This project will develop better plans for an effective deployment of medical supplies in response to a large-scale infectious disease outbreak. The multiple decisions involved in an efficient logistics response compounded with the uncertainty present leads to large-scale optimization/decision problems that are intractable using current methods. The proposed research will create models that provide robust solutions to the uncertainty present and develop adaptable decomposition and branching algorithms for large-scale mixed integer nonlinear programs. These algorithms will use new results on sensitivity measures for problems under uncertainty, and warm start strategies for interior point methods.
Recent events, such as the 2004 Indian Ocean Tsunami and the 2005 Hurricane Katrina, have highlighted the massive impact that large-scale emergencies can inflict on society. Careful planning can optimize the logistics response to such emergencies reducing their impact on the population. This project will develop new modeling methods and algorithms that will enable the development of better plans for the disbursement of medical supplies in response to an infectious disease outbreak. Ultimately, improving preparedness can help save lives in emergencies. The methods developed here can be applied to improving preparedness in other problems and, in addition, the optimization techniques developed can be applied to solve other convex mixed integer programs. This project includes an outreach component to local, state and federal stakeholders through the Governmental Advisory Committee of the CREATE Research Center, a Department of Homeland Security funded research center at USC. This project will lead to curricular developments in logistics and optimization, involve a Ph.D. student, and involve minority undergraduate students in summer research projects through USC's McNair Scholar's Program.
