When disaster strikes, having relief supplies in the right place to be deployed swiftly is critical.
Humanitarian relief agencies often position such supplies in advance to help ensure ready availability but lack a good way to gauge the effectiveness of such preparations.
It’s difficult to know what quantities of supplies will be needed and where they should be placed to be most effective, particularly given the uncertainty about where and when a disaster may occur, said Chris Zobel, professor of business information technology in the Pamplin College of Business.
In a recent research project, Zobel used analytics to develop a new approach to help the Red Cross address the challenges of more effectively pre-positioning resources that are needed to help open emergency shelters.
Zobel and co-researcher Andy Arnette, who received a Ph.D. from Virginia Tech and now teaches at the University of Wyoming, worked with the Red Cross in Wyoming and Colorado to build a model for allocating assets to prepare for multiple possible disasters in a region.
Effectively managing supply chains in humanitarian aid and disaster relief can be extremely complex due to demand uncertainty, shifting needs and priorities, and nontraditional objectives, Zobel noted. “This complexity means there is both a significant need and an opportunity to develop new analytic approaches for improving humanitarian supply-chain operations.
“Our project with the Red Cross is based on the organization’s data and its actual operations, resulting in a model that has practical applications and real potential to assist disaster response efforts for other relief organizations also,” he said.
“Not only will our approach allow assets to be more accurately pre-positioned to reduce immediate suffering, it also will save time and resources that can then be put towards other types of disaster response and relief activities.”
Their model can also be adapted to a variety of problems that involve exposure of a vulnerable population to a hazard, the impacts of which can be diminished through appropriate and equitable allocation of different assets. These problems include flooding and wind damage resulting from hurricanes.
Citing some other examples, Zobel said the model could be useful to such agencies as the U.S. Forest Service and CAL FIRE, which frequently pre-position mobile cache vans to support operations bases for fighting forest fires in the Western U.S., or to organizations, like the New York State Office of Health Emergency Preparedness, that pre-position medical emergency caches to mitigate the spread of infectious diseases, such as anthrax.
Despite the growing frequency and severity of disasters, the Red Cross lacked a systematic, analytical approach for determining the most effective locations to place its caches and trailers — the two types of assets stocked with the supplies needed for opening emergency shelters, Zobel said.
The computer model Zobel and Arnette created incorporates a risk-based formula that uses the likelihood of different hazards and the exposure and vulnerability of the population to determine the extent to which a given resource allocation can reduce disaster risk.
“Our approach helps ensure that such allocations will be made equitably, based on addressing risk, not just demand,” Zobel said. “In particular, if two populations have the same hazard likelihood and level of exposure, the more vulnerable population will be assigned more resources to offset its increased risk.”
Their approach, he said, also allows agencies to quantify the value of mobile assets in offsetting risk in adjacent locations, giving them important flexibility: “Lower-risk areas can have their needs covered by adjacent assets and more direct relief can be provided to the more vulnerable and higher-risk locations.”
An article by Zobel and Arnette on their work, “A Risk-based Approach to Improving Disaster Relief Asset Pre-Positioning,” will be published in Production and Operations Management journal.