The Immediate Challenge

By Michael Patrick, Director of Architectural Design, Leo A Daly, Inc.

April 2002

Recommendations for eliminating bottlenecks at screening checkpoints
Following is a proposal by a consultant who, with others, has been involved in analyzing security screening checkpoints at U.S. airports. The intent is to offer practical guidance on what airports can do in the short term to alleviate choke points that have arisen as security has tightened. This is an abbreviated version of a presentation made at the recent Airport Planning, Design, & Construction Symposium, co-hosted by the American Association of Airport Executives and the Airport Consultants Council in Denver.

In 2000, the draft of the recently published FAA Recommended Secur-ity Guidelines for Airport Planning, Design and Construction was essentially complete. When the Architectural Design Subcommittee completed the section on Security Screening Check-points for that document, we concluded that common sense layout rearrangements could improve checkpoint performance as much as new equipment could. Aspects of the basic idea include:

o creating a well-defined area in front of checkpoints to prepare passengers to move through the process smoothly;
o tables or podiums to encourage divesting (taking metal objects out of pockets) in advance of the metal detectors to limit congestion;
o a visual barrier to prevent x-ray screeners from being distracted by approaching passengers to increase focus and effectiveness;
o a distinction in materials (floor, ceiling, wall), lighting, etc. to deter accidental bypass of the check point through the exit lane (often called a "breach");
o a variety of ideas for arrangement of the equipment within the checkpoint itself.

It became clear in working on the checkpoints section that good design - in this case, careful organization, layout, and design of the physical components and spaces of a checkpoint - was an important element of success. Just as any design is based on a program, checkpoint layout is, or should be, carefully built around the procedures that will take place.

Post-9/11 Change 'A'

After failing the first (primary row) metal detector, the passenger must go directly to hand-screening.

This post-9/11 change in procedure eliminates the recycling process, with sequential divesting (taking more and more out of a passenger's pockets each time). Thus, moving through the metal detector is much quicker. However, those who fail the metal detector are conducted to an area for far more extensive and time-consuming hand searching than before. Sometimes their bags are left on the end of the x-ray belt, jamming up the works as noted above, and causing the initial queue to grow as people wait for the belt to start to move again. (At the moment, the x-ray machines are even further burdened by the practice of examining selected individual's shoes.)

In general, then, the time for metal detection processing has decreased greatly, while the x-ray belt is more congested than before.

Meanwhile, the handoff of the person who has failed the metal detector from the first screener to the next screener responsible for hand searching continues, as it was pre-9/11, to be a weak link in the processing speed of the checkpoint as a whole. At present, the first screener must blockade the metal detector (stopping flow and making the queue grow) until the next responsible screener arrives to take over, which can be a significant time. A better procedure needs to be addressed for this weak link.

One possible procedure is to allow for a retention space, possibly a glass "corral" between two x-rays, as a waiting area for passengers who alarmed the metal detector and are to be hand-screened. This would free the primary screener to open the metal detector flow again. In one case an increase in flow of 40 percent was measured using this procedure. This and other options should be considered and tested.

A major effort for the country should be the design of a more consistent, well-considered step-by-step procedure for the checkpoint screening process more along the lines of an industrial assembly line, in which tasks and responses to alternative situations are well-known and well-rehearsed. We hope that this will be a priority of the new Transportation Security Administration.

PRe-9/11 challenges
Before 9/11, secondary row metal detectors were becoming common to eliminate "recycling" through the primary row, which was causing long lines and delays. In short, if a person did not divest enough and the metal detector alarmed, he or she was asked to go back out of the metal detector, remove more items, and try again. For each person this happened to, and each time he or she had to recycle, the line grew longer. The secondary row metal detector was somewhat successful in shifting that "recycling" procedure out of the flow of people who successfully "passed" the metal detector.

This arrangement also proved to create a new difficulty relative to the procedures in place.

With two metal detectors in the first row, the screener faced a dilemma. When a passenger alarmed at a primary row metal detector, the screener was responsible to guide the alarmed passenger into the tunnel or corral for the secondary row metal detector, until the next screener took over. While doing so, the primary row screener's back could be turned to the primary row metal detectors for which he was responsible. If one alarmed while he was turned around, there was often confusion about which one had alarmed, resulting in delays as the situation was sorted out.

The basic principal is that the layout combined with the required procedure put the screener in the sometimes impossible position of being simultaneously responsible for
controlling three points which could never be in his or her field of view at the same time. This is a failure of integration of procedure and design, and illustrates the general principle of this article: Efficiency results from the skillful integration of design and procedure.

At around this time, still a year or more before 9/11, our subcommittee considered the possible benefits of using a single metal detector in the primary row. Members of our committee had experience at Heathrow Airport with such an arrangement that appeared to be working. Besides simplifying the screener's task, this arrangement seemed to allow the overall screening to take less space in terms of width. Further, there was less equipment to buy and there were fewer machines to maintain. While noted in the Guidelines, the idea remained mostly dormant.

Addressing an Imbalance
The Architectural Design Sub-committee also analyzed data on the amount of time spent by passengers in each stage of the screening process. We noticed that there was an imbalance between time spent in the x-ray process versus time spent at the metal detectors. In spite of the delays at metal detectors due to passengers not fully divesting and being required to "recirculate," the governing or limiting factor was still the x-ray machine.

Post-9/11 Change 'B'

All people and objects go through either the metal detector or x-ray.

Change B means the general banishing of the little bins that before 9/11 were passed around the metal detector. While this was a goal of some R&D leaders in the FAA before 9/11 to provide better screening of those objects and close a loophole, it had the consequence of increasing the number of objects, often complex electronics, that must be removed and placed separately on the belts, including laptops. Thus, the number of objects to be reviewed has increased greatly. The natural effect of this has been to slow the actual x-ray screening process for each passenger.

The reasons for this are instructive and deal considerably with human behavior. The fastest that the x-ray process will ever go is the rated belt speed of the unit (an average of 40 ft./min. ) when it is running continuously. However, even before 9/11, when an x-ray screener could not resolve a bag satisfactorily, the belt was stopped until the bag could be cleared or selected for removal and further screening. This took an average of six seconds and often longer.

Not only does this mean that one person who has passed through the metal detector(s) now had to wait by the end of the x-ray belt, but as importantly, the next person in line at the queue, if they are like I am and like many others I've observed, will wait until the belt is moving again and his or her bag fully goes into the front end of the x-ray tunnel before proceeding through the metal detector. The x-ray and the metal detector work as a system, but in this case the x-ray is holding things up while the metal detector stands unused.

Even before 9/11, hand-wanding and trace detection detained many passengers after they completed the metal detection sequence. In fact, with the procedure of recycling through the secondary (or primary) metal detector and sequential divesting, passengers often arrived to pick up their bags from the end of the x-ray belt only after a significant delay.

Post-9/11 Change 'C'

Each bag must be stopped in the x-ray machine for review (no continuous flow).

Change C exacerbates the situation by multiplying this greater number of objects with the increase in time that is required to stop the belt to review each bag. The general picture here is of an x-ray process that is groaning under the burden of more extensive and longer screening requirements, while metal detectors have become almost simple walk-through devices. Even the divesting process, which is now directed fully toward placing objects on the x-ray belt (either in bins or in bags) and not toward activities around the metal detector, has little effect on the metal detection procedure. It is recommended that divesting occur on podiums or tables away from the x-ray belt and metal detection arch to further leave metal detection unencumbered by last-minute divesting.

Proposed Reconfiguration

Perhaps one way to find some space for these additional x-ray machines is to revisit the idea of replacing two primary row metal detectors with a single one. This can create a more efficient use of space and of dollars. It is likely that this will work better at some installations than others.

This late arrival has deleterious results. As a few bags stack up at the end of the x-ray belt waiting for their owners, enough can build up that the belt has to stop until they are cleared. Again, as noted, people waiting to put their bags onto the x-ray belt will not go through the metal detector until they see their bag enter the x-ray tunnel, and the system is effectively shut down as the queue grows.

The above illustrates the interconnectedness of all the screening procedures at a checkpoint, but is included here to show that, even pre-9/11, the metal detectors often sat idle while passengers divested metal objects, or waited for the x-ray belt to move while queues grew to unmanageable lengths.

Post-9/11 changes
Since 9/11, there have been several changes in procedures which have increased this relative imbalance between the x-ray processing time and that of the metal detector. (Three of these changes are discussed in more detail in the highlighted boxes associated with this article.)

The net result of these is to greatly increase the time spent in the x-ray machine process, and decrease that spent in the metal detector. Now the metal detectors spend even more time idle relative to x-ray than before, while the flow through the x-ray has decreased substantially.

The facts and figures are disheartening, as provided by TransSolutions, which based its data on observations and simulation models at four major hubs and two large spokes in the U.S. Instead of an average of 400 to 500 passengers per hour per lane in a checkpoint pre-9/11, we are now seeing rates from 140 to 275, with a nationwide average of 200 passengers per hour per lane, a much reduced rate. If procedures stay relatively the same as they have been after 9/11, the increased burden on x-ray processing will require many more lanes of x-ray machines and space to accommodate them.

About the Author
Michael Patrick, AIA, is an award-winning aviation design architect who currently serves as the Atlanta director of architectural design for Leo A Daly, based in Omaha, NE. He has been involved with FAA and the Airport Consultants Council as chair of the Architectural Design Subcommittee in updating guidelines for security screening checkpoints. He has worked with airports at Boston, Albany, JFK, Providence, among others. He can be reached at or (404) 874-8333.
Others contributing to this report include:
o Cenk Tunasar, Trans-Solutions, Inc.
o Martin Fertal, TRW Systems