Robotic Arm

manipulators-tools

#1

How about a robotic arm to grab things and bring them to surface. This could actually also be a good mission for the OpenROV. One could actually retrieve objects someone has lost, in a lake for example.


#2

If you are going to add a grabber you should also think about stereo cameras. Grabbing something without depth perception is tough.


#3

That's one of the main things we'd really like to see for the payload module! I expect that there may end up being a bunch of different approaches, and I'm excited to watch this develop.

I suppose the most basic add-on would just be a stationary hook.

In my experience flying other ROVs in popular areas, the sea floor is LITTERED with pocket knives, sun glasses, etc. I'll bet someone could make a killing retrieving that stuff.

Eric


#4

If you are collecting things, you will want buoyancy compensation. If you are already doing pressure compensation for the electronics housing, this is simple. See my reply in pressure resistance thread.

Hi all. I am a diver and a mechanical engineer. Fascinating project(s). I will likely start an ROV next year with my son. In the mean time I'd love to participate. When I do build, it will definitely be a collector.

Enz


#5

That's a great point, I didn't think about buoyancy compensation after collection.

Glad to have you in the community and looking forward to seeing your collector strategy - sounds great!


#6

I wonder if the vertical thruster could be replaced with an inflatable oil bladder or bellows for both vertical movement and buoyancy compensation?

SherpaDoug


#7

I think, even with buoyancy control, I would still want the vertical thruster for fine maneuverability. That is partly because I am envisioning a simple claw on a rail collector, rather than a fully articulated arm. So you would need to "fly" the claw to the object you want to pickup. Once you grab the object,you can then retract the claw back up the angled rail into the ROV and drop the object into a basket. This arrangement only requires two servos. I will create some Solidworks models soon for review.

One other thing. Once you add a gas bottle to the ROV it pretty quickly wants to be arranged differently than the current OPENROV design. Will that get me drummed out of the forum before I even start? or could we have a gas compensated/buoyancy compensated branch of the family?

Enz


#8

That would definitely not get you drummed out of the forums. :)

I'd love to see it. Can you sketch up what you had in mind? Sounds like a fun discussion.


#9

OK, after a Solidworks session, I have a very preliminary concept of the "Compensated Collector ROV". See the attachment please.

The arm has a lead screw down the middle of a c-channel rail to extend and retract the claw. I like the lead screw, since it naturally has a low gear ratio and hopefully won't need a gear head motor to drive it. The rail is fixed at some angle for simplicity, with clearance to drag objects in and drop them on the tank. Claw activation is still TBD, but I have some ideas.

The closed portion of the frame at the top is the buoyancy tank. Sizing is totally arbitrary and will need to be worked out. It will be filled by a solenoid on the low pressure line from the primary regulator. Dump valve needs some thought, but needs to fail closed to ensure ROV recovery.

The green disc is the dive regulator to compensate the electronics housing. Which will give a useful depth capability of something like 2000 ft with half the gas in the bottle available at the bottom. I think the tether would be pretty unwieldy by then anyhow.

I have not modeled electronics, batteries, vertical thruster, plumbing and much yet. As I said, very preliminary. More when I have time.

Enz

2111-ROVAssyPerspectiveonBottom.JPG (277 KB)

#10

That looks interesting. Consider two rods within the channel. One threaded rod to control extension and one square or keyed round rod to control the claw.

Also if the top is an open trough to provide flotation it will be unstable. When you try to lift something heavy the front will dip and the air will rush to the back. You need baffles or multiple bladders to control this.

Keep working on it.


#11

Interesting! Thanks for sketching it up. Having a visual cue is much easier to relate to.

What's the bottom cylinder? Collection?


#12

Thats the air cylinder.


#13

The keyed rod is one thought, though it might bind as you are trying to bring the claw back up with your item. Another small motor and lead screw might work better. Working on it.

The buoyancy chamber is only open in one small waer drain in the middle. Baffling, or multiple chambers might be necessary to stabilize this design. I wonder if we can make the craft tall enough so the change in Cg has a minimal effect? Kind of like a funky hot air balloon.

Another idea is some kind of landing gear or skids to support the ROV until the object is nearer the Cg. Then you just need to add buoyancy.

At any rate, thanks for the feedback. This is early days for this design idea.


#14

I have a preliminary design for a motor actuated claw itself attached. Two images, on of which is a cross section, show how it works. The threads on the lead screw are not modeled.

Another simpler idea is to have a ramp or something at each end of the arm that pries the spring loaded jaws open. So, open when extended, closing as you begin to retract, and open again when fully retracted over the basket (near the Cg). This would be significantly more difficult to maneuver for a capture tho.

Enz

2109-ROVClawAssyClosed.JPG (103 KB) 2110-ROVClawAssyxsection.JPG (101 KB)

#15

Maybe you could work a spare-air SCUBA tank (or something similar) into the design for buoyancy control and pressure compensation for electronics:

http://www.spareair.com/


#16

I don't think that the spare air tank has enough capacity, at least if you want real deep diving capability. This is about 10x the size of the spare air and is good for 3000 psi. The other reason I'm using it is that I have one, valve and all.


#17

The current ROV design might be feasible for retrieving smaller objects by using a fishing reel with a spring-actuated hook. The ROV would be used to engage the hook, then return to the surface unreeling fishing line. At the surface, the fishing line could be used to retrieve the object.

For heavier objects, the ROV could attach a hook with a ring and have the fishing line running through the ring and attached back to the ROV (run from the reel through the ring and back to the ROV). At the surface, the fishing line could be used to pull a heavier rope down through the ring to allow for retrieval. Obviously the fishing line and rope would have to be more than twice the distance to the object in length.

Just a thought. It might be simpler.


#18

I think this is the right idea. I met the team from the Rozalia Project (clean up trash in the ocean using their little Video Ray) and they use this technique for larger objects. They have the ROV drive around with a hook and attach it to things like tires and chairs.


#19

I wonder if magnets might work as well, if for somewhat limited applications.


#20

How about using a hook, a float and fishing line? Together they would have neutral buoyancy. When the item is hooked, the ROV would pull back to release the hook; this would at same time release the float. The ROV would remain unsnagged and naturally buoyant. The float would then head for the surface for retrieval.

The design should try to make the float visible and easy to retrieve, real in and reset. A solution for using it as a pilot line like Dennis suggests sounds like a good idea. The hook should also be secured in a way so it does not fall out before it’s ready, but can still be pulled out by the ROVs thrusters. Various hook designs could be considered.

The ROV could maybe carry a number of hook-and-float arrangements staggered in length or otherwise somehow out of each other’s way.

I've added sketches (original ROV model by dezbot), the fishing line is not shown.

2107-hookSmall.png (150 KB) 2108-hook2Small.png (136 KB)