ROV for search and recovery


#1

Any ideas on a ROV designed for the search and recovery of missing persons underwater in varying visibility. What I had in mind was a modular system where you could join up 2 or more rov’s with a pole that could adjusted to vary in length. All modules would be controlled by a master unit with 1 or more slave units that could be joined together to form a horizontal pole which could vary the spacing between the cameras depending on the visibility of the water .So say the visibility is poor and the largest distance you can stay from the bottom is 4 foot before you get a image of poor quality then with a 3 module set up you would be able to cover approx.( 3 x 4) 12 foot of a span while searching in poor visibility , now say the visibility is not so bad say 10 foot then you could extent the spacing between each cameras to 10 foot giving approx. 30 foot of coverage with 3 cameras all controlled by the one controller.
so if you were using the 3 module unit it could have 3 forward facing cameras and possibly a camera at 45 degrees at each end to get a bit more distance, these figures are approx and the idea is in the early stages but it may also be required to power the unit from the boat as extended searching would not be unusual when looking for the recovery of body’s, what is required is a ROV that can be adjusted out of the water to vary its length so as to get the maximum amount of search area for a given visibility and number of cameras and to cover a large area as quickly


Using a Humminbird Side Scan Sonar not to the bottom but to ahead of the ROV
#2

Have a look at what @Stretch is doing

Many objects (including bodies) have been found with side scan and what stretch and @craig_bates is doing with a simple sounder is just an extension of this and it will give significant better coverage than a vision based system (say 30m out on all sides at a best guess)


#3

Hi Joe,

I remember when the body search and recovery topic came up with the DIY side scan guys and the consensus is that if you really do need to look for a body, hire the professionals, don’t keep the family waiting.

Now with that being said, if you want to work on the research and development side of this, I agree with Scott that a very high resolution side scan system, usually around or above 800 kHz, is probably the best way to tackle this problem. There are a few towed systems, but the ROV mounted ones are few and far between. You might want to try attaching one of the recreational side scan units (Garmin, Lowrance, Humminbird) to an ROV and see where that gets you.


#4

Thanks for the replys so far and I am hoping for more help and ideas.
My interest is mostly in search and recovery applications so I am looking for the most efficient way of searching when speed of recovery is important .
There are many advantages to sonar but also many disadvantages when searching for bodies or anything lost underwater.
Some of the disadvantage of sonar are.
1 False positive signals signals that look like what your looking for but turn out to be something totally different.
2 Definition is still poor compared to a camera.
3 It can be very difficult to interpret the images you can go over the same object several times from slightly different angles or from a different height with a sidescan sonar and each time you can get a different image, some images can look totally different to what the object is.
4 When you get a image or number of images on sonar you have then have to get back to the exact location often in poor visibility to send a diver to check out the image, when you get a image with a ROV you can move closer and get a detailed picture from every angle and be pretty sure you are on the right track before you send a diver down.

What I had in mind is a modular system ,probably 3 slightly modified ROV’s that would connect together but would be controllable with one controller and would be ideally recharged from the surface with a buoy on top that would provide GPS data and remote charging that would give unlimited battery life or at least extended bottom time as long as there was power on the top side.This set up would consist of two or more ROV’s ,these would be almost standard ROV;s except they would have a input and a output port this would allow them to be configured for normal use or as a slave unit where they would take directions from a master unit which would have the tether to the surface while the slaves would have a shortened tether that would plug into the ROV designated as the master ,the slave units could at any time have the normal tether plugged in and they would be used as normal units.Each ROV would attach to a harness that could adjust the space between each ROV depending on the visibility from perhaps as little as 3 feet to perhaps more than 15 feet between each ROV, so for a 3 unit set up would be 30 ft long giving perhaps a sweep of 45 feet at the bottom, the software would be slight different than normal as would need to be able to left and right hand vertical thrusters.For me battery life of 3 hours is to short as if you after spending a lot of time getting ready for a day of searching to be limited to 3 hours of searching is just to short especially if you are involved in search and recovery and you can have other volunteers to take over the running of the ROV while you take a break.As you would be controlling perhaps 3 ROV’s and perhaps charging them by the tether you could have a slightly heaver tether that would allow for more current to carried by the wires down to the ROV.


#5

Joe,
You may want to take an avenue similar to what I am working on. I am using the humminbird 360 side scan-sonar mounted on a ROV to use as what I would call an on board radar. The data coming back to the surface is a very high resolution scan that is easy to interpret. The 360 is an IP based system that will merge quite nicely to the existing network that the ROV currently uses. I am in the process of building a ROV with both the 360 sonar, a camera pod that houses a hi-def camera that is also 360 degrees of rotation and 45 degree up and down plus the forward looking ROV navigation camera.


#6

Hi Joe
Don’t take anything the wrong way as always it’s open source your free to develop what you feel you need and your proposed system would make a great platform for structure from motion data capture


Given that, for the application that you are looking at I would still use a sound based system rather than visual (remember this is a ROV (not autonomous) with a camera on board so you can drive over to any targets and check them out visually as soon as it looks interesting)

The resolution (and range) of any sound based systems (side scan/360/forward looking) is all about the frequencies higher frequencies greater resolution less penetration/range
Frequency 100 kHz approx. 15cm resolution Actual Usable Range approx. Say 300m
Frequency 200 kHz approx. 8 cm resolution Actual Usable Range approx. Say 150 meters
Frequency 455 kHz approx. 4 cm resolution Actual Usable Range approx. Say 75 metes
Frequency 800 kHz approx. 2 cm resolution Actual Usable Range approx. Say 30 meters

If you are interested in getting a better understanding of how it interpret side scan data by far the leading books are from John Perry Fish and H. Arnold Carr and I couldn’t recommend them high enough (Sound Underwater Images and / or Sound Reflections http://www.instituteformarineacoustics.org/IMABooks.html )
Scott W