SQeX: a scenario-based and question-driven framework for designing explanatory interfaces in autonomous vehicles

As autonomous vehicles (AVs) become prevalent, clear communication of their decision-making processes is vital for driver trust. Despite advancements in explainable artificial intelligence (XAI), existing methods for designing explanatory interfaces often lack systematic support for scenario complexity and user-centred requirement elicitation, both of which are critical in human-autonomous vehicle interaction (HAVI). To address these gaps, we propose the SQeX framework, a Scenario-based and Question-driven approach that systematically extracts explanation requirements and creates tailored designs. Using the SQeX design canvas, designers and researchers visualised and structured the design process based on real-world AV scenarios. Explanatory interfaces created during the co-design workshop were evaluated through a video-based online questionnaire (N = 160). The results provide formative evidence that these designs can enhance users' trust and understanding of driving decisions within the tested scenarios. Scenario patterns and design implications were identified during the design review, while qualitative evaluations highlighted the framework's effectiveness in addressing scenario complexity, aligning with user goals and fostering collaboration.