I-Corps: Crowdsource sensing of earthquakes

University Of California-Berkeley, Berkeley CA

Investigators

Abstract

The broader impact/commercial potential of this I-Corps project it to reduce the impact of future earthquakes. This smartphone-based crowdsource sensing application provides earthquake alerts and shaking hazard data. The vision is to provide earthquake early warning around the globe, and collect data on building performance. Warnings can be used to slow and stop trains reducing the likelihood of derailment, to isolate hazardous chemicals and machinery to reduce damage, to move data to safe storage, and to signal individuals to move to a safe place. Shaking data recoded in buildings can also be used to assess the performance of a building during an earthquake. Recorded shaking in smaller non-damaging quakes can be used to predict building performance in larger events. The data recorded by application enabled phones can potentially be used by structural engineers to assess buildings, and can be used by the insurance industry to set premiums. This I-Corps project develops a global smartphone-based seismic network. The technology uses machine learning techniques to distinguish between earthquakes and other sources of shaking as recorded on personal/private smartphones, and is embedded in the application that has been downloaded by over 250,000 users around the world. When a phone detects an earthquake, it notifies a central server. The server confirms the earthquake based on multiple phone detection, and estimates the size and location. An alert could then be issued to users. At the same time, the phone records the ground shaking during the earthquake. This timeseries data can be used for a range of purposes including fundamental research on earthquake processes, and characterization of buildings.