The impact of transportation capacity in pre-positioning humanitarian supplies

Abstract

Humanitarian action saves lives by delivering supplies following a disaster. To effectively prepare, large humanitarian organizations solve optimization models to allocate inventory minimizing expected time-to-respond. However, these organizations also rely on transportation carriers to deliver this inventory, a feature often omitted from such models. Part of the reason is that capacity data on third-party carriers are scant. Nevertheless, there is value in humanitarian organizations asking themselves the following: how should such data be incorporated into a model; how important is it to incorporate transportation capacities; where is it worth increasing transportation capacity? Building on previous inventory optimization models, we partner with the United States Federal Emergency Management Agency (FEMA) to answer these questions by incorporating transportation market capacity. Using public and proprietary data sets, we first approximate the trucking capacity near FEMA's warehouses. We develop intuitive metrics that guide practitioners to make decisions, understand trade-offs in complex networks, and negotiate contracts. We show that when optimizing inventory, ignoring trucking capacity can lead to up to 18% higher response times. In discussions with FEMA, our research has guided strategic inventory deployment and has been a catalyst in a new initiative for transportation contracting.