The conducted research is focused on design and development of intelligent mobile robots - self-guided vehicles. The overall goal is to develop highly robust cognitive control system for this kind of robots and to bring novel ideas into particular solutions. For central tasks are considered: sensing of the environment, sensor data processing and data understanding, all leading towards automated world model building and/or updating. The used knowledge representations of the world are designed and optimized for planning of robot activities for self-navigation in real (indoor) environments. As such system requires highly complex control and decision-making the crucial parts is always the sensing and sensor data understanding.