Hidden sensor network detects explosives

Sensor network // Source: isa.org

A covert sensor system designed to identify and track people carrying explosives at busy transport hubs has been developed by researchers in Germany. Dr. Wolfgang Koch and his team at the Fraunhofer Institute for Communication, Information Processing and Ergonomics (FKIE) have built a prototype system named HAMLeT (Hazardous Material Localization and Person Tracking) to alert security staff to individuals intent on carrying out a terrorist event.

Ellie Zolfagharifard writes that the system works using two separate sensory networks that gather chemical and kinetic information. The first is made up of a series of four to six rotating laser scanners that send pulses through corridors, walkways, or escalators at airports or railway stations.

By calculating the time taken for the laser pulse to return the device, the scanner is able to measure the distance between the objects and construct a 2D image of the area.

This image is combined with information from a second network of electronic sensors hidden in air vents and wall fixtures that provide chemical data on explosive materials. Oscillating crystals on the sensor chips capture any threatening chemical molecules in the air and identify their composition based on the frequency of their oscillation.

Koch explained that used on their own, these sensors would be unable to provide the detail required to identify individual suspects. He said: “The basic idea of HAMLeT is to overcome the fundamental deficiency of sensors in assigning chemical data to a source by using sensor data fusion. The infrastructure at airports and railway stations always includes certain tunnel-like structures. If these tunnels are equipped with a number of sensors we can combine their output with kinematic data to identify high-risk individuals.”

Once all the information is gathered using these networks, it is analyzed by algorithms based on Probabilistic Multi Hypothesis Tracking (PMHT). Any patterns identified by the algorithms are fed into advanced CCTV software, which automatically marks members of a crowd with a green, amber or red status indicating their threat level. Security personnel are then able to use their judgment on whether further action is needed.

Koch said: “What we have in mind is a system that is hidden and causes as little disturbance to public life as possible. We think this is probably a better alternative to the proposed full-body scanners as it will not cause journey delays or invade the privacy of passengers - one of the main concerns about some terahertz body scanners.”

Zolfagharifard that in a trial involving the German armed forces, Koch’s team proved the system’s ability to track down five individuals carrying hidden explosives in a crowded entrance way. The team has also trialed the system in military harbor security and ferry traffic applications with promising results.

Koch said: “Of course, there are still some operational problems with the system. If a terrorist is aware of this type of observation they could try to prevent the smell or mask it with perfume. Another problem with the chemical sensors is that other things need to be taken into account, such as air ventilations that distort the results. We’re working on algorithms to solve this.”

The team is also working on a model that incorporates gamma-spectrometers for the identification of “dirty bombs.” These combine conventional explosives with radioactive material and could cause local contamination to an area. Further research will focus on the ergonomic design of tunnels and methods of reducing overcrowding to reduce false positives.

Koch said: “We don’t want to produce a George Orwell state with our technology. Although these systems are hidden, if correctly used they can maintain privacy while addressing the terrorist threat. Our results are encouraging and we hope to develop a robust system within the next few years.”