ARQUITECTURA DE UN SISTEMA DISTRIBUIDO PARA GESTIÓN DE EMERGENCIAS SÍSMICAS

[EN] This thesis project has a different and innovative approach to detect seismic events in real time gaining knowledge of the community through a hierarchical architecture in 3 layers: The first layer, a low-cost distributed network which takes advantage of the current huge trend; the smartphone; a multi-sensor, multi-network, multi-task device embedded into a small processing computer able to be reprogrammed, for example, in an "accelerograph" through an efficient in precision and power consumption Android application. The second layer called Intermediate Server, corresponds to a computer with sufficient hardware to handle incoming messages from users of the first layer and deduce on these samples, if a seismic peak has occurred, and if appropriate, notify in real-time to the users (smartphones) gaining extra time in making a better decision involving harm reduction, as well as economic and structural losses, and most importantly human lives losses. It considers spatial and temporal analyzes obtaining a customizable server to the specific characteristics of the area. Finally, the third layer called the Control Center is the place where all the information from the lower levels makes sense, being the leader in the post-event emergency management; it can extend to a bidirectional help: first, users (smartphones) attend to Control Center with multimedia information from their environment (comments, videos and images) thus achieving an overview of the disaster in order to efficiently manage the various aid-centers; and second, it assists users through their own smartphones, with information that a particular user unknown but the Control Center knows by other users: roads destroyed, landslides, roads offline support centers, places of danger, etc.). All validated through an extensive evaluation of each layer through seismic data obtained from the National Geophysical Institute of the National Polytechnic School in Ecuador (IGEPN), data which part of this research is based; obtaining promising and relevant results alerting until 12 seconds ahead at the epicenter, reducing false positives and this time could be increased by further afield. [ES] Este proyecto de Tesis presenta una diferente e innovadora pro-puesta para detectar eventos sísmicos en tiempo real obteniendo conocimiento de la comunidad mediante una arquitectura jerárquica en 3 capas: La primera capa, una red de sensores de bajo costo distribuida que aprovecha el gran boom electrónico actual, el sma