Deep ROV Implications

[Scott_W]

Just after a few people thoughts who have modified some of the OpenROV backbone for some deeper diving operations as I am on looking for a way forward

I have a project coming up where we are kicking around building a very customised ROV for doing some quite deep work (around the 800m mark) or alternatively getting our hands on an existing deeper water ROV unit (we have got access to 2 existing units one with quite a fair bit of cost involved and the other quite some time off till it becomes available)

I don’t think I would have too much issues with the pressure side elements but I am just looking into some of the technical / electronic hurdles that will have to be resolved

I see one of the main technical issues as data transmission (both length of tether and the amount of data) the tether (say round figures 1km/3280 feet [maybe 200m neutral buoyant and the rest a reasonable cable maybe AWG 22 with some sort of stain relief core])

The ROV’s main function is really just a camera platform but this should not be too much of an issue we are thinking 3 independent 4K cameras (I am assuming data is recorded on board) and just the standard style camera to pilot the ROV (ideally it would be nice to get the other camera data up to the surface even if it is just stills say every second or so to look at exposure) and additionally a scanning style sonar to find the wreck site when we drop “near it” would have to get back up to the surface

I see this in part may be got around by adding an Ethernet Desktop Switch similar to what @Darcy_Paulin did for this TetherCam Build, to get multiple data streams to the surface

TetherCam Build

@Scott_W It is a switch. D-link go-sw-5e. It is really small! I plug the raspberry pi, the baby and the homeplug into it. Then I have a router topside to assign ip addresses. It is powered by 5v so I run it from the openrov board. I potted extra wires out of my endcap and used them to attach ethernet and power to the raspberry pi. I have had zero problems with it.

And additionally thanks to @Walt_Holm ‘s great work on the tender teardown

Teardown of a HomePlug Adapter

So numerous people within the OpenROV project have mentioned to me that some flavor of Ethernet-over-Powerline (say, HomePlug AV or IEEE 1901) device represents the ultimate way to send data across the tether. They are fast (200 Mbit/sec for HomePlug AV and 500 Mbit/sec for IEEE1901), don't require an arbitrary cable impedance, and allow power to be sent over the tether simultaneously. Most importantly, they are inexpensive due to mass production- a pair of HomePlug adapters can be purchased fo…

It’s got me thinking about a similar teardown of a VDSL2 (Very high bit rate digital subscriber line 2) system

VDSL

Similar to

VDSL2 Ethernet Extender Kit Over Single-Pair Wire - 10/100Mbps - 1.5km

Kit - HyperXtender ER Ethernet Extender - NV-202EKIT

So the questions are
Does this sound reasonable? And is it likely to work?

Any other technical limitations that would have to be overcome? (I am assuming a larger motors 6 or 8 thruster design ROV) I don’t think ocean currents will be too much of an issue, but it is a lot of tether out (1000m by 2mm dia gives the equivalent cross sectional area of basically 2m2 of sail behind the ROV)

Scott

[Walt_Holm]

Hi Scott:

I’m guessing you’re on the right track using VDSL2 modems. Right now I’ve got a pair, clones of the Startech ones, but I haven’t gotten around to testing them yet as I’m too busy with Trident work. My intention is to put them into a BlueROV2 that I’m modifying with completely new electronics. I want to be able to run 500-600m tether lengths without using a repeater.

Note that VDSL2 modems require significantly more power than Homeplug AV2 gear - about 5-6 watts per side, if you believe the documentation. This shouldn’t be an issue if you’re making a big ROV, but it would be a concern if you’re trying to make a really small ROV that dives deep.

-Walt

[Scott_W]

Hi Walt

Thanks for the reply

Yes I have been tossing around both OpenROV and BlueROV2 electronics / software as a basis (not wedded to either currently just thinking the technical issues through currently - I’ve got a reasonable budget worked up)

I’m guessing it would be a bit bigger than a BlueROV2 with full Stainless machined pressure elements with a lot of potting and a lot of thread and seals

Ideally I would want the unit say sub 30kg’s (to still be sort of single person manhandleable) I still haven’t done the power calcs yet but say 6 or 8 T200 thrusters and multiple light systems based around the (say) Vero 29D- Gen 7 LED’s and I’m budgeting at running at least 4 of those for say 4hrs active bottom dive time plus powered accent and decent so power/ batteries is going to be quite considerable weight / size / cost (and logistics of shipping them on international flights)

Do you have any thoughts on thruster configuration? It will just be an inspection class ROV and I have been thinking maybe 6 vectored and then an additional 2 at the rear? For more push through current? But not sure if that plays into the strength of either OpenROV and BlueROV2 electronics

Scott

[Walt_Holm]

There are interesting configurations out there that use 8 thrusters, including vectored-thrust designs that put a thruster at every vertex of a cube. See, for instance, the Boxfish ROV. It might actually be a pretty good model of what you are trying to build.

The ROV electronics that I’m designing will support 8 thrusters through individual UARTs, for fine control of motor parameters. There will also be a couple more UARTs on it for grippers and the like.

-W

[Scott_W]

Hi Walt

Yes very much that style unit, the Boxfish ROV actually sort of started the project, as we were talking about the potential of this project whilst we were on the boat for the Java Sea Project (that caused all the Netherlands / Indonesian diplomatic issues) with the Sales and Marketing manager of Boxfish who was one of the divers on that trip

British second world war shipwrecks in Java Sea destroyed by illegal scavenging

They are keen to be involved but looking like still 12-18 months before there is a unit available and we are trying to get the project up prior to that timetable alternatively it’s a Woods Hole unit ($’s) or DIY

Scott

[Kevin_K]

Hi Scott,

Very interesting project and requirements. I’d recommend also posting in the BR forums as there are a few people who have gone really deep using towed camera sleds.

Regarding the tether, for those distances, have to thought about an ethernet-fiber optic-ethernet connection? There are a few duplex ones around, but not many high strength tether options.

And for the thuster configuration, I’ve seen an 8-thruster in action based on the “Heavy Lift” frame: http://ardusub.com/introduction/#supported-frames

It was very stable underwater and managed to compensate for the long lengths of tether we had payed out.

[Scott_W]

Hi Kevin

Yep if we can get it up (still chasing down some funding big 's WHOI unit small 's DIY) it should be a great trip with several interesting deep water wrecks

I had tossed around switching to fibre but made (now maybe wrong as I did not look into it) assumption that the system would be too expensive but given

I had assumed (sorry A$) say 200m neutral buoyant @ say $5/meter and 800m @ say $1/m

Given
http://au.rs-online.com/web/p/fibre-optic-cable/2908861/
So fibre optic @ say $1.2 /m

And given the hardware is sort of comparable cost

Ethernet Media Converter - Copper to 100Base-Fx Multimode Fiber | StarTech.com

or

Ethernet Media Converter - Copper to 1000Base-LH Single-mode Fiber, SC |...

The only real issue is the splicing (hopefully not needed in the field)

RE the 8 Thruster it is more to give “grunt” in current rather than lift so given the heavy lift configuration

Thrusters 7 and 8 would be rotated 90 degrees down to provide additional aft thrust

With a fore and aft vertical thruster configuration so you would be able to pitch the unit down to “drive” it down to the seabed (sort of like the Trident under speed)

Scott

[Kevin_K]

Ah ok, I see what you’re aiming for now with the frame configuration. That’s a good idea with the dedicated fwd/aft thrusters but still having vectored thrust when needed, that should get you going straight.

[Ion]

Hi all:
For deep works, the theter design can not be neglected. Where the tether is, whats are approximate shape and path are good things to be known in order to prevent entanglements.
There are a few works related to the design of underwater towing cables and tethers.
I recomend … Fundamental theory by Glauert, Landweber and Pode.
Further development by Michael C. Eames.

Dont know any link related to this topic. If you cannot find any, I can write an abstract of the involved maths.

Regards

[martinoliver_999]

I agree with lon. One thing worth considering is that the more negative buoyancy tether used, the heavier it is at the top end. Even if you are using it with a clump weight, the occasional float may be required to ease the strain at the top, without changing the purpose. If you are only using neutral buoyancy tether the problem doesn’t occur.

[Douglas_Butler]

From my experience inspecting ships by ROV, if you want to move fast to a target area then extra fore & aft thrusters are good. But if you want to be able to fight a current while staying in place and looking at something interesting, you want strong lateral thrusters instead. If you are trying to read a name on a ship the current will most likely be sideways parallel to the hull. Current is never going to be into or out of the metal.