Have you ever wished you could tell the TSA what to do with itself? Now, you have that opportunity — at least when it comes to how they organize and manage their queues. To make things even better, they might actually pay you! The Transportation Security Administration has posted a challenge asking for people to develop a simulation model to tackle the capacity management issues of getting people through airport security. If you are interested in the challenge, the official call is here. Here are some of the specific considerations that need to be tackled:
TSA is looking for the Next Generation Checkpoint Queue Design Model to apply a scientific and simulation modeling approach to meet the dynamic security screening environment. The new queue design should include, but not limited to the following queue lanes:
- TSA Pre✓™
- Premier Passengers (1st class, business class, frequent fliers, etc.)
- Employee and Flight Crews
- PWD (wheelchair access)
The Challenge is to provide a simulation modeling concept that can form the basis to plan, develop requirements, and design a queue appropriately. The concept will be used to develop a model to be applied in decision analysis and to take in considerations of site specific requirements, peak and non-peak hours, flight schedules and TSA staffing schedules. Solvers are expected to provide the concept and provide evidence that it works as described in the requirements.
According to Nextgov.com, there are specific performance targets for different classes of customers (Attention, Passengers: $15,000 Prize for Whoever Can Speed TSA Screening, Jul 18)
The line, in this scenario, extends from the point where a passenger joins the end of the queue to the metal detector or body scan machine.
The rules for the challenge state wait times cannot be more than 5 minutes for PreCheck and 10 minutes for standard lines.
So I should acknowledge upfront, that this is a challenging problem. They are dealing with lumpy capacity — there are a limited number of scanning machines and reallocating one machine from one set of customers to another can make a material difference for both types of customers. Further, they have limited control over how they can set things up across different airports. In Denver, for example, essentially all passenger for all airlines are routed to one location for clearing security. At O’Hare in contrast, there are multiple places to clear security just in Terminal 3 and there is no easy way to send customers in Terminal 1 over to open lanes in Terminal 3. Denver should be a simpler problem to solve than O’Hare and there is really nothing the TSA can do about change the physical layout of O’Hare.
It seems that the TSA is interested first and foremost in a simulation tool that they can use across different airports. However, there are some guiding principles that should be relevant for any proposal. For example, they should avoid having too many classes of service — by which I mean that they should not be serving premier passengers differently from anyone else. I devoutly concentrate all my flying one airline in order to score the meager benefits of being at the lowest premier level of that firm’s frequent flyer program. Among those is higher priority in clearing TSA. But that means that the airline scores a financial benefit from controlling access to a government service. Further, each flyer pays the same security fee so it is not clear why TSA should be supporting airlines’ efforts to discriminate between customers.
A second consideration is to exploit scale as much as possible. Queuing systems exhibit economies of scale — which is a fancy way of saying that as arrival rates go up, the capacity needed to provide a given level of service goes up more slowly. This is why the design of the Denver airport is so smart. It maximizes the flow of passengers through one location so they can be served efficiently. The arrangement at O’Hare will need far more agents and machines to provide the same level of service. Still there are things that can be done to make O’Hare run better. For example, if passengers in wheelchairs need special processing, then there should be a single location in each terminal that handles them. A similar case can be made for employees and flight crews. To the extent that Pre✓™ passengers and standard passengers require similar processing, the TSA should maintain flexibility to move machines between the two. My assumption here is that the mix of Pre✓™ and standard passengers will vary over time (e.g., early Monday should see a lot Pre✓™ flyers while and the spring break should see a spike of standard flyers). Hence being able to move resources back and forth will be more efficient that staffing up to meet peak traffic for each separately.