Taiwan start-up P-Waver announced during CES a new system of analysis data to design an Earthquake Early Warning (EEW) model to give advanced warning...
P-Waver is a start-up from National Center for Research on Earthquake Engineering of Taiwan National Lab. Pei- Yang Lin, the Founder of P-Waver, and his team announced during CES a new system of analysis data to design an Earthquake Early Warning (EEW) model for Taiwan Government to give advanced warning in the Island of Formosa (Taiwan).
During the Taiwan Tech Arena (TTA) with more of 100 start-ups, the team of P-Waver highlighted how massive data analysis is key to determining the efficiency of the system, along with state-of-the-art electronics. The service of team also covers structural safety monitoring systems and earthquake disaster prevention consultants as well as smart security solutions for people in buildings or at home through IoT devices or control systems. The spokesperson of P-waver said that is setting their vision to reinforce the security of the population and entire business. EEW technology has a strong impact on people’s lives and the financial losses that earthquakes can cause.
Technology for Earthquakes Science
The increasing urbanization and especially the strong dependence on the complex infrastructure for telecommunications and transport have led to a careful study of an early warning system for earthquakes by sending warnings to the population. An Earthquake Early Warning System (EEW) sends a real-time alert to people before the quake arrives. The development of such a system is a fundamental step in reducing the fear of the unknown and the unpredictable nature of earthquakes, while at the same time improving people’s safety.
When an earthquake occurs, seismic waves, including compression or longitudinal (P), transverse (S), and surface (R and L) waves, radiate outward from the epicenter. The faster but weaker P wave travels to nearby sensors, generating alarm signals to carry out protection operations before the arrival of the slower but stronger S waves and surface waves.
Operation of the ShakeAlert system used in Taiwan [Source USGS]
Sensors, big data, analytics, gateways and, more generally, all technological tools related to the IoT can be suitable solutions in cases of catastrophic events such as an earthquake. The evolution of artificial intelligence systems and the growing amount of available data are helping scientists develop models capable of simulating the earth’s crust movements with ever greater precision. Many studies combine artificial intelligence and neural networks to search for relationships between large amounts of data to gain time on a seismic event.
AI and Big data for Earthquake Early Warning Methods
During an earthquake, the health, mobility, security, and energy industries are likely to be severely affected. High-speed railways can benefit from the EEW system by initiating an early stop to protect people.
P-Waver built its EEW system based on seismic data from Taiwan’s Central Meteorological Bureau and AI technology comprising more than 250,000 earthquake models. The systems could provide 5-15 seconds of warning time to the region located 30-100km away epicenter for preventive action. In order to prevent false alarms, P-Waver implemented multiple sensors to capture the initial wave. The team estimates accuracy of the order of 98%. The system can predict p-waves in 1-3 seconds, with no false alarm through the multiple backup sensor set up and calibration.
P-Waver set their vision further for potential collaboration, one of the opportunities is to expand their innovative EEW technology to smart city development. As P-waver highlighted, “their ambition is to approach High Tech customers, smart city developers that not only in the United States, such as Intel, Micron, Tesla, google, Honeywell etc, but also partners in another region to work on smart city application business together.”
With more seismic data in databases and increasing computer computing power, seismologists are turning to big data and AI techniques to understand and improve complicated simulation models of seismic activity. Other researchers are using machine learning algorithms to sift through seismic data to better identify earthquake aftershocks and volcanic seismic activities, then monitor the tectonic tremor that marks the deformation at plate boundaries where earthquakes can occur.