Configuration Management and Performance Verification of Explosives-Detection Systems (1998)

The objective of explosives-detection equipment is to detect certain types and amounts of explosive material with a high detection rate, a low false-alarm rate, and a throughput rate that makes it practical to be used in commercial airports. An explosive is a chemical compound that reacts rapidly, generating substantial amounts of heat and pressure. Most chemical explosives are organic compounds or mixtures consisting principally of hydrogen, carbon, nitrogen, and oxygen. Plastic explosives are of particular concern due to physical properties that make them easy to conceal and difficult to detect.

Bulk explosives-detection techniques remotely sense some physical or chemical property of an object under investigation to determine if it is an explosive. Such techniques often exploit the high nitrogen and oxygen content found in explosives. The relative amounts of hydrogen, nitrogen, oxygen, and other elements (e.g., carbon) can be used to discriminate explosive from nonexplosive materials. However, for each density window in which there is a cluster of explosive materials, some nonexplosives, such as plastics, clothing, and narcotics, may also be included (see Table A-1 and Figure A-1 in Appendix A). It is therefore the comparison of a number of relative amounts, or windows in a multidimensional space, that will discriminate explosive materials more specifically.

In some bulk explosives-detection techniques, the measurement of material density is limited and is therefore not sufficient to distinguish explosives from nonexplosives. In these cases, the use of geometric information such as size, shape, and volume of material with a certain density are utilized to decrease the degree of ambiguity and increase the accuracy of the system. The ability to detect the physical appearance (pattern recognition) of certain materials or objects (e.g., wires and detonators) can also be used to overcome shortcomings in separating the chemical components of the substance in question from those of surrounding non-threat items.

Trace explosives-detection equipment is based on the physical transport of vapor or particulates of explosives from the object under investigation to a unit for direct chemical identification. In contrast to bulk explosives detection, which can identify both the type and the approximate amount of explosive present, trace explosives detection can only indicate the presence of explosive material and, in some cases, the type of explosive. The strength of the detection signal for trace detection equipment is not related to the quantity of explosives present.

Trace detection techniques are less likely than bulk detection techniques to misidentify common, nonthreat items as explosive materials, but they suffer from missed detections due to inadequate sample collection.

Descriptions of the broad categories of explosives-detection technologies are given in Appendix A, with more comprehensive discussion of the wide variety of explosives-detection technologies to be found in other references (NRC, 1993, 1996; OTA, 1991, 1992). In the following sections, bulk explosives-detection equipment is considered to be the archetype explosives-detection equipment. Where significant differences exist between the bulk approach and the trace approach, trace detection technologies are discussed separately.

Explosives-detection equipment can be delineated into operational subsystems (Figure 3-1), each of which performs a function crucial to overall system performance and, as such, is important for configuration management and performance verification. In addition, the explosives-detection equipment infrastructure consists of support utilities, power supplies, transport mechanisms, baggage-handling interfaces, communications network, and a physical structure to

Figure 3-1

Schematic block diagram of the operational subsystems  comprising an explosives-detection system.

house the equipment. There are likely a wide variety of approaches to selecting operational subsystems, from one manufacturer to another and, perhaps, between different models manufactured by one manufacturer. Therefore, the manufacturers of explosives-detection equipment are best suited to determine how the operational subsystems are identified. The following is one example of how the equipment could be broken down into four operational subsystems:

• Sampling—collecting information about the bag that is necessary to characterize it. A detector is used to obtain the responses from threat and nonthreat objects.
• Analyzing—converting the sampled information into an interpretable form, such as a visible image, a total mass spectrum, shape, density, effective atomic number or other output, as an input to the classifying operational subsystem.
• Classifying—using the properties of the object and its constituents, as determined by the analyzing operational subsystem to classify the sample as a threat or nonthreat object.
• Interfacing—displaying key detection information and system condition information to an operator, technician, or automated security system.

Detection is the process of combining all information from the operational subsystems and making a decision (by a computer) as to whether or not an alarm should be activated for subsequent resolution by other explosives-detection equipment or a human operator. Each of the operational subsystems relies on one or more software, firmware, or hardware components.

As discussed in later chapters, delineating a unit of explosives-detection equipment into operational subsystems, such as the four discussed above, with associated hardware, firmware, and software components enables effective performance verification and configuration management and may foster equipment development, maintenance, and upgrades. The details of which system components are assigned to which operational subsystems, however, should be left to the manufacturers' discretion. The operational subsystems, in some cases, differ between trace detection and bulk detection techniques, as discussed below.

The purpose of sampling is to obtain chemical or physical information from baggage. The critical components needed for sampling include a detector and sometimes an illuminator. For example, x-ray computed tomography (CT) relies on x-ray radiation to illuminate and elicit a response from explosive materials that differs from that obtained from nonthreat materials. In this case, the illuminator is an x-ray tube that deposits energy into the baggage under observation, and the detector is an x-ray detector that collects and analyzes x-rays after they have interacted with the baggage. The interactions of x-rays with the baggage are then used to measure the different physical parameters associated with individual objects within the baggage. Another technology for bulk explosives detection involves the use of nuclear particles (e.g., neutrons) to illuminate the sample.

Vapor and particle detection techniques use a collector to gather the vapor and particulates of the explosive from an object under observation. Sample collection is accomplished by using high-volume air flow to gather vapors or dislodge particles from surfaces or by wiping the surface.

During analysis, information that is acquired during sampling is related to a measurable parameter associated with the objects within the baggage. The analysis subsystem consists of the entire data-acquisition, processing, and analysis system to transform raw data to physical parameter(s). It includes data processing in preparation for signal detection and presentation to threat-decision functions and algorithms. The analysis subsystem relies on a substantial mix of hardware, software, and firmware components. For example, an x-ray CT EDS utilizes x-ray detectors, electronic preamplifiers and amplifiers, digital filtering and image reconstruction algorithms, and signal and image processing algorithms to analyze the contents of baggage. For these CT systems, the measured parameters include the x-ray attenuation coefficient averaged over the x-ray energy spectrum, which is a function of the physical density and average atomic number of the material, shape, and location of the objects within the baggage under observation.

Techniques for trace detection sample analysis employ a variety of chemical methods, including gas chromatography, chemical luminescence, and ion mobility spectroscopy. These methods can determine chemical properties such as molecular weight, absorptivity, retention time, fluorescent emission, and electron affinity of the vapor or particulate matter collected to distinguish the sample from nonexplosive materials (NRC, 1993, 1996).

The purpose of classification is to determine the existence of a threat through the use of data analysis and manipulation

algorithms applied during analysis. During classification, objects within the baggage are identified and classified, and their attributes are compared with those of threat objects. Although the classifier subsystem may include hardware, software, and firmware components, software components typically dominate this subsystem. Notification of the presence of a threat can be in the form of data input to the next stage of a multistage EDS or the presentation of information on the threat to the computer or human operator of an explosives-detection device.

For a trace detection system, only the existence of a chemical or set of chemicals can be discerned. No shape, volume, or weight information can be ascertained.

Even in the most automated explosives-detection equipment available at the time of this report (including the FAA-certified CTX-5000-SP EDS), human operators are necessary for detection and alarm resolution and therefore must interface with the EDS for proper operation. The user interface comprises those components of the equipment that provide audible and visible indication of the operational status of the equipment. These indicators include not only those associated with normal operation of the explosives-detection equipment but, more importantly, those involved with the presentation of information (typically in the form of visual images) concerning the possibility of a threat within the baggage under observation.

Although infrastructure is not defined in this report as an operational subsystem, it is critical in enabling the proper operation of explosives-detection equipment. Infrastructure includes the basic, underlying mechanical and electrical framework of the equipment. The infrastructure comprises those elements of the explosives-detection equipment that provide the mechanical structure, the interconnection of different electrical and mechanical components, the mechanical transport of the baggage under observation, and the overall control and coordination of the different equipment functions, the interface between different devices in a multiple-device system, and the airport or air carrier bag-gage-handling system.

To elucidate the architecture of explosives-detection equipment, two examples with a complete tabular description of each equipment module are provided in Appendix A. The first example represents a conceptual picture of an explosives-detection device whereas the second is based on currently used technologies.