Explaining conceptual AI recommendations in urban air mobility: a comparison of storytelling video-based and text-based presentation formats for explanations

As digital assistants become increasingly integrated into safety-critical environments, explainable AI is expected to support operator understanding and decision-making. From a human-machine interface design perspective, how explanations are presented is an important consideration. This exploratory, design-focused study investigates the usefulness and user experience of storytelling video-based explanations compared with text-based explanations in a simulated emergency involving drone and Urban Air Mobility traffic management. Explanation content was structured using Construal Level Theory to provide incremental levels of detail. Nine licensed air traffic controllers participated in a mixed-methods, within-participants study, assuming the speculative role of UAM Coordinator responsible for low-altitude airspace safety. Eight of nine participants preferred the storytelling video format, describing it as more engaging and supportive of understanding and decision confidence. However, some perceived that the system had already made the decision, raising concerns about human agency. Video also offered limited flexibility for revisiting information, whereas text enabled self-paced reading and review but was perceived as more stressful to process. Rather than establishing the superiority of either format, the findings provide preliminary design insights into balancing engagement, user control, and human agency in explainable AI interfaces for future UAM operations.