A Smart Device Using Low-Cost Sensors to Detect Earthquakes

Due to the significant development of hardware and software technologies in mobile and embedded computing, small hand-held devices such as smartphones have been used to detect earthquakes. In the past few years, there were efforts to detect earthquakes using a low-cost acceleration sensors inside a smartphone. However, it is not only costly to use a smartphone for merely detecting earthquakes, but also a waste of computing resources because smartphones comes with a powerful CPU, plentiful memory, and other auxiliary sensors. Also, a smartphone always needs to be connected to Internet. In this paper, we introduce a stand-alone earthquake detection device equipped with an acceleration sensor and Wi-Fi, First, we systematically evaluated a set of acceleration sensors by assessing their performance and accuracy to select the most suitable acceleration sensor, and then developed an designated device that can detect earthquakes and send alerts to nearby devices. Furthermore, to distinguish earthquakes from daily motions, we employed a Artificial Neural Network(ANN) technique with the earthquake dataset obtained from the Pohang earthquake in South Korea, 2017. Our result shows that a low-cost acceleration sensor can be used to detect an earthquake, thereby enhancing the public safety of communities vulnerable to earthquakes.

I. Introduction

In the last decade, mobile devices including smartphones and Internet of things (IoT) have made a significant development in both hardware and software. Such mobile devices are often equipped with a variety of sensors such as an accelerometer, a gyroscope, a GPS, a magnitometer, etc., and these sensors are widely used to get useful information (e.g., location-aware services), to improve user experiences (e.g., game), to keep track of their daily life (e.g., health applications), etc. Furthermore, recently there were attempts to use sensors inside smartphone to detect an earthquake. In particular, MyShake can detect earthquakes by capturing sensor data from smartphones [1].