Skip to main content


Project Dates
05/01/2015 - 12/31/2016
Principal Investigators
Project Status

Vehicle routing problems (VRPs) have been studied extensively and, therefore, there are many algorithms and techniques already developed for the VRPs. In VRPs, the common objective is to minimize the distance travelled or time spent for the travel. That is, the objective is to achieve profitability and/or quality by minimizing the total travel time and/or distance. In disaster relief vehicle routing problems, the objective is quite different; loss of life and human suffering need to be minimized, which may weaken most algorithms and techniques developed for the VRPs. As an illustrative example, let us introduce the following problem that is depicted in Figure 1. When the disaster area is modeled as a transportation network, a node can be used to represent a geographic locus and links (also called arcs) can be used to represent roads connecting nodes (regions). The task is to deliver critical supplies from depot at node A to beneficiaries at nodes B, C, and D. The VRP solution takes the route of A→B→C→D→A (or an opposite direction), taking 20 hours to complete. If the vehicle departs at midnight, the arrival times are 2 a.m. (B), 10 a.m. (C), and 6 p.m. (D). If the arrival times are to be minimized, rather than the total time spent, then the optimal solution is the route of A (midnight) →B (2 a.m.) → D (6 a.m.) → C (2 p.m.) → A, taking 22 hours. The impact of such change in routing can be extremely significant when it comes to humanitarian relief problems as early delivery of critical supplies can reduce the impact of aftershocks substantially. Campbell et al. (2008) study the impact of different objectives in the VRPs in the context of relief efforts and propose solution approaches.