LAUSR

Abstract Automated vehicles have become a popular topic of conversation. Initially, these conversations were limited to technology developers, innovators and engineers, as they worked to progressed the various technologies and systems that are required to create automated vehicles. Then, over time, these conversations extended to other communities; lawyers, insurers, planners, policymakers, social scientists, and various publics all began hearing, and talking about automated vehicles – also known as ‘driverless’, ‘self-driving’, and ‘autonomous’ vehicles. Levels of automation emerged as a way to depict gradations or categories of autonomy, with tasks divided between those for the machine and those for humans. In this paper, we critically reflect upon the dominance of levels of automation – up to seven sequential ‘steps’ - proposed by a number of industry organisations. Focusing on the Society of Automotive Engineers (SAE) Standard J3016, we signal the intended and unintended performative effects of these levels. We argue that current discourses on automated vehicles have been underpinned by a techno-centric, expert-dominated logic, and point to the benefits of more dispersed, geographically contingent, and socio-technical perspectives in re-framing the dominant discourse and allowing for more nuanced spatial and temporal understandings on future systems of (automated) mobility.