Automated shuttles and ‘negotiation in motion’ – A qualitative meta-synthesis of spatial interactions with human road users

Since automated vehicles (AVs) were first introduced in public imagination, the stated goal of developers has been to develop vehicles that would eventually operate in diverse contexts like any other vehicle. To understand what this entails in real-life traffic, data regarding interactions were extracted from three separately run trials of automated shuttles in low-speed contexts with human road users in Denmark (2018–21) using a qualitative meta-synthesis approach. The underlying data consists of field observations, interviews with road users, geolocalized event registrations, video tracking data, and responses to open-ended surveys. The synthesis suggests that 1) dynamic negotiation of space and timing, 2) handling of situational and traffic system ambiguity, and 3) human road user learning, go beyond what should simply be attributed to a transitory immaturity of the technology. Road users expect other road users to engage in a deeply social negotiation of space and timing. When AVs fail to negotiate, traffic flow is interrupted, and road users express confusion and impatience, until they develop strategies to obstruct or move around the shuttles. We discuss implications on planning in low-car environments.