An Engineering Application of Earthquake Early Warning: ePAD-Based Decision Framework for Elevator Control.

In a medium-to-large earthquake, there are often reports of people being trapped or injured in elevators. This study investigates using an earthquake early warning (EEW) system, which provides seconds to tens of seconds warning before seismic waves arrive at a site, to help people escape from the elevators before a strong shaking arrives. However, such an application remains as a major engineering challenge due to the uncertainty of the EEW information and the short lead time available. A recent study presented an earthquake probability- based automated decision-making (ePAD) framework to address these issues. This paper focuses on studying the influence of two commonly ignored factors, uncertainty of warning and lead time, on the decision of stopping the elevators and opening the doors when an EEW message is received. Application of the ePAD framework requires using the performance-based earthquake engineering methodology for elevator damage prediction, making decision based on a cost-benefit model and reducing computational time with a surrogate model. The authors' results show that ePAD can provide rational decisions for elevator control based on EEW information under different amounts of lead time and uncertainty level of the warning. [ABSTRACT FROM AUTHOR]