612289 - A Dynamic-data-driven Decision-making Framework for Efficient Disaster Management Strategy

Monday, May 20, 2019

4:10 PM – 4:30 PM

Location: Wekiwa 6

Primary Contact(s)

Establishing an efficient disaster management strategy against severe natural disasters is essential to mitigate and relive its catastrophic consequences. Overall comprehension of human’s behavior and their decision-making processes, which is essential for the better situation awareness under such catastrophic events, require information from various sources such as survey data, historical information regarding location and intensity, government’s policies, and information about the supplies in the affected region. In this work, we propose a dynamic-data-driven model for people’s evacuation decision-making processes to develop an efficient disaster management strategy by evaluating the causal relationships among people’s decisions and the information provided by the government, news media, and community. For this study, we consider a situation in Florida during hurricane Irma in 2017 and develop an agent-based simulation by combining multiple data about the hurricane, traffic flow, government’s actions for the disaster, and people’s responses. The quantitative relationship among people’s decisions, their evacuation patterns, and the traffic flows over Florida will generate the better gas allocation strategy to gas stations along major highways. Furthermore, what-if analyses are conducted to find the best alternative action to the government agencies to minimize the effect of the hurricane, which will help the preparation for future disaster situations.