Citation

Ozguven EE, Ozbay K. Transp. Res. C Emerg. Technol. 2013; 29: 171-196.

Copyright

(Copyright © 2013, Elsevier Publishing)

DOI

10.1016/j.trc.2011.08.012

PMID

unavailable

Abstract

An efficient humanitarian inventory control model and emergency logistics system plays a crucial role in maintaining reliable flow of vital supplies to the victims located in the shelters and minimizing the impacts of the unforeseen disruptions that can occur. This system should not only allow the efficient usage and distribution of emergency supplies but should also offer the ability to be integrated with emerging ITS technologies such as Radio Frequency Identification Devices (RFIDs) for commodity tracking and logistics. This paper proposes a comprehensive methodology for the development of a humanitarian emergency management framework based on the real-time tracking of emergency supplies and demands through the use of RFID technology integrated with a multi-commodity stochastic humanitarian inventory management model (MC-SHIC). First, logistics and management aspects of RFID technologies in the context of the emergency disaster relief framework are discussed. Then, MC-SHIC model proposed to determine the optimal emergency inventory levels to prevent possible disruptions at the minimal cost is presented. The solution of the model with several sensitivity analyses obtained using the pLEPs algorithm is presented and discussed. Realizing that actual emergency inventory levels can deviate from optimal values during the actual disaster relief period due to the possible stochastic disruptions such as fluctuating demand for vital supplies in the shelters, a comprehensive on-line inventory control framework is proposed to minimize impacts of these unforeseen disruptions, or at least to address the problem at hand as fast as possible. Within this methodology, we obtain an approximation of the MC-SHIC model using a simultaneous perturbation stochastic approximation (SPSA) based functional approximator, and compare the performance of these algorithms for solving the new unconstrained optimization problem. Finally, proposed model-free on-line control methodology is discussed using examples to understand the efficiency and practicality of both algorithms in terms of computational times and accuracy of results.