← Back to Sessions

A Consumer Internet-of-Things Device for On-Site and Regional Earthquake Early Warning

Session: Earthquake Early Warning Live in California! Current Status and Challenges I

Type: Oral

Date: 4/23/2021

Presentation Time: 10:00 AM Pacific

Description: 

While there has been significant development in earthquake early warning (EEW) systems, there are still challenges, one being large blind zones near the epicenter where alerts are most needed. Another issue is slow/missed alerts due to low seismometer density and variability in on-site intensity and user alert threshold preference. Even with emerging less-expensive devices targeting citizen scientists, costs of seismometers, installation, and maintenance are still high. This contributes to limited EEW growth, exacerbated by heavy dependence on public funding.

This research pursues a different consumer-based approach, developing a low-cost on-site IoT seismometer that can be mass deployed at homes and buildings, much like a standard smoke detector. In addition to an alarm at the user’s location with personalized settings, the IoT device sends text messages to local subscribers, uploads waveform data for further research, and is controllable via smartphone. A large number of these devices can form a dense network for regional EEW.

This paper presents such an IoT device integrated with seismometer, alarm, wifi, software, and packaging, for under $100 at prototype volume and much less with increased production. The device is palm-sized (7cm x 7cm x 5cm) and samples ground motion at 100Hz with a geophone and a 32-bit ADC. The device successfully detected recent earthquakes in Southern California (for example, the M4.6 earthquake on September 19, 2020 UTC) with high signal-to-noise ratio, sounded the alarm, sent text message alerts to subscribers, and produced standard earthquake waveform miniSEED files. The alert level is customizable by each consumer, and the alarm is managed by a smartphone application or on the device itself.

Further developments are underway, such as deploying many devices for better regional EEW, as well as an AI-based software to recognize ground motion signatures of human activities for security applications to further enable mass adoption.

Presenting Author: Vivien He

Student Presenter: Yes

Authors