Early Warning for geologic disasters became a subject of intensive research when observational and communication options allowed for dense sensor installation and real-time dissemination of signals and information. Initially most efforts focused on doing what seismologists had done before, just faster. In the current period real new research directions have emerged. Significant information, for instance on the earthquake process is contained in the first few seconds of observations and allow empirical extraction of warning messages. Novel sensors and intercommunication, specifically self-organizing networks, open new routes for implementing early warning systems. Models for community based early warning – capitalizing on these networks as well as on Google Earth popularity - are evolving. Science and tools for performance-based design of warning systems become available. Multiple-purpose systems for warning and for instance structural health monitoring have been designed. In short: Science moved towards the point, where it is translated in technologies.