OpenROV Gripper, Grabber, Manipulator

manipulators-tools

#1

Hello All,

My name is Patrick and I am interning at OpenROV this summer. One of the projects that I have been given is to design a gripper powered by a standard hobby servo for the OpenROV. My question to you all is what functionality would you like in a gripper for your ROV?

My current thoughts on this challenge goes as follows. When you are driving your robot you are never entirely sure what you are going to see and what you are going to want to pick up or want to probe. Because of this I see a modular design that combines a “scooper” and different types of “pinchers” as the best solution. Companies such as VideoRay and Seabotics already use this technique with interchangeable heads to their actuator (in our case a waterproof servo would be used to power the device)

Given this information, I have been designing a two finger gripper based off of the Yale Open Hand Project (found here) . This design works by having multiple pieces of rigid plastic connected with bendable joints inbetween. Cables run up the length of the pincher which are then pulled by a servo at the base. The idea behind this design is that it is adaptable to grabbing objects of different sizes and shapes with ease. See this video to see the gripper in action.

We would also plan to have some type of attachment slide on the gripper to enable scooping.

Thoughts??

What Our Current Gripper is Based off of

A More Traditional Gripper Design


#2

Patrick,

Based on everything I have seen with manipulators and grippers, I believe a “three-finger” gripper (with fingers at 120 degrees to each other) such as the Seabotix standard or the VideoRay Trident gripper provides the most utility. With three fingers, the object is less likely to fall off to one side. Two fingers work great in the lab and for demonstration purposes, but in field applications, objects have a tendency to fall out.

Now with that being said, we all have different requirements for grippers so if this is made as a modular product, two fingers might be the best way to start, but I can also see applications for parallel jaws, three-fingers, and your mentioned scooper/sediment sampler.

Kevin


#3

Since we are dealing with an accessory for OpenROV the power and size will be limited. Personally I would like to see a gripper that has a “wrist” joint allowing the “fingers” to be rotated through at least 180 degrees. Multiple hand attachments would reduce the need to fabricate an entire system to handle the various tasks one would want to accomplish such as gripping, tagging, digging, etc…


#4

@Patrick2 - Interesting video, looks to have a lot of potential. Another idea I gave some thought to was “soft robotics” where hydraulic pressure (via air, water or oil) is used to manipulate the grabber. There are several DIY examples on the web, here’s a video which illustrates the concept: https://www.youtube.com/watch?v=csFR52Z3T0I There would be several issues that would have to be overcome in order to use this underwater, but one big advantage would be that you wouldn’t need to deal with waterproofing servos or other electronics.


#5

I like the Festo fin gripper design very much, which can also adapt its shape to the object:


#6

@kevin_K The Idea behind using a waterproof servo as the driver is that it creates a modular system which ideally different gripper heads can be attached and detached creating a somewhat modular system. Fortunately, at OpenROV we are currently developing a waterproof servo, so the user would not have to deal with issues of waterproofing servos themselves. I’ll also throw together a three jaw gripper design once I get the two jaw working

@stretch Given that information, a “soft robotics” solution seems worthwhile to look into, especially for picking up objects that a normal servo would not normally be powerful enough to pick up. I’ll try to make one soon in order to test it.

@Ronald_Peters In the future having a multi jointed arm is defiantly something that we would like to start exploring. However to start with we are going to build just a single servo powered gripper, that although it will not be ideal for all solutions, will be a step in the right direction.

@Stefan this type of adaptable gripper design is exactly what we are aiming for. That is the main strength of a tendon-joint based gripper it should adapt around different shaped items. However, this Festo design is quite simpler and well designed. The challenge is figuring out how to mass produce it, ideally with a laser cutter.

I’ll keep everybody posted as I go. My next step is to build a proof of concept of a two finger grabber tendon grabber just to see if my concept will work. After that I will start building a 3 prong gripper


#7

We built a two finger tendon based grabber today and we quickly learned that a servo will not be powerful enough to enable it to grip. It is based of the design below.

                          The Original Design 

The version built today

The end verdict is that a servo will not be powerful enough to run this. We are instead going to focus on a more direct drive simpler system


Further thruster engine for manipulator/grabber
#8

What if we used a sort of corkscrew and gears, like in this picture. We could use a stepper motor to drive the screw. The motor controller could be potted or encased in a dry box and the connections could be waterproof. Basically we would use a motor & controller rather than a servo.


#9

Although electro-mechanically based grabbers are the norm in mini-ROV’s, a hydraulics based grabber might have some potential and would eliminate some potential issues. By using hydraulics (using water or mineral oil) you wouldn’t have to worry about sealing electrical motors, penetrating wires or rotating shafts.

Here’s a demo video of a children’s kit that illustrates the concept:


#10

@Ronald_Peters I am a big fan of the geared-drive gripper as well because it has a lot of holding power and stays locked. It’s fairly robust and pretty easy to manufacture. Here’s the one from Makeblock that I liked:

However as you mentioned, the motor is a problem. I would argue that we could use a stock OpenROV brushless motor (I would use an M100 for seawater), but the issue is, how would you limit travel so you don’t break the mechanics? Without limit switches, you run the risk of snapping parts with over-travel. You could always manually control it, (I’ve tried it with the camera mount control), but you just have to be careful.

@Stretch The hydraulic approach solves A LOT of problems, and is actually what I would prefer for a small/medium size ROV (not mini size, where would you put the equipment?) I have yet to find a system that doesn’t cost hundreds of dollars though. On the other hand, we know these systems work, and they work well. So size and cost are the issues here.

Example (see page 14): Akvator-3D


#11

Both hydraulics and pneumatics afford you some awesome power and mobility but you then need to have either a valve bank and pressure chamber on the ROV itself which would be problematic given the OpenROV small size. You could also thicken your tether with pneumatic or hydraulic lines but that compromises the driving experience.

The goal of this current gripper design is more of to create a “starter gripper” which is relatively simple and can be included with the purchase of a waterproof servo. I am aiming less for power and more of for something to get the user excited about grippers and the ocean so that they will then want to explore other design options in the future.

Has anybody so far tried a pneumatic or hydraulic gripper on their OpenROV yet?


#12

Hydraulics, or pneumatics are interesting concepts. However, given the relatively small size of the OpenROV I am concerned that there might not be enough space to mount and carry the required pumps, hydraulic fluid and tanks, servo valves to make it work. I was thinking a “stepper” motor might work because it rotates in divides the 360 degree rotation up in to sectors thus allowing the operator to use a controller to set the motor to turn to any given position as desired. The motor would likely work like the openrov thruster motors in terms of being exposed to the water. Like the thrusters you would just need to be sure to keep up the maintenance on them to prevent corrosion. The entire assembly could be fabricated in a small enough size to fit on OpenROV.


#13

RE: Using a regular rov motor/snapping parts.
The motor, if it was geared correctly could just stop spinning, so long as the gears were stronger that the magnetic force of the motor. Or the gears could be made to slip. I have looked for brushless stepper motors, but they seem to be much more expensive than the regular brushless we use. Or we could use a solenoid to jack hammer the gears via a lever and make the gripper ‘normally open’ with springs or gravity.

RE: Hydraulics.

Servo, motor or Solenoid might be used to power a hydraulic gripper.

Ideally the layout would be more compact…

Not sure how the lock would be released, but you could use a solenoid! Why so obsessed? I have a potted solenoid ready to do something. Anything at all really. It just wants to be wanted.


#14

@Patrick2 Sorry for getting off topic. The focus here is using the waterproof servo OpenROV developed. I’ve seen it, I just don’t know the dimensions.

Ok, if you don’t need much power and want something easily manufactured, then just take an existing commercial design and modify it for your own servo. You guys have laser cutting abilities, so something like these should be easy to cut out. Here are my recommendations:

http://robosavvy.com/store/dagu-2dof-robot-arm-with-gripper-and-servos-21cm.html

Or



#15

That is a very impressive little manipulator!


#16

The most recent version of the gripper that we have been designing uses 3 fingers spaced 120 degrees apart as suggested by Kevin_K. The gripper is able to be cut out of acrylic on a laser cutter and the design uses the current m3 nuts and bolts that OpenROV already stocks.

Personally I think that this a step in the right direction, but I’m not quite happy with it yet. Let me know your thoughts on it!


#17

@Patrick2 Well I like it and that is exactly what I had in mind! It looks like it would be easy to integrate a push-pull servo linkage with a bit of an extension to the wrist.

Would you mind elaborating on what you don’t like?


#18

It looks pretty good and is probably sufficient to let you move forward. :smiley:


#19

Yes a servo should be easy to integrate into the design, and that will be the next item that I will need to CAD. Their is nothing technically wrong with this design, but my hope is to make it slightly more robust and more multifunctional. One thought that I have been having is embedding magnets into the end of the gripper to enable it to provide extra force to remain closed around objects, but I have not tested this yet.


#20

I love this design! :heart_eyes:

I’m not an engineer or much of a developer, but I am a marine biologist and I have a couple of thoughts on improving its gripping ability (I know this is a WIP):

  • serrations on the inside edge of the tips
  • grippy* nubs on the inside edges
  • perhaps even a grippy* liner along the round inside edges

* Material that is grippy like rubber, neoprene or similar. It should perhaps be a bit loose, so that when the gripper closes, the material when form to the object being collected.

I can’t speak for everyone who would be using a gripper for collecting samples, but my use would require a firm but delicate touch to keep a hold on the specimen. That is, keep a hold of it without breakage.