Semantic annotations and formally grounded ontologies constitute flexible yet powerful methods of knowledge representation. Using them in a system allows to perform automated reasoning and can enhance the knowledge management. In the paper, we present a system for collaborative knowledge management, in which an ontology and ontological reasoning is used. The main objective of the application is to provide information for citizens about threats in an urban environment. The system integrates a database and an ontology for storing and inferring desired information. While a terminology of the traffic danger domain is described by the ontology, the location details of traffic conditions are stored in the database. During run-time, the ontology is populated with instances stored in the database and used by a Description Logic reasoner to infer new facts.