Emergency management has attracted the focus of mobile computing research in the last years due to the flexibility that it provides in critical scenarios. The lack of a pre-existing infrastructure or even a communication breakdown are important issues that mobile computing can deal with. In addition, Semantic Web techniques to handle the data in these scenarios, such as knowledge representation and reasoning, have been proven useful. In the SHERLOCK project we are developing a general system to provide users with interesting Location-Based Services (LBSs) that can help in emergency management. SHERLOCK executes on mobile devices and leverages their communication mechanisms to exchange information among them. The system uses: 1) ontologies and semantic reasoners to handle knowledge about LBSs and interesting objects, and 2) mobile agents to balance CPU consumption and communication load.