A medium-sized airport has a limited number of international f ights that arrive and require immigration and customs. The airport would like to examine the customs staff ng and establish a policy on the number of passengers who should have bags searched and the staff ng of the customs facility. The number of passengers who arrive on these large planes is uniformly distributed between 240 and 350, and the simulation is to run until all passengers on the plane have been completely processed (that is, the number of passengers on the plane determines the replication length). Arriving passengers must f rst pass through immigration (immigration is outside the boundaries of this model). They then claim their bags and proceed to customs. The interarrival times to customs are distributed as EXPO(0.2), with the f rst passenger arriving at (or just barely after) time zero; all times are in minutes. The current plan is to have two customs agents dedicated to passengers who will not have their bags searched, with service times distributed as EXPO(0.56). A new airport systems analyst has developed a probabilistic method to decide which customers will have their bags searched. The decision is made when the passengers are about to enter the normal customs queue. The decision process is as follows: a number is f rst generated from a Poisson distribution with a mean of 7.0. This number is increased by one, to avoid getting a zero, and a count is started. When the count reaches the generated number, that unlucky passenger is sent to a second line to have his or her bags searched. A new search number is generated and the process starts over. A single agent is dedicated to these passengers, with service times distributed as EXPO(3). Develop a simulation of the proposed system and make 20 replications (that is, process 20 planes, since a replication ends when the last customer from a plane clears customs), observing statistics on the system time by passenger type (searched vs. not searched), the number of passengers, and agent utilizations. kel01315_ch05_207-278.indd 274 05/12/13 3:28 PM Modeling Detailed Operations 275