Security Screening

Security screening is a very large potential market for millimetre wave technology. This is because over the last two decades the demand for individuals to be screened at entrances for illegal or forbidden items has risen dramatically, and shows no sign of abating.

The capability for screening individuals for items concealed under clothing is enabled by the almost complete transparency of clothing to millimetre waves. It will be the security screening technology of the future, being unobtrusive and fully automated using AI, to detect and identify metal and non-metallic threats. The technology offers a touchless screening capability, deployable in any scenario where there are people. 

Systems research over the last decade indicates deployment scenarios for this technology fall naturally into two categories, full-body scanning and stand-off screening:

1) Full-body (or surround imaging) portal scanners need to scrutinise closely all surfaces of the human body with millimetre scale precision. At airport departure lounge entrances, the market requirement is to detect threats such as (sheet) explosives and contraband within about one second, in a walk-through system, with a high probability of detection and low false alarm rate. At entrances to transport networks, shopping centres, arenas, schools, corporate headquarters, office complexes, churches and public servant buildings the requirements will be to detect a diversity of metal and non-metallic weapons and threats.

2) Stand-off screening scanners provide an initial first layer of screening of individuals for larger threats (knives, guns, bombs, improvised explosive devices), at ranges out to tens of metres. Typical scenarios are approaches to checkpoints and entrances, and in public places where people congregate.

Cooperative and non-cooperative are two further categories of screening. The former is overt, where the subject is aware of being screened and may have to divest items in pockets, remove shoes or remain stationary for a few moments while being scanned. The latter is covert, where the subject is unaware of the screening process, respecting the legalities of the particular deployment scenarios.

Sensor systems may be active (as in radar) whereby man-made millimetre wave radiation is reflected off a subject and analysed or passive (like a long-wavelength infrared camera) whereby thermal (Planck) radiation emitted by a subject is measured. Both methods have complementary attributes, discussed more fully in Sensor Science.

Knife and gun crime is an area now being addressed for non-cooperative screening of people in public places. Evolving millimetre wave sensors are becoming more effective at recognising these types of weapons concealed on persons. Security services are immediately alerted about the positive results of non-invasive pat-down searches.

Rucksack bomb detection is also enabled by the transparency of fabric to millimetre waves. Should a device be wrapped in an opaque material, the sensor will detect this as an anomaly.

Body screening videos from a 30 GHz millimetre wave imager are shown:
1) Metal pot in a rucksack
2)  Empty rucksack
2) Beeswax disk - one centimetre thick 
3) Polarimetric millimetre wave radar simulation

Recognising threats on the body, particularly in areas difficult to screen using existing technology, is key to the newest methods of screening. The methods need to be unobtrusive and protect personal privacy. Machines running algorithms process sensor data to recognise threats. Human operators cannot assimilate the throughput data fast enough at entrances, so machine learning takes over and passes cues to security officers, so potential suspects can be engaged at an appropriate location. In future, these systems will be omnipresent, fully automated and screen adults and children alike, and all groups of people without discrimination or prejudice. The technology will be accepted as a necessary part of society and blend seamlessly into building infrastructure and street furniture of future towns and cities.

Clothing penetration of millimetre waves varies considerably over the band and this dictates the screening capabilities. Greater penetration at lower frequencies enables screening through thicker clothing and greater numbers of layers, whilst at the high frequencies screening is guaranteed only through thinner clothing.  Thicker materials, such as (shoe) leather and rubber, and materials which contain moisture, can be penetrated at the lower frequencies.

Penetration into human skin of millimetre waves is only a fraction of a millimetre, with 10% to 40% of this radiation being reflected from the body, the precise percentage being dependent on the thickness and moisture content of the skin. This varies with gender ethnicity, body location and age. The skin-model for humans is a key element of machine anomaly detection algorithms.

Recognising threats inside packages and concealed behind walls, under floors and ceilings is another capability of millimetre-wave sensors. Since many materials (paper, cardboard, plastic, thin plasterboard, carpet, thin dry wood) are transparent to millimetre-waves radar sensors can categorise types on objects concealed in enclosures.

Combatting human trafficking by screening fibre glass (refrigerated) and canvas sided vehicles for stowaways at road and ferry port (border) checkpoints is a current application of MMW sensors. The technique uses passive (or radiometric) imaging. The capability is enabled by the transparency of the fibre glass and canvas, and the opaqueness of the human body. Stowaways have high contrast against haulage vehicle contents and the canvas and fibre glass sides.

Health & Safety: Our environment continually bathes us in millimetre wave radiation wherever we are, as a natural phenomenon; it is non-ionising and extremely low in intensity, therefore completely non-dangerous to life. Active systems like radar deliver doses of this radiation, lower than that from a mobile phone.

Receiver Operating Characteristics (ROC) of detection-probability and false-alarm-rate are the performance metrics of security screening sensors. A good sensor has a high detection-probability and a low false-alarm-rate, meaning it can detect many threats without generating false alarms. Specific ROC characteristics have been set for the different screening scenarios based on the estimated costs of the risks. For a given sensor, the detection-probability threshold may be lowered at periods of heightened security, to detect more potential threats. However, this raises the false alarm rate, which slows the personnel throughput rate. For this reason End Users, equipment manufacturers and governments are demanding the highest detection-probabilities and the lowest false-alarm-rates for emerging personnel security screening sensors.