of road use practices that are translatable into AV software and the development of a concept, called the Intention Indicator, for how the AV might communicate with other road users. We continue to investigate road use to uncover and describe the ways in which the social interpretation of the world can enhance the design and behavior of AVs.

Investigating the Influence of External Car Displays on Pedestrians' Crossing Behavior in Virtual Reality

“…Further aspects include the need to help pedestrians, in particular children, to identify whether a vehicle is (highly) automated or manual [6]. The idea of socially acceptable AV behaviour has also been raised, noting that often human motorists and pedestrians break traffic rules [39].…”

When will most cars be able to drive fully automatically? Projections of 18,970 survey respondents

“…That is a very very complex problem to solve." (Sierhuis, 2016, 49:38; see also Vinkhuyzen & Cefkin, 2016).…”

On the Road with an Autonomous Passenger Shuttle

“…How people make sense of AV actions can be complicated in high-density areas such as busy dual use vehicle/pedestrian roadways. Communicating the intentions of an AVs forthcoming actions has been identified as a priority for improving Human-AV interaction specifically in relation to driver-vehicle, passenger, and pedestrian interaction [8,14,24]. We examine another communication requirement that involves making a request to another road user to modify their behavior.…”

“…To date, most research has been conducted in laboratory settings using prototypes, future scenarios, and in some cases onsite at test facilities [20]. However, as AVs move from test facilities to public roadways, researchers now have an opportunity to investigate Human-AV interaction in real world traffic situations [1,11,23,24].…”

“…Recent studies in AV Human-Machine Interaction (HMI) provide insights into how algorithms may need to be calibrated to take into account the norms of conduct between pedestrians and vehicles [24], the types of information drivers might need prior to an AV action [7], and implications for transforming the role and skills of drivers [17]. With so much valuable research underway, it is worth noting that enquiries into Human-AV interaction are nascent, with in-car research focusing primarily on driver-vehicle, and passenger-vehicle interaction in ordinary cars with AV technology added to them.…”