Using card-sorting to determine information requirements for remote operators of high-level automated vehicles

High-level connected and automated vehicles (CAVs) may encounter edge cases which require a human to intervene remotely, particularly during the initial implementation and deployment. The nature of these interventions will vary, demanding different operator tasks and competencies—and, in turn, distinct information requirements (IRs). We present an empirically derived, prioritised classification of IRs for remote operators of high-level automated vehicles and translate these into actionable interface design guidance. Seventy potential IRs were initially collated based on established use cases and supporting literature. Following iterative refinement, 39 IRs were evaluated in a card sorting exercise, in which 20 participants (recruited as pairs) were presented with two remote operation use-cases (taken from ten possible use-cases) and asked to group and prioritise the cards—each showing a different information requirement. Two hundred and sixty-five categories were initially generated. These were subsequently refined using cluster analysis (similarity matrices, dendrograms) and verified using inductive thematic analysis, resulting in eleven distinct classifications: Occupant Information; CAV Interior; CAV Systems; CAV Movement Status; Communication Channels and Connection; Route Information; Immediate Surroundings, Vehicles and Events; Wider Surroundings, Traffic and Events; Obstacles and Obstructions; Other Agents; and Secondary Vehicle. The insights were then applied, supported by interface design principles, to inform the layout of a future remote operator workstation.