Optimal allocation of aviation security screening devices

Recent advances in aviation security screening technologies have presented a new challenge in optimally allocating these devices across a set of airports. Following check-in at the ticket counter, self-service kiosk, or airline website application, each passenger is assigned to a security screening class through an automated passenger prescreening system based on their measured perceived risk level. The class to which a passenger is assigned can be used to determine how this passenger's baggage will be screened by a set of security devices and procedures. In this paper, an explosive screening device allocation model is formulated as a nonlinear integer program to assign both the type of and number of devices to each class at each airport such that the total security is maximized, given a set of budget, resource, and passenger throughput constraints. A Dantzig-Wolfe decomposition approach is used to transform the nonlinear program into a binary integer program by redefining the binding constraint associated with the number of individual devices allocated across all airports. Computational results are provided for several randomly generated problems to demonstrate that the resulting binary integer program can be quickly solved to optimality.