Operations research models developed by a team at the Georgia Institute of Technology helped CARE International pick three locations worldwide to supply relief quickly to victims of earthquakes, floods, and other natural disasters, according to a paper in a journal of the Institute for Operations Research and the Management Sciences (INFORMS).
"Pre-Positioning of Emergency Items for CARE International" is by Serhan Duran, currently at the Middle East Technical University in Ankara Turkey, and by Marco A. Gutierrez and Pinar Keskinocak of the H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology. It appears in a special issue of the INFORMS journal Interfaces that is dedicated to the new, growing field of humanitarian logistics, which relies on industrial concepts like supply chain management to benefit the public sector.
The special issue is edited by Ozlem Ergun, Prof. Keskinocak, and Julie Swann, who are directors of the Georgia Tech Center for Health and Humanitarian Logistics.
"The team's work not only gave us excellent recommendations regarding locations, inventory levels, and an expansion strategy for the network, but is also serving as the basis for funding proposals for the network," writes Rigoberto Giron, Associate Vice President, Strategic Initiatives and Supply Chain Management, CARE, in a post-script to the study. "By pre-positioning we expect to reduce response time from weeks to 48-72 hours, reduce procurement costs by buying in larger quantities, reduce freight costs by using transportation resources more efficiently and improve coordination with other responding organizations."
The models created by the team helped CARE International review 12 suggested international locations for opening new CARE warehouses and finalize three, in Dubai, Panama, and Cambodia. They also helped CARE determine that, although more is better, the benefits of multiple relief-supply warehouses declines after the number reaches three to four, thus helping CARE make maximum use of its limited resources.
In their research, the authors considered two kinds of capacity constraints: the number of warehouses to open and the inventory amounts to keep throughout the pre-positioning network. They ran their model for the option of opening between one and nine warehouses and for three levels of inventory high, medium, and low.
With funding limited, they also helped make the critical decision which warehouse location would be most valuable to open first. They recommended that CARE open its first depot in the Middle East, expand to Central America, and then to Southeast Asia. Given a gradual roll-out plan, they were able to determine that once all three warehouses were operational, the supplies should be divided 35% in Dubai, 15% in Central America, and 50% in Cambodia.
If CARE obtains the resources to open a fourth warehouse in Africa and a fifth in Europe, the authors' sensitivity analysis shows that the relief organizational will be at its highest possible state of readiness to respond to unforeseen disasters anywhere in the world.
The authors were able to make recommendations by modeling the frequency, location, and magnitude of future demand based on historical data about earlier CARE relief operations.
One of the first applications of the research took place during the 2010 Haitian earthquake. With a million water purification kits and other supplies at depots, CARE was able to rapidly deliver water purification tablets to victims of the earthquake from its Panama warehouse.
The team's model is a mixed integer program that was run on a 4 x 900 MHz processor using ILOG OPL Studio with the CPLEX solver.
|Contact: Barry List|
Institute for Operations Research and the Management Sciences