Finding the needle in a haystack

   Senior Editor

Since 9/11, security has risen to an entirely new level of importance. One security goal is keeping weapons, biologic and chemical agents, and radioactive material out of the U.S. and off public-transportation systems. But consider these facts:

• Last year an estimated 130 million cars, 12 million shipping containers, 11 millions trucks, and 2 million railroad cars crossed U.S. borders, not to mention nearly 500 million people.
• 700 million people and nearly one billion pieces of checked luggage travel through U.S. airports each year.
• Authorities seized more than 3 million items from U.S. airline passengers in a seven-month period last year. And while that total includes nail clippers, sewing kits, and other relatively innocuous devices, it also includes 800 guns, 80,000 knives, and 31,000 box cutters.

How can the U.S. and other countries possibly cope with a problem of this magnitude? They not only have to thoroughly screen passengers, luggage, and cargo, they have to do it without adding onerous costs or delays. One answer is technology.

X-rays

X-rays have long been used to search bags. High-energy beams sent through an object and on to a detector array create an image of an object's electron density, which correlates well with its material density. X-rays let security guards see into carry-on and checked luggage for metal guns, knives, and suspicious looking objects.

Engineers at L-3 Communications, Woburn, Mass. (www.dsxray.com), have combined X-ray and computed tomography (CT) technology, and taken them both a bit further. Their half-serious motto is "In God we trust, everyone else we X-ray."

The 3DX 6000 eXaminer, one of their top-of-the-line machines, uses X-ray-image density and variations in material density, along with a proprietary algorithm, to detect explosives. Federal regulations prevent them from revealing exactly how this is done or even what the variables in the algorithm are, but they do say their machines meet Federal certification criteria. One criterion is that "detection must not be dependent on the shape, position, orientation, or configuration of the explosive materials."

The 3DX 6000 can handle up to 600 bags/hr. Key to the high throughput is its helical-cone X-ray beam. Most CT systems use a fan beam to illuminate a single plane of the target (suitcase). The beam and detectors rotate around the target to create an image of one entire plane, collecting about one to two images/sec. If more planes are needed, the target is moved, stopped, and the process repeats.

With a helical-cone beam,

X-rays illuminate portions of several planes at once, and detectors simultaneously acquire data from several planes as well. The target moves at a constant speed past the rotating beam and detectors, generating data for all planes of the target and producing 40 slices/sec. L-3 also has a patented method of reconstructing CT images using a limited number of images.

When a bag sets off alarms, operators on older, single-slice machines must stop the line and take additional slices of the suspected bag. The 3DX, on the other hand, takes all the info it needs at once, so when a bag alarms, the on-scene operator or a remote operator linked to the networked 3DX can double-check images of the suspected bag while others continue through the detection process. This means that while the 3DX can operate at full speed, the single-slice machine's throughput is determined by its false-alarm rate. And remote screeners will not have all the images they need until the machine takes them.

Time is an important issue. Screening luggage and passengers shouldn't impact airline schedules and passenger satisfaction, and there is not that much time to spend training operators on complex detection systems. Therefore L-3 studied the screening process to see which functions on the 3DX machine were used most often in determining whether alarmed bags needed a manual search. "We assigned these functions hard keys and placed them near the pointing-device controls where operators routinely rest their hands," says an L-3 spokesperson. "Less frequently used functions are further out, and final-decision buttons are larger and placed on the outside edge of the panel." Display panels have also been programmed with operators in mind. They provide multiple views of the bag being screened, so operators intuitively know what they are looking at.

Another system, the PX-M baggage screener, uses dual-energy detectors, with one optimized for low-energy X-rays and the other for high-energy beams. It takes image data and calculates the atomic numbers for objects in the bag. The atomic number is the same as that on the periodic table of elements if the object is a pure element. If not, the number is an average, the effective atomic number, which helps classify objects as threats or contraband. PX-M also lets operators assign colors to specific classes of objects such as organic, inorganic, and metallic for quick identification

PX-M operators can adjust display characteristics such as contrast, edge enhancement, and level of penetration, until they get a clear image of what's in a piece of luggage. And authorized security staff can customize threat alert criteria. But operators also have to contend with Threat Imaging Protection (TIP). TIP randomly inserts threat images into displays, creating fictitious alerts. It keeps operators on their toes and lets supervisors evaluate their performance.

L-3 also builds mobile and fixed-site systems to screen trucks and cargo carriers for contraband. Their mobile, truck-mounted systems, for example, move the X-ray beam and detectors around stationary trucks. These systems use stronger, pulsed X-rays capable of penetrating denser cargo than carry-on luggage. Baggage systems operate at about 160 keV, which is considered low energy, and will penetrate about 1 in. of steel. Medium-energy systems work at about 450 keV, which penetrates 4 in. of steel. And high-energy devices put out 3 to 6 MeV, enough to see through 8 in. of steel, but they are usually powered by linear accelerator sources.

The boom-mounted detectors are optimized to receive pulsed signals, and data is digitized to expand the system's dynamic range to accommodate the wider variety of densities in cargo. And since they work outside, they have heating and cooling subsystems. The trucks are also slightly larger than they have to be. "This gives room for access," says L-3.

The trucks are self-contained, so they also carry generators. The boom carrying the detectors is electric and controlled by a simple joystick. "We avoided hydraulics and stabilizer jacks because both add complexity and can be problematic," says L-3.

Mobile X-ray screeners are more flexible and can go where the cargo is. But fixed sites are better suited to higher energy X-rays and let operators control potential radiation hazards more tightly. Fixed sites can also use dual-view systems that give operators a look from two perspectives, usually a side view and one from underneath (or above). This lets them quickly determine whether objects are contraband or not.

Although these systems are expensive, costing well above $1 million, they obviously pay for themselves if they prevent loss of life. And L-3 points out that in some instances, ROI can be relatively easy to calculate. "One government customer generated $10 million in one year from a high-energy system located between the mainland and a major city port. The system detects tariff evasion schemes, forcing smugglers to pay up. They used the money to build similar sites at all their ports."

Mass spectrometry and ionization

Mass spectrometry (MS) identifies compounds based on the mass of the molecules in them and can do it with samples measured in nanograms. The engineering trick is to take this complex process and package it so it can be used to quickly and reliably screen for chemical weapons, drugs, and other contraband. And that's what designers are doing at Syagen Technology Inc., Tustin, Calif. (www.syagen.com).

They use photoionization, shooting high-energy ultraviolet photons (about 10 eV) at sample molecules to create ions, trap them in a quadrapole ion trap (QIT), then send them through a time-of-flight (TOF) mass spectrometer. The spectrometer calculates the ions' masses and thereby identifies them.

In the first step, photoionization, the photons' energies must exceed the molecule's ionization potential (IP) to knock loose an electron and create an ion. (One rule of thumb is that the larger the molecule, the lower the IP.) "Almost all molecules of interest are large with ionization potentials below 10 eV, so they are all efficiently ionized," says Jack Syage, president of the company. "But explosives contain molecules with electronegative nitro groups, which raises the IP out of the 10-eV range. However, they have a very high affinity for electrons. So when screening for explosives, we use a patented discharge ionization process that creates negative ions of explosives and we then analyze them just like other ions."

Fortunately, photoionization doesn't effect small molecules with IPs above 10 eV, such as the main constituents of air and most solvents, so they don't swamp the analysis. Syagen uses high-output discharge lamps as a source of ionizing photons. They are less costly and maintenance intensive than lasers, and are much "cleaner," creating fewer fragments. Fragments are generated when ionizing photons have enough energy to break molecular bonds. "Consider the nerve agent VX, which has a molecular weight of 267," says Syage. "Ideally, you want it to show up in analysis as having a mass/charge ratio of 267. But if it breaks up into smaller, ionized particles, it creates two possible problems. The signal intensity of the mass/charge ratio at 267 diminishes, or, if all of the molecules in the sample break up, some might wind up with a 267 ratio and give erroneous readings. Photoionization doesn't have the energy to break molecular bonds and create fragments."

On the analysis end, Syagen uses quadrapole ion traps (QIT) to hold up to 10⁶ of the ions generated by photoionization and selectively eject them. Ions are held in a 3D electrode driven by RF fields. The fields determine which ions are held and which are ejected based on ion mass. Manipulating the RF waveform lets Syagen eject any combination of ions. They routinely pulse all ion masses out of the QIT and into a TOF mass spectrometer in less than 5 msec, getting a snapshot of the sample's constituents in 40 sec.

Syage and his team are still working at bringing QIT and TOF spectrometry to the market. They've used it to build an explosive detector for screening people, a chemical-weapon detector for monitoring water, and they'll soon have a combined chemical/biologic weapon detector.

Airtight security might seem a hopeless task. But with improved technology and wider dispersal of current technology, U.S. borders and travel will remain safe.