Shortly after the Transportation Security Administration (TSA) took over responsibility for screening all passengers and their carry-on baggage at 450+ US commercial service airports, an aviation industry expert commented that the problem is not a security problem but an industrial engineering problem. That observation was true then and remains true today. The challenges of thoroughly screening the hundreds of thousands of US passengers and their carry-on baggage at a rate that facilitates commerce is immense, making process-thinking essential.
The rate of passenger screening at US airports this summer has received significant focus by the media as a result of the strongest growth in US air passenger traffic since 9/11 occurring at the same time as reduced TSA screener funding. As all IEs know: seconds and minutes count, especially when arrival rates exceed screening rates. This results in queues and wait times that grow exponentially. We have seen queuing theory validated by multiple stories of US travelers suffering 60 minute plus waits at screening checkpoint at US airports nationwide, leading to passengers missing flights. Beyond annoyance, the Airlines for America estimate that the passenger value of time is $39.74 per hour, making airport screening delays a worthy concern as we contemplate US productivity.
An airport passenger security screening checkpoint (SSCP) is a system of systems that are dependent, with high process and product variability, and demand that is transient and highly peaked. IE tools, especially discrete-event simulation (DES), are well suited to identify security process enhancements that will expedite security screening of passengers and their carry-on bags even within the closely regulated and controlled limits imposed by Congress acting through the TSA. The following sections discuss each of these challenges one-by-one.
In operational terms, the passenger enters the SSCP when they present their boarding pass for visual and electronic inspection by a Travel Document Checker (TDC). Subsequently, the passengers must remove all controlled items from their person and baggage to prepare for screening—called divestment. The TDC inspection rate must match the passenger divestment rate to prevent subsequent screening activities from being starved. Likewise, the passenger divest process must keep up with the x-ray inspection of the items they divest. The x-ray inspection of passenger items is accomplished by technology-assisted visual review by the Transportation Security Officer (TSO). If the x-ray image is complex, or if the TSO identifies images requiring additional screening, the serial process will be delayed, potentially halting the passenger divest process.
From a systems perspective, the system to screen passengers to ensure no contraband is concealed on their person is relatively simpler. A passenger enters the screening device—either a walk-through metal detector (WTMD) or a device that uses Advance Imaging Technology (AIT)—to detect contraband. In the case of a WTMD, passengers are permitted a second attempt at divesting and trip through the metal detector before being subject to a secondary screening protocol. When the AIT identifies elements requiring secondary inspection, the passenger is held in a controlled area at the AIT exit. In both cases, the primary inspection of subsequent passengers is delayed by secondary inspection of passengers. Since passengers must claim their belongings from the x-ray inspection system, if they are held up by the WTMD or AIT inspection process either because of excess demand or system delays caused by secondary inspection, the passenger inspection system may interfere with timely completion of x-ray inspection of passenger belongings. Conversely, if passengers are delayed in submitting their belongings for x-ray inspection, then at times the WTMD or AIT is starved, creating an overall reduction in throughput.
There are a number of strategies for keeping the throughput of these systems in balance. Assuming adequate capital, operations and maintenance funds, and space, this balance may be achieved by oversupply of inspection equipment and staffing. However, Lean principles tell us this is ultimately wasteful and not sustainable. IEs are pursuing other, more productive ways to reduce the dependency of the process elements and component systems.
Variability: The variability in large part comes from humans’ interaction with the system, both as subjects – or passengers—as well as components—or TSOs—of the screening system. The variability of a passenger’s speed of document presentation, divestment and re-vestment controls the speed of the screening process. Frequent travelers become trained in the best methods for accomplishing each of these processes; however, infrequent travelers who do not understand the system present challenges. Since formal training of all passengers is not realistic, and since there is natural variation in passengers’ speed of accomplishing all these activities, reducing variability by increasing the repeatability of the process is largely not an option.
Conversely, through thorough training and specification of inspection protocols, the TSO speed of performing inspection tasks is constantly being optimized. However, the magnitude of inspection tasks is immense and creates significant fatigue. Although TSA procedures are in place to manage this fatigue, over the time span of a shift, variability in process rate does occur.
As a side note, process thinkers prefer for the airport security screening process to be repeatable—at every airport at all times - and the lack of that repeatability is the subject of derision expressed in the media. For the highest levels of security, the detection of threat items by the screening technology must be, and indeed is, highly reliable and repeatable. However, intentionally injecting randomness in security screening protocols or practices is a proven effective strategy to dissuade potentially bad actors who observe the system to identify ways to defeat it.
Transient Highly Surged Demand: The US aviation scheduling network is predominantly dependent on the hub and spoke strategy. This scheduling approach calls for flights from spoke cities to arrive to a hub city so that passengers may be consolidated on larger planes that will transport them to the next segment of their journey. This approach conveys the most passengers when all of the flights from spoke cities arrive within in very narrow time windows. This scheduling approach is the most profitable for airlines and it offers passengers the most travel destinations at the lowest cost. However, it causes passenger demand to be very concentrated, leading to dramatic surges in screening system demand. Of course, highly peaked demand coupled with the highly variable nature of the screening process elements conspires to make the design and operation of screening systems a significant technical challenge. The complexity of this challenge is deceptive, since the SSCP process is not highly automated.
Although screening is technologically increasing, the screening technologies’ integration into the SSCP to date has been low. Artifacts of these low tech integration approaches include the serial nature of screening processes and the ultimate reliance on human review of screening images. However, due to the perfect storm created by significant passenger growth and reduction in TSO funding, the TSA has been encouraged by Congress to aggressively pursue innovations in both screening and product development. A recent example is the implementation of innovation lanes at Hartsfield-Jackson Atlanta International Airport in cooperation with Delta Air Lines. Technology advances include strategies to make passenger divestment and items requiring further inspection to be processed in parallel rather than serially, as well as collecting and returning the empty bins back to the divestment area automatically.
The complexities of the SSCP make DES an excellent tool for evaluating potential system enhancements. To accurately evaluate the impact of system enhancements, it is important to use a tool that can quickly be configured to reflect all of the interactions, process distributions and passenger behavior characteristics accurately and in great detail. Of course DES is perfect for the speedy and cost-effective evaluation of these system enhancements.
It is important to note that this discussion is primarily applicable to US airports. The US airport security is governed by Federal law, executed by the TSA. TSA security policy and protocols are applied to US airports that are owned and controlled by local governments, and operated in cooperation with various stakeholders. All airport security screening equipment is approved, based on stringent performance requirements, specified, and paid for by the TSA. Demonstration of new security technology and systems is closely managed through limited pilot programs. Consequently, advances in airport security must be accomplished within this highly-regulated operational environment that challenges the nimbleness of today’s rapid technology life cycle. Regardless, process thinking must prevail!