.
6 years earlier, similar blog post: https://defense-and-freedom.blogspot.com/2016/10/acoustic-sensors.html
The typical forward observer of the Second World War did not enjoy a great vantage point, had no mast-mounted thermal camera with zoom and did not fly in an aircraft. The typical forward observer had thus a limited ability to see what's going on. I recommend looking around when you're driving in Europe; you can usually not see much farther than 2 km, unless there are some mountains or hills involved. There's usually something breaking up the line of sight, usually trees or buildings. There's additionally smoke from pyrotechnics, smoke from other burning stuff and dust involved, too. The typical WW2 forward observer needed to be able to assemble a situational picture mostly from what he heard. Only occasionally did he actually see enemies and could identify them as such.
The typical way of detecting howitzer batteries to counter them by counter-battery fires during the World Wars wasn't aerial reconnaissance (only). It was sound ranging, at night assisted by flash spotting when observers had a good vantage point (or were using a tethered balloon). The flash spotting could be countered with additives placed in front of the propellant charges, though. The sound ranging used the time delay between shot sounds arriving at different listening stations. It's a very capable approach when not too many guns fire at the same time, and not for too short a time (the sound takes about a minute to travel 20 km). Radar is much quicker and more accurate in calculating the battery positions, but it's dependent on treacherous emissions.
Counter-sniper detectors were all the rage before the jammers against radio command-controlled mines (fashionably: "IED") took over the lead as most fashionable must-have tool for moving around as soldiers in distant countries that didn't attack us and where people want you dead. The principle was not always focused on the muzzle noise; at least some counter-sniper acoustic sensor systems were rather using the sonic boom of the supersonic bullet, an approach that would fall flat against subsonic bullets as I pointed out before.
I will argue that we should pay more attention to acoustic sensors despite the availability of radars. I will further argue that the acoustic sensors are complementary to radars and gain much utility by digital datalink networking that's been all the rage for two decades anyway.
First, let's look at what benefits a battlefield vehicle (tracked AFV, wheeled AFV, scout car, unmanned ground vehicle) could have from employing a 3D microphone array:
- sniper detection (at least against supersonic bullets)
- approx. detection of relatively nearby (~20 km) artillery
- detection (triangulation) of AFV main gun shots (probably down to 30 mm calibre)
- detection of firing 60...160 mm mortars (normal mortar pattern)
- more accurate direction-finding (not triangulation) to tracked AFV noises
- detection and direction-finding to multicopters
- aiding in detecting and confirming subsonic anti-tank guided missiles
- helicopter detection, including non-line of sight
- limited non-cooperative IFF against helicopters
- giving AFV crews acoustic sensing of their surroundings when the hatches are closed
The mass and power demands of such acoustic sensors are negligible in the overall picture, the costs affordable. The volume of such systems and how the clutter for example a tank's turret roof might be an issue in some cases - especially when we're talking about upgrades that get patched onto a tank.
And here's the beauty: We do not need to develop, produce and retrofit or issue a gazillion different acoustic sensors to make practically full use of the technology's potential! We just need to develop one type and field it, and even its integration into a tank with an already cluttered rooftop won't be too much of an issue, either.
I've argued that we need a large quantity of decentralised anti-drone shunter-killer systems to cope with saturating drone swarms of the near future. The old approach of fielding one dedicated anti-air system such as the Gepard SPAAG of the 1970's won't cut it because drones can attack from within trees, below trees, from within buildings. The drones may switch between flying and ground movement. The typical attack of a drone swarm could be in areas of short lines of sight rather than over open fields. We need many defence systems to cover those areas, and their range can be short without this being much of a downside.
My recommendation is thus a remotely controlled weapon station (RCWS) with a rapid fire weapon (7.62 mm MG3 with high rate of fire might be enough, the maximum would be a 30x113 mmB autocannon with AHEAD shrapnel with .338 machineguns and 20x102 mm autocannons with self-destructing HE-T in between). Such a RCWS needs identification and targeting sensors (thermal camera with zoom, electro-optic camera with zoom, laser rangefinder), but this wouldn't be enough. Even an automated search (rotating gimballed sensor package on top) would not suffice in my opinion. The challenge wouldn't be the targeting, it would be 24/7 readiness and ability to detect threats in time. Diving loitering munitions that have already shut their engine down are a huge challenge, but at least all those rotary flying drones could very well be detected by acoustic sensors at useful distances. So the base of the RCWS should have a microphone ring and the gimballed sensor head on top should have a microphone as well (for calculating the elevation angle towards the noise source).
The very same RCWS could be mounted on a MBT, IFV, APC, other AFVs, 8x8 logistic vehicles, 6x6 logistic vehicles, 4x4 vehicles and some unmanned ground vehicles (UGV). Almost all vehicles except assault bridgelayer tanks and cabrio low profile scout cars would qualify (remember, I'm against red cross-marked vehicles on the battlefield). An American-style combat brigade has approx. 4,000 personnel and approx. 1,000 motor vehicles. Imagine it having approx. 900 such RCWS! Saturation attacks by drone swarms would still be possible, but the bar for them would be much higher.
All these vehicles would have a digital radio datalink capability anyway (yes, even the logistic vehicles), so all these sensors could transmit their interesting sound readings with timestamp and direction to a server at a battalion command post which would then be able to determine movements and firing positions of hostiles (and friendlies). Target acquisition drones could be tasked to look at the acoustic contacts and would find the exact position of AFVs and mortars. Air threat warnings could be issued about moving multicopter swarms or helicopters, and fibreoptic-guided missiles (or fast radio-controlled loitering munitions) might be launched to engage hostile helicopters and artillery beyond the line of sight.
All this, and not one radar would have directed its beam towards hostiles. Just the (anyway quite unavoidable) radio communications would be required, and those could use tiny sub-kilobyte message packets.
Quite importantly, the sensors could also confirm the absence of noise sources. A rumour about a tank assault could be debunked by the absence of tracked AFV noises in that area, for example. Reports about the absence of hostile activity are hugely useful to clear the fog of war, one should keep that in mind.
related:
/2017/08/very-low-level-air-defence-against.html
/2018/05/summary-modern-air-defences-for-europe.html
Part II: Soon, promised, I already know what it will be about.
S O
edit March 2024: Ukraine uses a microphone network to detect and track flying kamikaze drones.
.