SOUND DETECTION: Microflown says its acoustic vector sensors can be mounted on drones, helicopters and other aircraft as well as on vehicles, buildings or even soldiers’ epaulets. Image: Courtesy of Microflown AVISA
More In This Article
Sensors originally designed to alert pilots of single-engine planes to the location of nearby aircraft are instead finding a military role locating unseen battle threats as far away as 40 kilometers.
The basic technology—the single hot-wire anemometer—has been around for decades, measuring heat dissipated by air (or fluid) movement. A new technology from Dutch acoustics firm Microflown Technologies, BV, uses two far smaller heated wires to measure individual air particles affected by sound waves. About the size of matchsticks, the devices are called acoustic vector sensors. They passively glean information about the source of a sound—including its location, what made it, and its movement, if any. Sounds picked up can emanate from howitzers, helicopters, sniper fire or even conversations.
The sensors' small size and relative simplicity have militaries worldwide, including those in the U.S., India, Singapore and the Netherlands, testing or deploying them. "I was flabbergasted when I first saw this system" in 2011, says Col. Harold Jacobs of the Royal Netherlands Army. “I was really surprised about the simplicity, the amazing accuracy, the size and all the possibilities,” he says.
Microflown says they can be mounted on drones, helicopters and other aircraft as well as on vehicles, buildings or even soldiers’ epaulets. On drones, the sensors would augment what Jacobs calls the "deaf camera" onboard. When an acoustic vector sensor detects a disturbance in the air, it could signal the camera to rotate quickly and capture images of a nearby mortar crew, for instance, that has just fired a round. In such a scenario, that initial mortar round, detected by an acoustic vector sensor, could be used to gather enough data to accurately direct return fire before the enemy relocates or even reloads.
The sensor's probe is one millimeter wide, two millimeters long and 300 micrometers thick. It consists of two resistive platinum strips, each 200 nanometers thick by 10 micrometers wide, stretched parallel across a gap and heated to 200 degrees Celsius when operating. Air particles flow through the gap causing temperature variations in the strips, enabling the system to do two things: First, it counts the air particles to measure sound intensity. Second, without the need for two other sensors to triangulate, it records particle movement, which reveals the x, y and z coordinates of the sound’s source. Researchers have measured its precision at less than 0.1 degree C within one kilometer. A garden-variety PC running Microflown's sound-filtering software—which eliminates engine, propeller and environmental sounds—performs real-time calculations, says Hans-Elias de Bree, co-founder of Microflown and inventor of the sensors.
Engineers have used heated, single-wire anemometers to study airflow for at least 50 years, says Rich Lueptow, a Northwestern University mechanical engineering professor. Microflown seems to be the first to have seen the promise of a miniaturized, double-wire probe, he says.
Andi Petculescu, an associate professor of physics at the University of Louisiana at Lafayette, says he considered buying one of Microflown’s sensors, but found it too expensive. Petculescu, who studies topics including quantitative acoustic gas sensing, says he wonders if Microflown would need an ungainly database of sounds against which to compare noise in the field.
De Bree, however, says, "We analyze certain classes of sound and distill the properties, which are sets of statistical features." Classes include the shock wave of supersonic bullets, muzzle blasts, propeller-driven aircraft and so on. "A shock wave has a set of 10 properties," assigned to it in the system, for example, de Bree says. The system has a sharp memory, too. It identifies a sound source's unique signature after hearing it once. This information could be used, for example, to create a profile on an elusive sniper by plotting when and where attacks happen.