The Pentagon identifies people from afar with a heartbeat

The Jetson laser prototype can pick up heart rate signals from a distance of 200 meters, even through clothing.

The Jetson laser prototype can pick up heart rate signals from a distance of 200 meters, even through clothing.

Everyone's heart is very different. Like iris or fingerprints, our unique heart rate sign (or "signature") can be used to differentiate each other. The important thing is that heart rate discrimination can now be done remotely.

According to Technology Review, the US Special Forces is very interested in this issue along with other long-range biometric techniques including gait analysis, determining someone by how he or she walks. The method is believed to have been used to identify an infamous ISIS terrorist before the US military launched a drone attack. But it is not always possible to use gait, the face to distinguish one person from another. Only the heart rate is unique, and unlike the face or gait, it is constant and cannot be changed or disguised.

Picture 1 of The Pentagon identifies people from afar with a heartbeat Photo 1 of The Pentagon identifies people from afar with a heartbeat
The device detects their unique "heart signature" with an infrared laser that has a range of 200 meters.

Long range detection

A new device, developed for the Pentagon at the request of the US Special Forces, can identify people without seeing faces. Instead, it detects their unique "heart signature" with an infrared laser that has a range of about 200 meters ((better lasers can have a longer range).

Infrared contact sensor is often used to automatically record patient's blood pressure and heart rate. They work by detecting changes in infrared light reflection caused by blood flow. In contrast, the new device, called Jetson, uses a technique called laser vibration to detect surface motion caused by the heartbeat. It works even in cases where the person in need of identification wears casual clothes such as a shirt and thin jacket (not applicable to thick clothes such as winter coats).

The most common way to perform remote biometric identification is face recognition. But this needs a good angle of view like a face to face, and this is hard to do, especially from drones. Facial recognition can also be mistaken for a beard, sunglasses or headscarf.

"Heart signature" has been used for security purposes. Canadian company Nymi has developed a wrist-worn pulse sensor to replace fingerprint identification. This technology has been tested in the UK. Jetson extends this approach by tuning a device commonly used to test vibration from a distance in structures such as wind turbines. For Jetson, a special gimbal has been added to target even an invisible spot, as small as a ¼ laser. The device will take about 30 seconds to process the information, so the device only works effectively when the object is sitting or standing.

Better than face recognition

Steward Remaly, of the Pentagon's Office of Counter-Terrorism Technology Support, said his team has developed algorithms capable of extracting heartbeats from laser signals. He stated that Jetson can achieve 95% accuracy in good condition and this could be improved further. In fact, it is likely that Jetson will be used along with facial recognition or other identification methods.

Wenyao Xu of New York State University in Buffalo has also developed a remote heart sensor, although it only works up to 20 meters away and uses radar. He believes that the method of recognition by heart rate is much more accurate than facial recognition . "Compared to the face, cardiac biometrics are more stable and can be 98% more accurate," he said.

However, heart rate recognition has a clear limitation that a database of cardiac signatures are needed . But U.S. armed forces in Iraq and Afghanistan regularly collect biometric data, so heart data can be added to that library.

In the long run, this technology can be applied more in other areas. For example, hospitals can monitor a patient's condition without having to connect them to machines.