The research aims to develop a system that can control a mobile robot remotely. Robot is equipped with only a webcam and a wireless network adapter. During the research it was found out that many buildings are equipped with video surveillance, but not all buildings are ideal for video surveillance. There are places that cannot be seen in video surveillance camera. Therefore, this paper proposes a method for remotely controlled robot in bounded space, which uses a robot, equipped with a web camera, and video surveillance camera data. The method is based on the simultaneous localization and mapping algorithm to accelerate the creation of maps on the computer. Processing is performed with special software, which also is described in this article. Processing result is a virtual space, which is used to determine the robot position. Using wireless network, web camera data are sent to the computer for their treatment. It is possible to extend the proposed system to manage multiple robots, each equipped with a web camera or some other sensors, as well as the use of ad-hoc network. The system also provides cases when mapping fails, e.g. it is not possible to receive data from video surveillance cameras due to failure in communication, or room is not equipped with surveillance cameras. When map is created it is possible to take a variety of other tasks for which the localization is required, for example, the robot must go from one office to another. There is one drawback in the proposed system – it is hard , or impossible to detect transparent, i.e. glass, office walls. This disadvantage can be avoided with additional ultrasonic distance sensors. These ultrasonic sensor measurements are then processed and associated with the camera data. The main advantage of this system is that it can be extended to create a 3D space model to match the real space.