Decision Assessment Algorithms for Location and Capacity Optimization under Resource Shortages

Emergency response preparedness requires strategic and operational planning. As natural disasters, disease outbreaks, and bioterrorism continue to threaten public health and the global economic system, policy makers and businesses develop plans for improving their readiness. These emergencies present challenging problems that require evaluation of several trade-offs due to insufficient resources and capacity with limited response time pressure. Finding an optimal strategy to these problems can be computationally difficult, since they are formulated as large-scale service network design models with location-allocation decisions. The facility location problems in disaster management literature are typically known as NP-hard problems. Therefore, heuristic algorithms can help find practically feasible solutions, but decision makers may have to sacrifice one of their objectives while trying to satisfy multiple of them. This paper presents a flexible algorithm to evaluate trade-offs in resource scarce situations with a given response time. We formulate the location-allocation problem with capacity and time constraints, and the objective of minimizing the service time for individuals in an affected area. Due to the complexity to solve the problem for large-scale scenarios, the presented algorithm relaxes capacity and time constraints, simultaneously, and presents flexibility to assess trade-offs. A modified NSGA-II algorithm is used with a penalty function formulated to leverage resources. We analyze how values of the penalty parameters interact with a number of open sites in defining an efficient resource allocation strategy. Therefore, decision makers can expand their choices to design an emergency response network with their preferences taken into consideration.

Automated summary

Emergency response preparedness requires strategic and operational planning, involving trade-offs due to limited resources and time pressure. Facility location problems in disaster management are known as NP-hard problems, and heuristic algorithms can help find solutions in such scenarios. Decision makers may have to sacrifice certain objectives when trying to satisfy multiple goals in emergency response planning.