By Kris Osborn*
What if a small recon unit of Army soldiers was advancing through hostile terrain on a high-risk scouting mission to find points of entry for a major follow-on armored ground assault, when it was suddenly hit, even overwhelmed, with a massive array of enemy small arms fire? The soldiers take casualties, run for cover and immediately engage the enemy, yet the attackers are deliberately in defilade or obscured from view. Where exactly is the enemy fire coming from? What if soldiers could immediately know the source of incoming fire?
Perhaps a sniper is, by design, hidden in leaves, brush or other kinds of camouflage? Wouldn’t it be useful to destroy the attacker before more soldiers were hit? If a small area of attacking fire were precisely identified, perhaps the recon unit could call for immediate air support to blanket the enemies with suppressive fire from above?
These prospects are now a reality.
Traveling at supersonic speeds, a bullet exiting a gun barrel generates acoustic “shock waves” propagating through the air from the tip of the projectile, producing a sound “signature” which can be detected by specially engineered sensors, according to Raytheon BBN engineers.
This technical process, simply put, saves lives…. as it enables soldiers to instantly know the exact location of incoming enemy small arms fire, offering an opportunity for a precise and lethal counterattack amid high-intensity combat. A technology which does this, made by a Raytheon subsidiary called BBN, already exists and has been deployed with U.S. Army soldiers. It’s called Boomerang, and a set of six different sensors can instantly find the source of incoming bullets from moving vehicles and stationary locations.
“The way a shock wave works is it generates and propagates at the speed of sound. While the bullet is moving, a stream of waves comes off the tip of the bullet that propagates through the air. From six sensors I can locate exactly where it came from,” Brad Tousley, President at Raytheon BBN, told Warrior in an interview.
Now part of Raytheon, Boomerang-maker BBN began with innovative ideas from three MIT professors who envisioned a way to engineer these kinds of advanced acoustics years ago.
“As a bullet travels down range, it generates a shock wave. The sensor locates the shock wave from the muzzle blast and the shock wave from the bullet,” Tousley added.
A second variant of the technology, called Boomerang II, can locate enemy fire from moving vehicles by using sensors on top of the vehicle and small sensor box processor compartment underneath the passenger side or rear of the vehicle. The acoustic signature is converted into a small 5-inch by 5-inch display providing audible warning of shots fired. A Warrior X variant is carried by an individual soldier, wherein sensors and processors are worn by the soldier.
“As the source of sound, a shock wave travels through free space. A sensor detects it and converts the acoustic energy into an electrical signal. We have a set of algorithms that know where the sensor is located and we take the data feeds from those sensors and fuse those…convert those into an algorithm which is fed into a processor,” Tousley explained. “The shock wave expands as a cone behind the bullet, with the wave front propagating outward at the speed of sound.”
Interestingly, Tousley’s description of the sensor analysis of supersonic shock waves aligns with a 2006 Montana State University essay written about acoustic gunshot detection, which explains that shock-wave detection is more accurate than purely tracking a muzzle blast.
“The supersonic projectile causes an acoustic shock wave that propagates away from the bullet’s path. The shock wave expands as a cone behind the bullet, with the wave front propagating outward at the speed of sound,” the essay, called “Modeling and Signal Processing of Acoustic Gunshot Recordings,” states. (Robert C. Maher)
Extending its analysis, the essay explains that a muzzle blast alone is not necessarily as reliable source for analysis.
“For most firearms the sound level of the muzzle blast is strongest in the direction the barrel is pointing, and decreases as the off-axis angle increases. The blast may also be obscured by barriers and other obstacles blocking the direct path between the firearm and the microphone location,” Maher writes in the essay.
Also, Maher makes what might be a lesser known point, that an acoustic suppressor such as a silencer might remove or decrease a muzzle blast, yet not interfere with the shock waves described by Maher and Tousley.
The Next Frontier…What about tracking anti-tank missiles and RPGs?
What if this kind of technical application could be used to track the source of even larger, more dangerous enemy attacks from RPGs and even anti-tank missiles? Taking the success of Boomerang to the next step to accomplish this is now the focus of Raytheon BBN to advance the mission functionality of source detection sensing.
The concept, now underway and being demonstrated by Raytheon BBN, is to engineer a multi-mode sensor able to synthesize acoustic detection with infrared detection to precisely identify the sound and heat signature of various forms of hostile fire such as attacking anti-tank missiles and RPGs. RPGs and anti-tank weapons naturally generate a larger heat signal than small arms fire, and some of them operate at very long ranges….much farther than small arms fire in some instances. Merging acoustic and infrared detection, therefore, could compare or analyze an acoustic signature in relation to a heat emission to offer new levels of precision tracking.
“Some in the market are focused on acoustics only and others do infrared. We are doing R&D to integrate our infrared technology with acoustic sensing,” Tousley said.
Merging, analyzing and organizing otherwise disparate pools or streams of sensor data, such as those generated by infrared or acoustic sensors, aligns in a seemingly optimal way with ongoing AI and Machine Learning applications.
“Almost everything we are doing is being threaded with AI and ML to make our current systems even better,” Tousley explained.
Advanced, AI-enabled algorithms can now, in near real time, discern multiple streams of sensor data, feed them into a vast database and instantly perform analyses, answer questions and generate possible solutions. Perhaps an infrared signature offers higher fidelity in a particular kind of attack? Better yet, perhaps renderings from a series of small acoustic sensors can combine with infrared signals to bring unprecedented levels of precision to a long-range incoming missile attack from miles away? What if an attacking round fires from a weapon which generates a heat signature much larger and more distinct than an acoustic signal? Or, a weapon might also be engineered with some kind of advanced IR suppressor and, by contrast, only generate a detectable acoustic signature? If an attacking weapon generates both, and the data is synergized, then perhaps detection can reach previously unavailable levels of precision.
Analysis can only be as effective as the database it operates with or draws from, a circumstance which explains why AI-empowered information systems can perform procedural searching and data-mining functions in seconds. All of this can be done by bouncing new, arriving information off of or against a seemingly limitless data repository. Not only that, but the computers can draw upon previous scenarios and analyze a number of otherwise disconnected, yet highly relevant variables in relation to one another such as wind speed, terrain, climate or the trajectory of an incoming projectile. This is where Machine Learning can bring new dimensions, as it involves a technical effort to perform near real-time analytics upon newly arriving, yet previously unrecognized information. Advanced algorithms can quickly integrate, analyze and help identify new details with the aim of placing them in context for presentation to a human decision-maker.
All of these calculations are engineered to, among other things, offer human commanders optimized choices for counterattack based upon a particular circumstance. For instance, perhaps certain weapons or methods of attack might be best suited to effectively respond to a particular attack? An AI-capable system could determine these kinds of informed solutions in many instances, based on its compiled database of prior circumstances and a host of interwoven, yet previously disaggregated variables.
It goes without saying that all of this pertains to current Army (and multi-service) efforts to massively decrease sensor-to-shooter time while lives hang in the balance. Part of this parallels the Army’s recent Project Convergence exercise, a demonstration at Yuma Proving Grounds, Ariz., which shortened sensor-to-shooter times from 20 minutes to 20 seconds using networked sensing and advanced AI.