Has anyone here worked on some kind of emergency surfacing device for underwater vehicles?
I’m building this AUV, and we have some work coming up where we’re attaching some very expensive sensors too it, and I’d like some piece of mind that there is a back up plan if this doesn’t come back up on it’s own.
We have a timer internally that tells the AUV to go back up after a set amount of time, but ideally I’d like to find something that’s independent of the main control system to add redundancy.
I am working on something that sounds similar;
A small self-contained ‘antenna’ with LoRa, GNSS reciever and lights. Will flash lights and transmit position periodically when activated, can be activated by a number of mechanisms including some fail-safes such as mission time-out, lack of FCU activity, lack of sufficient main battery voltage, leak detection. Potted with a hole and loading path for easy recovery when you find it.
Not super far along with it but I’m happy to compare notes.
do you have any pictures of this you can share? And does this deploy subsea and float to the surface on it’s own?
I’m using and ELRS system on the surface, but I’m concerned if it fails underwater I won’t be able to find it.
It’s just a breadboard of the electronics at the moment so no pictures, I’ll try and remember to circle back to this and post them when its further along. I’ve not used ELRS but my understanding of it is that its a protocol you can send over LoRa. What I’m building is intended to be rigidly attached to the vessel like a fin, the assumption is that the vessel is slightly positively buoyant.
Edit: I believe there are such things as emergency floatation devices that use small CO2 cylinders to inflate an airbag. I imagine they’d be a pain to integrate without ruining the hydrodynamic profile of your AUV though. Slight positive buoyancy is a more normal approach.
I was looking at the little Swim rescue things as well. They would work in shallow water, but I need a way to activate it reliably, ideally without connecting into the main power supply.
Have you looked at these SPOT trackers? if you could solve the problem of making it pressure tolerant, then it might be a quick fix to get a surface position.
Most AUVs are positively buoyant for this exact reason. but they use a bit of power to stay at depth, so not ideal for my application. My AUV has a buoyancy engine to let it go up and down like a glider. but if this fails, it could get stuck at the bottom. most gliders have a drop weight on the back to get around this. when they’re in trouble they drop the weight, but they also rely a lot on the internal electrons working to tell it to drop the weight.
I might be over thinking it some, but a simple solution would really give me some peace of mind when it’s deployed.
Can you mechanically bias your buoyancy engine? That way when the power is removed it will tend towards positively buoyant
It’s possible to have chemical releases as well, by attaching it with a water-soluble material, but then there’s a maximum submersion time per dive (which may be somewhat variable), and a new consumable to manage.
If you’re specifically concerned about the buoyancy engine getting stuck in a non-positive configuration, you could potentially add a mechanism to agitate it if it’s not changing, and/or add an acoustic beacon to make it easier to find the vehicle (which could even be triggered to start pinging when the vehicle loses power, if that’s another concern).
I can see how that could work, but I think it may add a lot of inefficiency to the normal operation of the AUV.
My current setup has one internal oil bladder, one external oil bladder, and a pump between the two. I had thought about putting the internal bladder inside a hydraulic accumulator, so that as the bladder fills it builds external air pressure. That air pressure would act like a mechanical bias or spring, pushing the oil back out and back-feeding the pump.
The issue I see with that approach is that every time the vehicle wants to dive, it would be fighting against that spring force from the compressed air. To create enough force to overcome the resistance of the gear pump and external water pressure at depth, I think the accumulator pressure would need to be fairly high, which could make the normal dive cycle less efficient.
There may be some combination of components that makes this work well, but I’m not sure what that would look like yet. It’s probably worth exploring further though.
Thanks for the suggestion!
Thanks, Eliot.
I’ve actually been looking into something like a drop weight attached with this magnesium ribbon. If I clamp the ribbon with stainless screws, the galvanic corrosion should eventually dissolve the magnesium and release the weight. Some testing would be needed to dial in the timing and reliability.
I used to work at a factory making sonobuoys, and they used a similar concept with a magnesium plug and nickel housing. After roughly 24 hours in the ocean, the plug would dissolve and the buoy would sink most of the time.
I also like the idea of using a beacon. I’m using a lot of drone parts, and buzzers are the standard way to find your lost drone, so it makes sense to think about an underwater equivalent.
Would it be possible to determine the direction of an underwater beacon without specialized equipment, or would that require something like a hydrophone array?
Fair enough. I imagine that’s quite reliant on the salinity of the water you’re operating in.
Another option to consider is
e.g. using 3D-printed PVA filament as structural elements to hold the weight.
The biggest challenge with chemical processes is likely the variability of release time, particularly since it may be difficult to control exposure to water flow/agitation. That said, assuming the emergency recovery time can be “set” to a fair amount longer than the intended dive time, that’s perhaps not a huge concern if the range of timings is well-established beforehand.
There’s generally a tradeoff between sensors, time, and spatial resolution/knowledge:
Depending on whether you expect a vehicle failure to result in a stationary vehicle or one that’s drifting with a strong current, the speed with which you can update directional knowledge may be low priority or critical.
Then of course there are variations of the signal, vs the reliability you expect to receive it with. @AndrewMI’s suggestion of a sophisticated signal that communicates exact position can allow recovery with very few received transmissions, but is then reliant on a high enough signal to noise ratio over the period the signal is being received that that position is actually decodable/interpretable by the receiver. It’s also reliant on both the beacon and the recoverer actually knowing where they are, which may work fine, but could also cause problems if one of those positions aren’t well-aligned.
Thanks for all the info Eliot,
I like the idea of using PVA because it’ll also work in the freshwater lakes that I’m currently testing in, and I already have some. The hard part of testing underwater robots that don’t have a tether is that when they fail, you typically don’t get them back. so having a recovery system is more of a cost saving measure, and we can remove it once the reliability have been proven.
as for locating a sounds underwater, this sounds a bit more involved than I’d like it to be, and a bit dependent on having a diver/ROV to go get it once you’ve found it. I’ll stick to a drop weight, or detaching a floating spool with a recovery line. I’d still need a boat, but this might be a bit more manageable at my scale.
Thanks again, I’ll keep you all posted with the results.
If you are in a marine environment and have enough spare wet volume, you might consider compressed salt and ballast. You can buy it cheap on ebay as cattle salt licks and the blocks are machinable like wood to fit any compartment. During a dive, you wait until enough salt has dissolved to eliminate the negative buoyancy for an emergency ascent. You control the salt’s dissolution rate, either by restricting the salt housing’s inlet and outlet port sizes (outlet facing down as the saltier water is denser), or by soaking the compressed salt in vegetable oil which slows, but does not stop, the dissolution. Compressed salt is not a dense material, so you need a relatively large volume of it, but then you are also not polluting the sea floor with your dropped ballast. I had originally worried that the saltier water would repel sea life, but I’ve seen no evidence of that. It has been a very reliable technique and has saved my drop camera when it lost all power.
Very cool idea!
Have you seen
This is a great idea! i love the simplicity.
I don’t know if this will work for my application, but I love the out of the box thinking. I’m trying to measure conductivity as well, so the salt would likely influence that, but it would be helpful for testing in the short term as long as the main compartment doesn’t flood.
This is also really cool. It’s similar to an old prototype I’ve built. thanks for sharing.
