As a former Submariner and Sonar Technician (trained in oceanography and underwater acoustics), it's interesting that such low power could be used. Long range underwater transmission of sound isn't a new thing. It's possible to passively track targets for up to 3000 nautical miles when the audio gets trapped in the deep sound channel.
The device discussed in the article has only achieved a distance of 300 meters using a slightly modernized version of sonar transducers and receivers that have been around for a very long time. I've seen MIT ocean projects meet the real world and go poof, but it's good to see people are out there trying to figure things out again, in the world after the pandemic.
Acoustic modems are fascinating devices— you're transmitting at most tens of bits per second before error correction. To make such a paltry amount of data do anything useful is a pretty interesting challenge. You're limited to just bare telemetry and a few status flags / commands.
>you're transmitting at most tens of bits per second before error correction.
You can get in the kb/s range over a couple of kilometres. You don't have to drop to tens of bits per second until you're trying to transmit over many nautical miles.
but why do buoys or sensors need to transmit that data underwater? They could easily do that via radio transmissions above the surface.
to me, anytime underwater communication is concerned, it is solely to be in the realm of military submersed vessel communication. There's no compelling reason for any other communication to undergo the expense of underwater transmission.
I guess the other use case that comes to my mind is the oil and gas industry who have a lot of underwater equipment, which could conceivably use an acoustic modem as a backup device.
also substantially slower than the speed of light, while not the biggest factor is also slower than other forms of transmission.
As you said, there is a lot of noise that needs powerful error correction to be used, such as how the Reed-Solomon code was used in deep space communication. Prepackaging information before relying on wireless communication is usually the most necessary part of any reliably complex system.
My natural first reaction is what unknown effect will this have on underwater life? We know other types of sonar have had negative effects. Knowing that, this should be studied and considered when making any new underwater sounding equipment. You don't get to go "we never thought about it" now that we know about it.
If many of the people replying to this comment had read the article, they would know that the low power device in question sends data by altering its reflectivity. It's metaphorically like covering and uncovering a retroreflector to send signals to a person a mile away with a powerful spotlight or laser pointed at you.
It seems reasonable to be concerned about what this significantly stronger and likely highly-directed signal might do to biological things it encounters on its way.
I wish the article had a reference to the actual power used. It SOUNDS like it's quiet minimal. In which case, it'd be hard to imagine you are dealing with biologically devastating energy.
That said, the unidirectional nature of this approach also means that actual life that could be affected is a lot smaller than what happens with sonar, where the signal is blasted in all directions at a very high power.
This, or something very similar, was posted a long time ago here. I cant find it anymore, sadly, but it was a backscatter-based communication platform that used very low power in the underwater part...
And illuminated that with 210 dB sound.
I guess this is similar, and equally disruptive to sea life.
The innovation here is to make that unidirectional rather than omnidirectional. So seems reasonable to conclude that even if the beam is gunshot noisy, outside the beam is basically silent.
Maybe I'm not understanding, but isn't this reflecting sound back, generated by something else? It seems the device uses one millionth the power, but the thing stimulating the device uses an unspecified amount, that would increase dramatically with distance. Unfortunately this is an MIT press release, so it will be impossible to understand what's actually going on.
> ... fix the input electrical power to 150 W ... The 20 dB DI transducers push the uplink decoding range up to several kilometers.
That probably is close to a baby whale fart, but I doubt competing systems are running at 150,000,000 W, as the above comment suggests. In their test to get 60m, they used 1.8W. 1.8 MW seems unlikely, for the same distance, with competing tech.
There's a fundamental misunderstanding of what's going on here. But, that's to be expected, with how these press releases are written.
This is asymmetric communication. The node itself uses backscatter thus operates at few micro-Watts. The remote acoustic projector is the 1.8W, and can communicate with 100s/1000s(?) of micro-watt power nodes - similar to RFIDs, but this tech works underwater.
I understanding that, but claiming that the backscatter energy, or the nodes power usage, is all the ocean life (the context of this comment chain) will see is incorrect, as the previous comment did. They primarily see the excitation energy, which appears to be ~23db greater (harvesting efficiency) than the backscatter, at the node. This is the same as something sitting between an RFID chip and the reader would.
So, the sea life near the transmitter would still have a bad time. The sea life near the nodes would see 23db more* than an active node, assuming the same power could be used to transmit from the node. Correct? This seems logical, since the energy harvesting will come at a coupling and efficiency cost, which means significantly more energy in the water at that node. If you had a battery powered node, you wouldn't need all the extra energy, and instead could just transmit.
All these numbers being thrown around are the power usage of the node, not what the sea life actually sees.
* Maybe more, since the signal path is twice as short, meaning your SNR starts higher at the midpoint (node).
Does it matter if it is generating the sound or just intentionally bouncing it back? It is still a source of unnatural sound that as of now has unknown affects on the natural inhabitants of the medium the sound is traveling. While you may think the hairless apes on the surface can do whatever they want regardless of repercussions, others of the hairless apes choose to be more conscientious of their effects on the surrounding environment.
There is no more information, as described in TFA, for reasons of this information having not been discovered yet by the researchers studying the new tech.
This sounds like an acoustic analog of the radio tracking device that was disclosed by Snowden a decade ago involving a quarter wave antenna that was selectively grounded. It was trackable at 20 miles, and totally passive.
It could run for months on a single coin cell battery.
Observing the parameter being monitored and deciding whether to ground or not ground the resonant element.
"Passive" in this case meaning that none of the battery power was turned into RF. It did not produce any RF. It merely absorbed more or less RF to encode information.
The trove of information released around the time of the Snowden leaks are not all necessarily from Snowden. There was a second leak that everyone likes to attribute to Snowden. Most of the info on the gadgets that I had seen seem to all be from the second source. But, yeah, they had some pretty cool toys.
"Passive" in this case meant it didn't emit its own signals per se; active sonar emits an audible noise (which can easily be traced back to an origin), passive sonar just listens.
It might work at distances longer than 300m, but "they ran out of space on the dock." That seems like such a silly limitation to bring up twice in the writeup.
It's an MIT press release. Technique probably doesn't work at any interesting distance, but this way they can make it sound as if they've discovered something incredible without ever having to acknowledge that they know it won't work as described.
It seemed pretty clear that they did a prototype scale test and then the rest was computer modeling. I'm actually shocked they calibrated it to within 1dB of reality and am curious how they managed to prove that one.
I agree about academic press releases generally, but it seems to work as described? If their model is right and they build a larger prototype, which is normal scaleup from academia, a kilometer scale underwater RFID tagged sensor buoys responding to a transmitter seems like a reasonable if optimistic claim. They can't build that before modeling, and they can't model without a small scale system to cross-calibrate with.
This press release actually seems better than most to be honest...
This isn't a product that's already on the market, so what matters is how far it can get if actually commercialized - which is pretty impressive at such low power.
They were probably testing this in the Charles River, where the MIT sailing club meets— that's also where the marine autonomy lab [1] does their field work.
If it’s “long distance” put the unwieldy piece where you have reserved space and move the less complex piece to borrowed space. Like another jetty, or a beach, or here’s a crazy idea: on a seaworthy vessel. You know, to test your marine communications device on a marine vehicle?
Yeah, I've worked in this space for many years (more on the autonomous vehicle side, but certainly work with acoustic folks) and am very familiar with the WHOI affiliates and the ONR sponsors. You're not wrong. The multiple reference to the dock being too short to perform longer range testing also struck me as odd. It's very simple to get an academic or industry partner to strap a device on a two boats for a test. It's done all the time. WHOI is always testing vehicles. Something's not adding up.
Wonder how well it would work for things like RC submarines. The distance increase isn't all that valuable for RC but the lower power required sounds very useful.
This is cool. I'm not sure if I missed it, but did they address how a signal is generated by the array in the first place?
My current understanding is that the device will echo a signal back to the source, not propagate it forward or create a signal.
Is it right to assume low power sensors exist and would be hooked up to this array, and then when the sensor triggers the array echos the signal to the other receiver?
The device discussed in the article has only achieved a distance of 300 meters using a slightly modernized version of sonar transducers and receivers that have been around for a very long time. I've seen MIT ocean projects meet the real world and go poof, but it's good to see people are out there trying to figure things out again, in the world after the pandemic.