Bee-Nav is a robotic experimental dataset created to support the publication “Efficient robot navigation inspired by honeybee learning flights.” It contains flight logs, structured data files, and omni-directional images captured during both learning flights and navigation flights, with materials provided in formats including CSV, JPG, ULG, and DB3. The dataset is intended for research on bio-inspired robot and drone navigation, especially methods that use principles from honeybee learning behavior to improve autonomous movement, visual localization, and efficient route learning in artificial intelligence and robotics applications.