Stereoscopic vision for robots: 3D position of objects are inferred from 2 or more simultanious view of the scene. The scene is mapped to monitor and using the inferred range information, robot is navigated in the environment with the help of GUI devices.