Group , Poster #245, Seismology

A Low-Cost Internet-of-Things Device for Earthquake Early Warning

Vivien He, & Robert W. Clayton
Poster Image: 

Poster Presentation

2021 SCEC Annual Meeting, Poster #245, SCEC Contribution #11283 VIEW PDF
A low-cost Internet-of-Things (IoT) device for earthquake early warning (EEW) was presented at the Seismological Society of America (SSA) Annual Meeting in April 2021. The IoT device is integrated with a geophone, alarm, wifi, and customized hardware and software, and costs under $100 to make. The device is about the size of a Rubik’s cube, runs at 100 SPS, uses a 32-bit ADC, and is controllable via a smartphone or computer. The device successfully detected earthquakes across Los Angeles metro area and issued EEW alerts through the onboard alarm as on-site warning and text messages to local subscribers as regional warning.

This report summarizes recent progress on the IoT EEW...
device. The device was tested at a single-story house in Los Angeles (LA) continuously for 9 months from September 2020 to May 2021, during which the device detected all earthquakes over M 3.0 magnitude around Los Angeles, as well as nearby earthquakes down to M 2.3. The waveforms were validated against outputs of broadband seismometers in Southern California Seismic Network and accelerometer-based stations in Community Seismic Network.

By analyzing data of all 12 detected earthquakes, empirical formulas were developed linking ground motion amplitude and earthquake magnitude, with and without epicenter distance. For the 8 earthquakes in the LA area with magnitude ranging from M 2.3 to M 4.5, and epicenter distance ~13 to 40 km, the formula showed highly linear relationship between amplitude and magnitude without considering distance. This could be used to quickly estimate magnitude (or effective local impact level) using a single device.

Ten of these IoT EEW devices have been built and tested, which showed performance consistency for the same earthquake. The devices were also tested for time sync accuracy among the devices, which range from low milliseconds on the public internet down to sub-millisecond level on a Local Area Network synced to the same time server. The power consumption was measured with the device continuously running the EEW program in a steady state. The average power consumption was ~620 mW. The low power consumption is helpful as the device can be powered by a battery or power bank as an Uninterruptible Power Supply during power outage or be used in the field.

Future plans for this research include deploying the devices to test an IoT EEW network across LA and exploring other applications such as earthquake education.

SHOW MORE