Big Data Application for Emergency Evacuation in Multi-Hazard Events

Faculty Mentor: Dr. Ying Huang

Since the 1190s, studies show that the annual damages caused by disasters are more than 100 billion USD, affecting more than 600 million people, and the upward trend in such losses is expected to continue (Laframboise and Loko 2012).  Well-planned evacuation operations and identification of appropriate rescue routes before and during a disaster play a significant role in saving lives and minimizing casualties.  Generally, transportation planning departments consider the peak traffic demands during normal workdays and on special occasions.  However, it is almost impossible to conceive transportation plans for emergency situations, in which large volumes of traffic from mass evacuations are likely to exceed the capacity of road networks.  The delay of evacuation induced by the ineffective plans may lead to loss of human lives and properties.  For example, due to the lack of proper evacuation plans, 25 people lost their lives in the first 30 minutes while attempting to flee Oakland Hills, CA during a 1991 wildfire.  To prevent future incidents, this project will use large-scale big data-based computation to develop real-time emergency evacuation plans during multi-hasard disaster events to ensure the availability of safe, efficient evacuation routes for residents (Khan et al. 2015; Khalid et al. 2016).  The proposed emergency evacuation plans will be based on cloud computations designed to: (a) act as a decision-making tool for transportation departments to evaluate and review their emergency evacuation plans, and (b) recommend preferred, efficient routes during a disaster using high-end sensors and intelligent transportation systems.

Teacher and/or Community College Faculty Component: The teachers will be trained to understand the basics, including: 1) understanding the needs of emergency evacuation in multi-hazard events; 2) reviewing state-of-art and state-of-practice evacuation plans in major US cities; 3) obtaining and analyzing the sensor data for evacuation applications; 4) modeling emergency traffic management planning; and 5) cloud computation fundamentals for this application.