Stochastic Dual Dynamic Programming Solution of a Short-Term Disaster Management Problem

Abstract

Every year, more than 500 disasters are estimated to strike our planet, causing the deaths of around 75,000 people, and impacting more than 200 million others [1]. To decrease the number of deaths and disaster-affected people, the design of an efficient and effective emergency logistics system is vital. The inefficiency in such a system in Haiti after the 2010 earthquake had the following results: Various on-field journalists [2,3] reported that relief efforts were stalling in the logistics network and that, therefore, much aid remained undelivered. This caused more than 3 million affected people, 200,000 deaths, and more than one million wounded.Disaster operations represent the set of all activities performed before, during, and after a disaster in order to diminish its impacts. Traditionally, the life cycle of disaster operations is divided into four stages, namely risk mitigation, preparedness, response, and recovery. The risk mitigation operations are performed before the disaster occurs, and they aim at reducing vulnerability to the disaster impacts such as injuries, and loss of life and property. The preparedness operations involve those activities performed before the disaster occurs in order to accomplish a more efficient response. The response operations are performed upon the occurrence of the disaster, and they address immediate threats presented by the disaster, including saving lives, meeting humanitarian needs (food, shelter, clothing, public health and safety), evacuating people from the disaster affected area, etc. Finally, recovery operations are performed after a disaster occurs, and they refer to short-term or