A fully integrated hardware–software platform consisting of a wheeled robot equipped with an embedded vision system and an image-to-text model, enabling autonomous movement, visual odometry, environmental observation, and verbal description of visual inputs.