Monitoring Changes in Breath Rate Can Prevent Terror
This post is also available in: 
עברית (Hebrew)
Olga Grosman-Ezrahi
Will crime be abolished in 20 years? Is there a way we can know in advance about the next murder or crime? In recent years great resources were directed to huge projects in the US dealing with just these issues. Dr. Nimrod Kozlovski, a partner in the JVP venture capital fund, explained technologies to predict crime during the Video Analytics conference, organized by I-HLS at the Israeli air force center in Herzliya.
After the US realized that perhaps the solution for crime and terror lies in predicting technologies, countless resources were directed towards research of this field. The massive American project called Precrime completely changed the way we understand the world of crime, seeing as how in the past 2,000 years, punishment for criminals was given after the crime was committed. Today, however, may people fail to believe the efficiency of law enforcement due to many failures to enforce tax offenses, internet crimes, etc.
This situation enabled in recent years several fascinating breakthroughs in the field of crime prevention using prediction. So, for example, credit card offenses were prevented, using a search engine that checks the client’s history of transactions up to a certain point and analyzes whether the purchase right there and then is normal. The system checks, among other things, which computer and mail address the client was using before, whether it fits their monthly expenses and so on. If it recognizes anything out of the ordinary it alerts the bank, and the credit card owner will receive a phone call to ensure that it was indeed their transaction. This way, by gathering information, creating profiles and crime patterns and basically predicting the nature of the deal, the system manages to prevent the crime before it takes place.
Another fascinating system meant to prevent crime is called FAST (Future Attribute Screening Technology). The system builds an integrative picture based on continuous video footage and sensors documenting human behavior from afar. It can detect physiological and behavioral indicators in humans that usually suggest a suspicious behavior and can alert a possible malicious intention to commit a crime or an act of terror. These parameters include dry lips, sweat, rapid breathing patterns, change in voice tempo or eye movement and more. The indicators, examined using external sensors, are compared to normal behavioral patterns of the average, calm person and point out suspicious individuals, to whom security services can be directed in just moments.
Today some of these inspections are being done by security personnel at the Ben-Gurion Airport, but the goal is to have a computer program do it automatically. The system’s accuracy reaches about 85%, relying on a statistical level, so you can’t say that it is perfect but it’s definitely a start from which the only way to go is up.
The systems owe their success to a number of different factors which were also developed in recent years, such as sensors broadcasting all the data to the process point in real time from a distance. But don’t think that is where the line is drawn, as more ideas for more innovative technologies in crime and terror prediction, such as using MRI images, are already taking form.
