Quake detection system using smartphone-based wireless sensor network for early warning

Detecting disruptive events using Commercial off-the-self (COTS) sensors like the ones embedded in smartphones is a difficult challenge but also an interesting opportunity. In this paper, we present a reliable and scalable hierarchical architecture of smartphones acting as opportunistic sensor nodes. Using a low energy-consumption application, we have used the smartphones inertial sensor as an accelerograph. The deployed smartphones and the application form a low-cost wireless sensor network, that detects, analyzes and notifies a seismic-peak. The systems optimizes the distributed calculations in the smartphones; communication capabilities and integration in order to provide extra time for early warning in disaster scenarios (e.g. earthquakes), although the architecture may be extended to other disruptive and rare events. We propose an innovative real-time solution which considers time and spatial analyzes, not present in another works, making it more precise and customizable, coupling it to the features of the geographical zone, network and resources, so as providing evidence of the feasibility of earthquake early warning using a dense network of cell phones. The architecture has been validated by extensive evaluation and the most relevant result has been the improvement in notifications delivery about a seismic-peak 12 seconds earlier than previous works in the epicenter zone, and a reduction in the number of false positives. Additionally the proposed architecture includes a post-event management to help users and strengthen coordination between aid-agencies in order to optimize human resources and time to implement measures in order to eliminate negative effects on the population.