We have previously developed the Cognitive Robotic System, or CRS, to use the Soar cognitive architecture for the control of unmanned vehicles. The system was able to demonstrate the applicability of Soar for unmanned vehicles by controlling a ground vehicle to navigate to a target location using GPS in the presence of obstacles, including a cul-de-sac. However, only fairly basic information about the environment was available to Soar during this mission. The Soar cognitive architecture will be most interesting for the control of mobile robots when it is coupled to perception systems that are capable of extracting high level information from the environment which Soar can use in its reasoning. Therefore, recent work on the CRS has focused on the development of perceptual systems capable of generating relevant information from the environment for use by Soar agents in the CRS. This paper describes the integration of a system for generating occupancy grid maps in the CRS. The occupancy grid algorithm and a simulation environment used to develop and test the algorithm are described. Occupancy grids of common intersection types generated using the CRS and results from a system that uses fuzzy logic to detect common intersections from occupancy grids are shown.