Veteran submariner on how sonar crews tirelessly hunt enemies they can't see | Hacker News Reader

Veteran submariner on how sonar crews tirelessly hunt enemies they can't see

thedrive.com

117 pointsclouddrover79 comments

I'm not normally one to promote Hollywood but the film Greyhound (2020) with Tom Hanks does an _excellent_ portrayal of this technology (albeit slightly more primitive) in fighting German U-Boats in WW2. Well worth a watch if this interests you.

I was just going to say exactly this.

It’s amazing to see them doing this with such incredibly basic technology.

- Having to draw their ship and the submarine and constantly update its movement.

- sonar guy is just looking at a screen that blips every now and then.

- all communication by morse code - must be horrendously slow in the heat of battle

- even communication within the ship is passed between multiple people so big chance of Chinese whisper type errors.

I don’t know how realistic it is to real life but it gives you an appreciation for how smart and situationally aware the captain of the ship needs to be.

They wargamed live during North Atlantic convoy attacks on the floor of a building in Liverpool. Players received instructions through letterboxes around the room perimeter and each could only see a fraction of the the battle. They could not see the u-boats at all as those were same color as filters they looked through. Information was radioed in from convoy escorts and scenarios were played out. Battles could last a couple of days.

Game of Birds and Wolves! Great book

There is a nice video of Lindibeige about that, with some old photos of the setup https://www.youtube.com/watch?v=fVet82IUAqQ

The book by C.S. Forester that the movie is based on is even more realistic, though the movie is pretty good.

To really dig into it, Samuel Eliot Morison's magisterial History of Naval Operations in World War II does a good job talking through the development of both the technology and organization of naval operations.

Not all communication was by morse code. We did use ship to ship voice radio, but it was a tradeoff. RDF allowed the enemy to triangular the position of the broadcasting ship, and (often) the enemy could overhear messages and infer some tactical details.

Sensor systems changed wildly throughout the war. Sonar improved markedly, and radar went from non-existent to able to identify altitude and distance to individual planes. The weight of the additional sensor systems and controls systems made some ships unstable, so we removed weapon systems from them.

Being a captain (or, heaven forbid) and admiral, was very demanding. Sleep was hard to come by, and the fog of war was intense. We had meaningful problems with friendly fire in night engagements even with radar and radio communications.

> Samuel Eliot Morison's magisterial History of Naval Operations in World War II

Wow, you weren't kidding. Can you recommend any shorter tomes on history of naval technology?

https://www.amazon.com/History-United-States-Naval-Operation...

It's surprisingly readable. Good history is storytelling, and Morison was there for much of the history he relates. You could get a great picture of the variety of modern naval warfare by picking up the volumes on Midway and the Struggle for Guadalcanal.

C.S. Forester's novels, while historical fiction, are also excellent.

I'd also recommend Japanese Destroyer Captain by Tameichi Hara, another powerful first hand account. Reading Hara and Morison's accounts of a battle in which both participated is really interesting.

Man, that looks like a hard series of books to actually get a hold of. Either spending multiple hundreds of dollars for a used set, or struggle to find individual copies for $15+ each (a quick google search seems to show Vol 2 unavailable anywhere I looked).

I think there's a new paperback printing under way. Happily, it's not the end of the world acquiring them piecemeal, as each volume tells a coherent story and takes some time to read through!

Many libraries have it. I remember spending many hours in the 940.54 section of my local public library thumbing through volumes when I was a kid.

What you are seeing is the very evolved version, the later-war version of this battle. Early in the war it was much more primitive. There were incidents of ships literally hunting their own propellers. Only once procedures and training evolved did the real cat-and-mouse game become even possible.

The military has a partial solution to the “chinese whisper” type error where the person receiving the command will repeat it back so the other person can make sure it was understood correctly. This has the downside of slightly slowing down communication.

That movie and Master and Commander are two movies that helped a digital nerd like me understand how an incredibly complex system can be managed by lots and lots of humans playing very specific roles.

Your comment made me remember the book Cognition in the Wild:

https://mitpress.mit.edu/books/cognition-wild

A film set is the same way. Also, sounds like a lot of "machine learning" shops. Hire a squad of humans to do the work until the ML/AI graduates from schooling.

Massive history nerd and I really enjoyed Greyhound because they didn't over dramatise it (much) and the attention to detail in the roles, speech etc was all very well done.

I personally haven't seen Greyhound, but if you found this article interesting, the 2019 French film Le Chant du Loup is entirely about sonar crews on modern submarines. It's definitely a fun movie that seems (to an uninitiated) pretty technically accurate.

https://www.imdb.com/title/tt7458762/

That movie is indeed very good to watch, intense, and portraying what is described in the article: war on a submarine through the eyes (or ears) of a sonar man.

The French trailer looks good. It looks like the film has been dubbed in English. Thank you for the suggestion!

Tom Hanks wrote the film as well. He has done a lot of good in terms of WWII history. Greyhound was one of those movies that really brought to life just what combat looks like on the water. To an outsider, naval combat seems rather benign and “far away” compared to Infantry, but the reality is that naval combat is terrifying and up close. It was an incredible film in capturing the horrors of naval warfare. Just seeing how hard it was to be a captain (or even a sonar man,) gives me redoubled respect for those that serve on the seas.

And Greyhound depicts a relatively calm series of engagements, compared to what actually happened. Consider the U.S.S. Borie:

https://en.wikipedia.org/wiki/USS_Borie_(DD-215)

which rammed (!) a U-Boat in high seas, was repeatedly picked up by waves and dropped on top of the surfaced U-boat, whereupon a small arms engagement broke out (!) until the U-boat attempted to disengage and was damaged by the Borie's main guns. Both ships sinking, the Borie attempted to rescue the German submariners until a torpedo from another nearby submarine caused them to reconsider. The Borie then promptly sank from seams opened by being repeatedly bounced off the top of a submarine and most of the crew were rescued.

the British cracked Enigma in 1941, and from then they were able to completely destroy the 300-400 German u-boats, because they were intercepting and decrypting u-boat radio traffic, so they always knew where the u-boats were going. The only reason the Allies allowed the u-boats to continue winning small victories was to preserve the illusion that Enigma was still secure to prevent the Germans from switching to another cipher system. So when you see depictions of exciting historical battles against u-boats, keep in mind that it was not a level playing field due to the strategic advantage of superior SIGINT by the Allies.

IIRC 75% of U-Boat crew members were killed in action during WW2.

The book "Marching Orders" by Bruce Lee tells the story of Magic intercepts of the Japanese military and diplomatic code and reviews the daily summaries read by Marshall and other US leaders in the context of what was happening in the war at that time The point of the book, c.1995, was that signals intelligence played a huge role, but unknown role, in influencing the actions of the U.S. and U.K. And that without understanding what the leaders knew there is a tendency to overestimate their skill / leadership ability.

And Poland cracked Enigma in 1933, and then shared their techniques and Enigma clones with the British in 1939!

Having read the book and then watched the movie, it’s amazing how much technical detail you don’t even get to approach in the film. This is just the nature of the medium, but the book is very focused on the captain’s inner thoughts and technical descriptions, awareness of limitations, and homages to the mental math involved. Highly recommend the book if you at all appreciated the movie.

Which book did you read?

I just looked and it looks like there are two books by Forester. One is The Good Shepherd which is the original book and the other is Greyhound which is also by Forester and is sold as the movie tie-in. They have different number of pages, so I assume they are different.

Edit: I ended up buying The Good Shepherd because it was only $0.99 for the ebook.

If you enjoy The Good Shepherd, The Ship is a similar book about surface combat in the Med.

For a detailed and in places dramatic book on the primitive technology of WW2 surface naval warfare, I can recommend "Neptune's Inferno": https://erenow.net/ww/neptunes-inferno-the-u-s-navy-at-guada...

While radar existed, it had similar problems of manual plot, was only on a few ships, and the commanders had not updated their training and doctrine to match.

Several of the key engagements around Guadalcanal were fought at night. Without radar, and without any reliable means of telling friend from foe other than eyeballing the shape. Total chaos.

Thanks for this reference to Neptune's Inferno by James Hornfischer, it was a fantastic book. I like two quotes in particular

"The epic of the Pacific war found new chapters for everyone. The endless game of personnel-rotation musical chairs saw the continuous replacement of the experienced by the inexperienced, until, by the end, only the experienced remained."

"Ascending to theater command had never been his ambition, for ambitions, he felt, were meant not for personal gain but to pursue common goals within the established order of a group."

A description of Chester Nimitz

I like this description of ambition as a model for leadership: "pursue common goals within the established order of a group." I think it's a good one for startup CEO's as much as any other leader.

There's actually a former US Navy Sonarman who has a youtube channel called Sub Brief and it's amazing. He left the service in 90s (IIRC) but he shares all kinds of interesting knowledge on how torpedos and mines work, all the different kinds of accidents and mishaps that happen - he even analyzes different disasters like the Kursk and how they likely happened. I highly recommend it if you're into subs.

https://www.youtube.com/c/JiveTurkeyCrew/featured

JiveTurkey and Aaron Amick (author of this article) are the same person :)

I recently got into realistic combat simulators (Squad, Arma, DCS, Cold Waters ... there are some other sub sims out there too) and I had no idea JiveTurkey and Aaron were the same person. Maybe I just missed it or wasn't paying attention.

Anyway, if you're a military geek and want to get into realistic naval combat sims play Cold Waters (Uboat is also good, but way different play style). Like all realistic sims the learning curve is pretty steep, but during lockdown having something to concentrate on and enable a bit of a disconnection from the rest of the world for a while is nice.

688(I) Hunter/Killer with the Jane's manual is great too. I find it more immersive than Cold Waters as you are in charge of the waterfall and have to move the cursor like a true sonarman to find the good signature.

If you liked 688(I), I strongly recommend giving Dangerous Waters a shot. It’s a sequel made by the company, featuring submarines as well as surface and helicopter platforms, and is currently on sale.

https://store.steampowered.com/app/1600/Dangerous_Waters/

My father used to help design and build sonar of varying types for the Australian Navy (RANRL) when I was a kid. I used to love hanging with them all and even got to go on a submarine a few times. He had to be fully qualified for sea trials. I assume a lot of his work was adapted from other countries, but he still will not really get into details about what he did for them (70s and 80s). One little nugget he told me a few years ago: their entire UI/UX team were psychologists, engineers played almost zero role in that aspect.

Oh man, I worked on the software for these systems for 6 years :)

specifically the ones from this image https://www.thedrive.com/content-b/message-editor%2F15972209...

During the holidays we'd pump Christmas music through the sensors. We could do this since that's what the raw data is anyway -- sound. Made for interesting sonar displays as well as background music.

If you can say, what does the dev version look like, physically? I assume you don't have a dev ship? So same HW, just in an office? Or remote access to HW in a server room? Or is there a more desktop friendly form factor? Or all virtualized?

It would be fun to be that guy saying but dev/QA/prod should all look the same.

This[0] youtube channel is quite good at explaining how subs work

[0] https://www.youtube.com/c/JiveTurkeyCrew

The author of this article (Aaron Amick) is Jive Turkey, the same as the host of that YouTube channel [0].

His other articles for The Drive [1] are all worth reading.

0: https://www.thedrive.com/the-war-zone/24638/retired-submarin...

1: https://www.thedrive.com/author/aaron-amick

I came to post Jive Turkeys YouTube channel. The highlight are his episodes where he takes an underwater recording, uses some Adobe audio program, and extracts all sorts of interesting detail about what we’re hearing. When you see how much he exploits from a generic public tool, makes you wonder how much more comes from super advanced, specialized tools the navy has.

> The second step in classification requires analysis. It takes 30 to 45 seconds to get a look at a new detection and discover what kind of engine she’s running, what kind of hydraulic pumps are online, and what screw blade configuration she is using.

One can imagine ML making this job somewhat obsolete in the not-too-distant future.

https://xkcd.com/1897/

Not sure why this was downvoted - it almost certainly is heavily automated already, and the article hints at it. Tom Clancy described a primitive version in The Hunt For Red October in the 80s. You'd expect the Navy to be very interested.

That's if they aren't already.

Without having the data I can't try but I'd guess that Fourier analysis (or something a bit more clever to deal with noise e.g. Welch) would get you a lot of the way there even without invoking AI

Even if you're also using ML, Fourier transforms can be pretty useful in this sort of analysis.

This is not based on any classified information and sonar signal processing is not something i know much about, but cyclostationarity (and other higher-order statistical signal processing) does tend to be useful when processing signals that are generated by processes where there's multiple periodicities at work: https://en.wikipedia.org/wiki/Cyclostationary_process#Angle-...

Signal separation techniques for characterizing cyclostationary signals are also more robust to noise. For example, analysis of the second order FFT can resolve the signals of incipient faults in rolling bearings because it can resolve the signals of the bearing components rotating (at a much lower frequency than the shaft) whereas simple FFT analysis would generally only measure the magnitude of the structural resonance of the machine, with spikes at the mains, rotational, and slip frequency.

If you ever watched the hunt for red october, they already had computers doing this (doubt it was ML'd tho).

It is ML depending on what we're calling ML. Is gradient descent/MLE ML? If yes, then the Dallas (the US sub) in Red October was using ML.

Clancy did embellish a few things and speculated rather a bit to make an entertaining scenario.

All told, a good solid read, and accurate in a bunch of not-quite critical details, but don't lean too hard on the edges of the story.

Same with Red Storm Rising. Very well told, close to a solid "what if", but lots of little touches to make it entertaining as well.

Is it really something for ML? Shazam isn't ML, and it's the same basic thing. Take the current sound and then compare it to a huge database of known sounds. Report the result.

Agreed. ML and AI are (currently) useful in situations where there is a lot of data, of different types, usually from disparate sources, often sparse, with no known pattern or causal relationship. Shazam and sonar analysis don't really fall into any of those areas.

Can confirm that USN and sonar systems have been using "ML" techniques since the 90s. Y'all were late to the game!

I was at a day of research presentations in January before everything shut down. One of the presentations was on using ML to identify individual dolphin speech signatures--with some degree of accuracy. One of my thoughts was that the more sophisticated navies probably already have this sort of thing pretty well developed.

From TFA: "Patterns are matched and measured with onscreen tools and cursors. Automatic artificial intelligence algorithms flag potential targets for the sonar operators to verify. "

Somewhat related. I went to the Oregon Museum of Science and Industry (OMSI) in Portland last fall and did a tour of their submarine, the USS Blueback. Our tour guide was a retired Cold War-era submariner and he was fantastic. He gave us tons and tons of details about what life and working on an active sub was actually like. It felt like being there, and you could tell he delighted in being able to share his past with people.

I've been meaning to check it out, it's like right there too.

This[0] is an interesting account of an accident with a submarine and how sonar played a role in the event.

[0] https://en.wikipedia.org/wiki/Ehime_Maru_and_USS_Greeneville...

Obviously, being a skilled sonarman would seem require a good physical sense of hearing (frequency range, etc.), but as as the article mentions, the ability to recognize patterns is also important.

Have there been any studies or analyses to see if there's also a correlation between skill at, say, music (relative pitch, etc.) or language (detecting subtle differences in accents, etc.) and sonar skill?

The best sonar operators have perfect pitch, except instead of saying "that is an A-flat but there is a ghost harmonic because the slot in the bridge for that string is too wide" they would say "that is a 415 Hz sound with up doppler. sounds like a 3 bladed prop with a notch on the second blade". It was pretty amazing to watch the really good ones in action.

They would go to a bar and comment that the ceiling fan was spinning at 23 rpm, and the compressor motor on the cooler had a bad bearing.

Blades are numbered? I would think each blade would be indistinguishable from the next, so any of them could be the 2nd one depending on where you started counting.

"Over selectivity" is considered a handicap (to the point of disability) by some musicians; when they play two different tone sets and say "thats the same note"... if you can't hear it that way but hear 440hz and 850hz or whatever, it is difficult to communicate.

I used to be a part of this when the tracking party was manned and I was engine room sup.

Using the computer tools For fusion was definitely interesting (90’s X11 gore), with a large palette of modeling tools, but the coolest experience was being taught to time/range plot on paper by the mustang weps.

Articles like this leave a bad taste in my mouth. They venerate the subject and participants involved and leave out the vast amount of detrimental effects they cause.

Whales, other marine mammals, and other wildlife are affected by these systems, and they are never considered. The Navy is currently still trying to increase its weapons testing off the coast of the Pacific Northwest, bringing further harm to the already decimated and declining southern resident orcas.

We have absolutely got to start understanding how our technology affects the world and making changes.

The systems described in this article are "passive". They are simply listening to the sounds around them. This is much more common for submarines than "active" sonar - which ensonifies the water around you and then you listen for echos. It is those loud pings which can threaten or frighten wildlife. However they are also like holding up a big lit sign that says "there is a submarine here!" - and the whole point of submarines is stealth. They say that if you hear a ping from a submarine, it is too late - there is already a torpedo heading your way.

source - used to be a sonar tech

These articles leave a bad taste in my mouth because they give far more credit to the sonar shack than they deserve, with the implications that they're motivated and nobody is asleep at their console.

I've heard the sonar ping from a submarine is so powerful that it will kill you if your a diver near the sonar dome.

235db! Apparently also kills wales and dolphins.

https://physics.stackexchange.com/questions/93222/could-subm...

The only saving grace is that they only use it as a last resort since sending out that kind of noise would let everyone in that hemisphere know exactly where you are.

Keep in mind that's likely not related to the dB sound pressure level in air that we're all familiar with. dB is a logarithmic ratio to pressure to a reference. In air, that reference is 20µPa, acoustic pressure in water typically uses a different reference (1µPa). There is also an adjustment of 36 dB to account for the higher impedance of water. So 235 dB in water is roughly 173 dB SPL -- still loud, but many orders of magnitude lower.

[0] https://fas.org/man/dod-101/sys/ship/acoustics.htm

This is what you can expect https://www.youtube.com/watch?v=sCmyZYYR7_s&t=90s

This diver was near a sub base for training.

Comments say this was very likely not a sub‘s sonar.

Oil and Gas exploration in the sea routinely uses extremely loud sounds, because explosive geosensing. Its screwing up cetacean migration and mating. It travels miles and miles.

I can attest to this being much louder than any active sonar in the water.

Usually it is surface ships which get that loud. In harbors where sabotage is a concern it is not uncommon to run a regular ping as a deterrent to underwater saboteurs.

Reply on news.ycombinator.com