Kantenbouki VR: A Virtual Reality Authoring Tool for Learning Localized Weather Reporting

Localized weather reporting based on human observation is a skill practiced by fishermen, seafarers, airport ground staff, among others. Weather reporting largely depends on accurately identifying cloud types, and judging visibility and distances. This skill is developed by first-hand experience of various weather phenomena. As such, beginners would have difficulty practicing this reporting task without experiencing the phenomena, without proper guidance of experts, and without appropriate tools to support their learning. We present Kantenbouki VR, a virtual reality authoring tool, which can solve various issues in the process of learning weather reporting based on human observation. Using a 360° camera capture as an image background in the application, users can create annotations via freeform drawings accompanied by a cloud type label. To test our idea, we conducted an exploratory design study. Observations and results show that users learned how to use the system quickly, and approached the annotation tasks differently. Features that can be improved include the fidelity of the cloud images, as well as mapping of actions to controller buttons.