Thesis Archive

A Multi-echelon disaster relief supply chain model considering stock-prepositioning and demand customization (2010)

Agonias, Kristianne J.
Marquez Karla Louie M.
Remigio, Kylie E

Abstract:

-This study investigates the modeling of a humanitarian supply chain specifically designed for disaster response. Various supply chain models as well as literature regarding humanitarian supply chains are reviewed. Among the many challenges that most humanitarian supply chains are currently facing, distribution processes, and satisfaction of end-receivers are the most noteworthy. Additionally, timeless on delivery of goods is crucial because lives are at stake. Also, the human flows aspect is important in humanitarian supply chains they are the ones responsible to mobilize supplies and distribute them to the end-services.

A disaster relief supply chain model that allocates multiple commodities in a multiple period, multi-echelon system such that the total coverage benefit and fairness satisfaction are maximized was formulated. At the same time, this model considers stock pre-positioning, demand customization and volunteer allocation since these are also needed to reduce the time taken to respond to the needs of disaster victims. The model dictates the warehouses and depots to be utilized, the quantity of relief packs to be delivered and the quantity of additional commodities to be distributed in the relief centers. The model was run using General Algebraic Mathematical Model System (GAMS). An optimal solution was found and the decision policy was used to determine the configuration to use. It was found out that demand for the relief centers were not completely satisfied but a considerable portion of the demand were satisfied. The warehouses and depots delivered standard relief packs and additional commodities to the relief centers. The model has addressed the problem of a humanitarian supply chain integrating decisions on a multi-period, multi-echelon environment that takes into consideration volunteer allocation.

For future studies, it is recommended to incorporate a post-disaster strategy to the model to make the service to disaster victims more thorough, use actual data instead of hypothetical values to generate more accurate results, and incorporate volunteer scheduling to be able to maximize their utilization and thus, have a positive impact in meeting the demand lead time.

Adviser:

Mr. Dennis Cruz