Passenger Screening Technologies
Perceptions of increased threats from explosives and nonmetallic weapons have prompted the investigation of new passenger screening technologies, including chemical trace-detection techniques and imaging methods that can see through clothing. The development of these technologies has reached the stage at which operational implementation can be contemplated. However, possible negative public reaction toward many of these new detection technologies will have to be addressed before these systems can be used in airports. Demands for additional space, utilities, labor costs, and increased operator skills that these technologies could impose on air carriers and airports also will have to be considered. This chapter describes the technological aspects of the systems under consideration.
Screening procedures currently used in U.S. airports, at least during routine operations, involve metal-detection portals for screening passengers and x-ray imaging systems for screening hand-carried baggage. Metal-detection devices impose a time-varying magnetic field in the space within the portal that induces eddy currents in metallic or ferromagnetic objects passing through that space. Various methods are used to detect these eddy currents, and when they exceed a preset level, an operator intervenes to ascertain the presence or absence of a dangerous object or weapon. The effectiveness of this security screening system depends not only on the performance of the equipment, but also on the performance of the personnel operating the equipment and resolving the alarms.
The detection instrumentation can be adjusted to optimize its ability to detect specific metals or alloys. The FAA utilizes a number of specific weapons to test the proper operation and performance of these metal-detection units. Typically, instruments are tuned for the optimum detection of these FAA test objects, at the expense of reducing instrument sensitivity in detecting weapons or objects not contained in this set of specific test weapons.
The FAA is considering for future application several new technologies capable of detecting a wider range of threat objects. The following section describes an approach used by a foreign air carrier, followed by sections describing the new technologies being considered by the FAA.
Among the presentations to the panel was a particularly enlightening one by Dan Issacharoff, the former head of security for El Al Airlines of Israel. Because of its national affiliation, El Al has frequently been a terrorist target. In response, El Al has developed an extensive security program requiring inspection of all baggage and the face-to-face questioning of all passengers. Issacharoff said the El Al security system emphasizes the identification of people who could be a threat, rather than the detection of objects that could be used to hijack or destroy an airplane. The system used to identify people who could be a threat is illustrated in figure 3-1, which identifies five types of people who could pose a threat to an airplane. ranging from naive terrorists, passengers who are unaware that they are carrying dangerous objects, to suicide terrorists who intentionally carry dangerous objects to destroy the airplane and kill everyone on board, including themselves. El Al has also developed psychological profiles of these individuals and a passenger-interrogation technique designed to identify them during check-in and before boarding.
This system has worked well for El Al, but Issacharoff pointed out that El Al flies out of approximately 200 airports worldwide. In contrast, U.S. air carriers operate more flights out of more than 400 airports in the United States alone. In addition, a large number of El Al passengers travel on international flights and arrive at the airport early enough to allow time for security screening. The time- and personnel-intensive system used by Ei Al may be suitable for international flights, but it would be inappropriate for an operation the size
TABLE 3-1 Passenger Screening Technologies Based on Imaging
Requires more than a single view
Requires more than a single view
Source: Jankowski (1995a, b).
Several emerging technologies can detect metallic and nonmetallic weapons, explosives, and other contraband material concealed under multiple layers of clothing by creating images that can be examined to discern these materials. No physical contact is involved. These are already used for a wide variety of security applications, such as screening visitors to correctional facilities and exit-screening employees to deter theft. Table 3-1 outlines some of the ways imaging devices could be implemented in an airport environment.
Imaging technologies either scan subjects for natural radiation emitted by the human body (passive imaging) or expose subjects to a specific type of radiation reflected by the body (active imaging). In either case, materials such as metallic weapons or plastic explosives, which emit or reflect radiation differently from the human body, are distinguishable from the background image of the body. These screening systems generate television-like digital images that can be evaluated by image processing and analysis methods. Images are viewed by an operator trained to identify potential threat objects in these images, sometimes with the assistance of image enhancing software that highlights unusual features. Although these technologies cannot detect objects concealed inside the body or in skin flaps, they are being considered for airport passenger screening because they would enable air carriers to screen for a wider variety of materials than they can with present screening systems. Two technologies under consideration use either x-ray or millimeter (or submillimeter) wavelength electromagnetic radiation. Figure 3-2 shows the electromagnetic spectrum, indicating the wavelength of x-ray and millimeter wave radiation in relation to other common sources of electromagnetic radiation.1
Passive Millimeter-Wave Imaging
A passive technology under investigation by the FAA operates in the millimeter-wave range (near 100 gigahertz) of
1 Energy, frequency, and wavelength are fundamentally related. Energy is inversely proportional to wavelength (e.g., long wavelength radiation is low energy and low frequency; short wavelength radiation is high energy and high frequency).
the electromagnetic spectrum. Passive millimeter-wave imaging is based on the principle that any object not at absolute zero temperature emits electromagnetic energy at all wavelengths. This energy can be detected by an appropriate receiver and can be used to produce an image. An important feature of this technology is its ability to accomplish imaging by gathering the radiation emitted naturally from the human body without artificial radiation. Thus, no health risks are associated with this technology. (An in-depth discussion of health concerns is presented in chapter 6.) The display methods and privacy considerations of passive millimeter-wave imaging techniques are similar to those of the x-ray methods described later in this chapter.
Figure 3-3a shows the visual image of a person being scanned using the Contraband Detection System produced by Millitech. Figure 3-3b is the resulting screening image showing the outline of two guns hidden under the subject's sweater. Images of this type reveal items a passenger might normally carry, such as a wallet, keys, a pocket knife, and belt buckles. Effective image analysis and interpretation requires these and other common and nonthreatening items to be distinguishable from threat objects. Images of subjects carrying threat objects, such as the ones shown in this report, were taken to illustrate the capabilities of the technology. They are not representative of typical airline passengers carrying a variety of objects that clutter their screen images. Image-analysis software is being developed to facilitate interpretation, but current technology requires interpretation by human operators.
Active Millimeter-Wave Imaging
Active millimeter-wave imaging technologies operate as short-range radar systems that project a narrow beam of millimeter-wavelength energy against the target and detect the reflected rays. The beam is scanned from head-to-toe or
toe-to-head to produce an image of the subject. Battelle Pacific Northwest Laboratories is developing a system for screening people based on active millimeter-wave technology. This method involves illuminating the subject with millimeter-wave radiation, but at a level low enough to prevent adverse health effects. However, the popular perception of the dangers of microwave radiation may cause public concern over this imaging technique.
Active X-ray Imaging
Active x-ray imaging technology uses low-energy, low-intensity x-rays reflected from the subject to create an image. The images are interpreted to detect the presence of metallic and nonmetallic weapons and explosives concealed under multiple layers of clothing. Two companies, Nicolet Imaging Systems in California and American Science and Engineering, Inc. (AS&E), in Massachusetts, now produce systems used for personnel screening.
Figure 3-4 is a representative photograph of an x-ray screening system. Figures 3-5 and 3-6 are examples of images from active x-ray imaging systems.2The images are revealing, but, as mentioned earlier, the subject is probably not representative of passengers who usually carry more nonthreat items on their persons. As in passive imaging systems, the presence of these nonthreat items makes images produced by x-ray imaging systems difficult to analyze and interpret.
Assessment of Imaging Technologies
The effectiveness of these technologies depends on how distinct the threat objects can be made against the background and how much of the body can be screened. Although the technologies being considered cannot create a photographic quality image, they can produce images that trained operators can interpret. All current imaging technologies require operators to view the images because humans can interpret complex images and identify anomalous objects more efficiently than available software. The wide variety of human shapes and sizes that can be expected during everyday screening also complicates automated image interpretation because software cannot simply be taught to recognize and discount a standard human body.
Another factor affecting the effectiveness of imaging technologies is the amount of time required to obtain enough information to make a decision. In screening passengers, the
target processing time is six seconds, which is the approximate time required to examine a passenger's hand-carried baggage. Each image shown in figure 3-5 was produced in three seconds; this plus the time required for image interpretation determines total processing time. In current applications of imaging technologies, including systems at correctional facilities, images from at least two sidesfront and backare usually taken. Images may also be taken from both sides, often with the subject's arms raised, to provide a 360° view. It is likely that there is less time pressure when screening people entering a correctional facility than when screening people preparing to board an airplane.
Trace-detection technologies are based on the direct chemical identification of either particles of explosive material or vapor containing explosive material. Thus, the presence of a threat object or bomb is inferred from the presence
2 These images are created using x-rays reflected off the surface of the body X-ray images used by doctors for diagnoses are created using much higher energy x-rays transmitted through the body.
of particulate matter or vapor. The main difference between trace-detection and electromagnetic or imaging is that in trace detection, a sample of the explosive material must be transported to the instrument in concentrations that exceed the detection limit. Trace-detection technologies cannot be used to detect the presence of metallic weapons.
The two distinct steps in trace detection are sample collection and chemical identification. To identify the presence of explosives, both steps have to function at the same time. The sample-collection phase of the procedure is the main point of contact between the technology and the subjects being screened. Table 3-2 shows potential applications of trace-detection equipment in airports.
Explosive substances can be transported from the carrier to the detection instrument as vapor or as solid particles. Initial efforts in the development of trace-detection technology were focused on collecting vapor around the person or baggage. However, because many modern explosives do not readily give off vapor at room temperature, the focus has expanded to include detection of particulates of explosive materials on the skin and other surfaces.
If traces of explosive material are to be detected, they must be concentrated from an air sample (vapor technologies) or dislodged from a substrate (particulate technologies). In vapor detection, large amounts of air must be collected, from which small amounts of the substances of interest must be extracted. In particle detection, pieces of explosive material must be removed from the surface to which they are adhering. Both trace-detection approaches have strengths and weaknesses, depending on the type of explosive material being sought. Vapor technologies are more effective for detecting explosive materials with high vapor pressures, while particulate technologies are more appropriate for explosive materials with low vapor pressure, such as military plastic explosives.
Samples can be taken either by having the subject walk through a portal or by passing a hand-wand device over the subject. Either method may be implemented as a contact or noncontact technique. In contact portal sampling, passengers walk through a portal by pushing open a door or by rubbing against paddles or brushes. In a noncontact system, an air stream passes over the passengers as they walk through the portal. Hand-wand devices may be used to sample air around the person or to make physical contact. In general, contact methods focus on gathering particulates of explosive material from the hands or clothing of the subject. Noncontact methods may use the air stream to dislodge particles, or they may distill a sample of explosive vapor from the air stream.
Although using a hand-wand device is a potentially efficient sample-collection technique, it is more labor intensive and more time consuming than collecting samples using an automated portal. The optimum solution may be to attach a hand-wand device to a portal-based trace-detection system as a higher-level surveillance accessory. This is a common technique used with metal-detection portals. In trace detection, a
TABLE 3-2 Passenger Screening Technologies Based on Trace Detection
Involves high-volume airflow to gather vapors or to dislodge particles from surfaces
Passenger opens saloon doors with hands
Passenger passes through a portal lined with brushes or fronds and brushes against them
Involves high-volume airflow to gather vapors or to dislodge particles adhering to surfaces
Technology currently in use
Boarding card is scanned after handling by passenger for particles of explosive materials
Source: Jankowski (1995a, b).
single chemical-identification instrument could be served by both the portal and the hand-wand device sample-collection mechanisms.
Because it is difficult to extract explosive vapors from large volumes of air or to gather particulates of explosive materials from the great variety of materials on which particles of explosive materials might be found, it is not surprising that no sampling technique that is universally adoptable has been identified. Developers have tried dislodging material mechanically with air-brush and air-vacuum devices. Although every approach is effective under specific sampling conditions, none of the techniques has shown itself universally effective. The lack of a specific sampling approach makes it difficult to discuss a generic trace method. It also creates problems in designing a standard testing and certification protocol for comparing the effectiveness of various technological approaches.
A problem in all trace-detection approaches is clearing vapors or particles of explosive materials from the sample-collection mechanism so that subsequent readings are not influenced by previous traces of explosive materials. The baseline readings must be monitored to alert operators to elevated levels of contamination before the contamination results in the shut-down of the equipment.
Identifying Explosive Materials
After a sample is collected, a variety of commercially available chemical-identification technologies may be used to determine if the sample contains any explosive materials. The detection limit of most technologies under consideration by the FAA is generally sufficient to identify explosive materials in a sample. Even average-performance mass spectrometers are capable of measuring and identifying ultra-trace quantities of relevant chemicals, but more highly trained operators may be required to maintain a high level of detection capability. Much of the work sponsored by the FAA in trace-detection technologies concentrates on integrating a particular chemical-identification technique with a sampling technique.
The chemical-identification part of the trace-detection instrument is likely to be smaller than the portal or sample-collection portion of the system. Therefore, for portal-type systems, airport accommodation is generally not dependent on the size of the chemical-identification component of the system but will more likely be affected by how the sample collection is implemented.
Some chemical-identification technologies may be small enough to be incorporated into hand-held instruments and thus have potential use in hand-wand devices. Technologies under consideration include pyroluminescence (for detecting solid particles of explosives, e.g., gun powder), chemical sensors, and ion-drift spectroscopy. As discussed above, these technologies must be combined with a sample-collection technique. The need to move large amounts of air in collecting a sample most likely means that available sample-collection techniques may limit the application of these technologies to continuous surveillance methods relying exclusively on the use of hand-wand devices.
NONIMAGING ELECTROMAGNETIC TECHNOLOGIES
Nonimaging electromagnetic screening technologies are used in places as diverse as libraries, court houses, schools, sports stadiums, and clothing stores. These mature technologies function as metal detectors to deter theft and to ensure safety. For airport use, a potential improvement would be to make these technologies specifically sensitive to weapons. As most travelers know, current metal detectors can be triggered by common, nonthreat objects, such as belt buckles or shoes
with metal shanks. Resolving these false alarms consumes time and resources and fosters an air of complacency regarding the detection of real threat objects. Other improvements include making the metal detectors (1) more versatile in detecting a broad spectrum of alloys, (2) more specific in locating suspected threat items, and, (3) more tolerant of electrical noise from nearby sources, such as video display terminals and fluorescent lights. These improvements would probably go unnoticed by passengers and cause little concern to airport management, from the standpoint of increased requirements for utilities and space.
The FAA is also considering a nonimaging dielectric portal designed by Spatial Dynamics Applications, Inc. Discovery systems based on this technique have been used for many years by the U.S. Customs Service to search for contraband in cargo being brought into the United States. This technique uses microwave irradiation and a transmitter/ receiver pair to determine the complex dielectric constant of the object being screened. The dielectric constant measured is compared to known responses for humans and threat objects to determine the presence of dangerous items. While the system being developed for consideration by the FAA is based on a simple signal comparison, the same technology could be used to produce an image in a manner similar to the technologies discussed above. For passenger screening, the person being screened steps into a portal and is scanned from head to toe to reveal the presence of both metallic and nonmetallic objects. The levels of microwave power used for weapons discovery are less than 0.1 percent of the levels established by the U.S. Food and Drug Administration (FDA) for microwave energy safety (Burnett et al., 1992). A single 360° scan is completed in approximately four seconds.
Hand-wand electromagnetic screening devices are used for locating specific items that set off alarms in portals and for screening persons who, for one reason or another, cannot or will not pass through portals. These devices, which can be manipulated with one hand, are slower than walk-through portals in screening passengers. The health and safety concerns associated with these devices are minimal, and their widespread use in airports apparently elicits little negative reaction from passengers.
Nonimaging electromagnetic screening technologies are unable to detect nonmetallic objects or materials. Technologies based on microwave irradiation are capable of detecting both metallic and nonmetallic threat objects.
CLEARING AN ALARM
The panel limited the definition of clearing the alarm to the equipment itself, that is, returning the instrument to its pre-examination state. The steps taken to determine whether the alarm was set off because dangerous materials or objects are indeed present are procedural. Therefore, they lie beyond the scope of this report.
Only trace-detection technology is expected to be affected by possible problems related to clearing the alarm. Neither imaging technologies nor nonimaging electromagnetic technologies have a memory effect. Therefore, settings usually return to the baseline state after the subject exits the inspection area. For these technologies, alarms are cleared simply by resetting the equipment to its pre-examination state.
In contrast to imaging technologies and nonimaging electromagnetic technologies, trace-detection equipment interacts with the vapor or particulate form of the materials of interest and signals the presence of these substances when their concentrations are above a threshold level. Ideally, the instrument signal should return automatically to its baseline value once detection has been completed. However, if the area around the sampling inlet is contaminated or if the trace compound lingers in the intake part of the instrument, the equipment will be unable to immediately return its settings to their baseline levels. Lingering contamination could result in persistent elevated signals, possibly above the alarm threshold value, and could cause the equipment to continue to indicate the presence of dangerous materials long after the original alarm-triggering event has passed.
Contamination of the intake part of the system may occur (1) if a vapor has high affinity for the material(s) in the sample collection part of the instrument and is present in high concentrations during the subject-screening stage, or (2) if the particulate trace is dislodged from the subject by mechanical means (e.g., an air brush), has high affinity for the sample collection materials, or lodges in the sample collection mechanism and has significant vapor pressure.
For the equipment to be contaminated by the material being detected, all conditions in either scenario I or 2 above would have to be satisfied simultaneously. Although these scenarios are unlikely, the probability may be minimized by the judicious choice of the materials used in manufacturing the sample-collection component of the system. The materials should have a low free energy of adsorption for the molecules of the explosive materials the instrument is designed to detect to prevent the molecules from adhering to the inlet walls. As an additional feature, the instrument should be equipped with a mechanism for bypassing the sample collection inlet. This mechanism could be used to provide uncontaminated ambient air to verify the proper operation of the instrument after the identification of explosive materials. These are issues that manufacturers of trace-detection equipment should address.
Aside from dealing with problems related to lingering contamination, manufacturers also have to address the tendency of trace-detection systems to react to the presence of materials, particularly certain medications, that are chemically similar to explosive materials. This tendency leads to false positives, which are likely to be more common than the
detection of true threat materials, and the need for operators to clear the equipment before being able to proceed with passenger screening.
The trade-off between technology performance and acceptance by air carriers or the public is a feature common to imaging, nonimaging, and trace-detection technologies. For example, a sharper (i.e., less ambiguous) image could be obtained by increasing observation time, but doing so could add to passenger delays. The intensity of the incident beam could also be increased, but this could lead to increased passenger concerns about exposure to radiation. Similarly, a more aggressive sampling for trace detection could lead to a higher rate of positive identification of explosive materials, at the expense of making the sample-collection phase more personally invasive to the passenger. Ultimately, the performance capability and quality of a passenger screening technology is unlikely to be the limiting factor in its implementation or application. Limitations on the technology will instead be imposed as a result of passenger intolerance for invasion of privacy, delays, or discomfort.