Robotics

[peter grimes]

The last project I shared here really got me interested in microcontrollers and programming.

I was sharing a lot of my trials and tribulations with my brother and cousin, both of whom do some sort of computer voodoo stuff. They liked the idea that it was a little bit about physical computing - having an effect on something other than a computer screen.

So we've decided to undertake a group project:

We want to make an underwater ROV, similar to Randall Monroe's:
http://blog.xkcd.com/2010/11/05/submarines/

We're still working out the parameters of the project, which has me worried, since we'd need to have something in the water for July. That's the only time this summer that all three of us will be at our grandfather's lake house

I was just wondering if anybody here as done anything similar? It would be interesting to share project notes and such.

 

[Perfection]

I'm an electrical engineer who has worked on several projects with microcontrollers in a professional capacity. Feel free to ask me any questions you have, and I'll see if I can give you some pointers.

A great resource to get started on robotics is http://www.sparkfun.com



I should note that if you want to get this done by July. You probably need to be ordering all the crap right now.

I'd just buy the project kit: http://nventivity.com/ROVIAB.html

Once you get that, then you can start modifying to suit your fancy.

 

[peter grimes]

Yeah, I've been trying to pare down the parts list so that I can get the order in before tomorrow.

I've also spent about 15 hours poring over the stuff on SocietyOfRobots.com It's a rich resource. Their forums are deep as well.

 

[Perfection]

So if you don't mind me asking what's your basic plan?

 

[peter grimes]

Well, as with most projects that are undertaken on a whim with ill-defined goals, the basic plan is still in flux.

-There will be a tether.
-There will be a video feed through the tether to the surface.
-There will be some directional control and perhaps propulsion on the craft
-Operating depth is 100'
-Dive time will likely be less than 2 hours

Beyond this we're still uncertain. For example, we're not sure if neutral buoyancy is necessary, since we'll have a tether. But if we use the tether for altitude control, then our other axes of movement are further impaired.

Propellers attached to through-hull shafts connected to internal DC motors?
or submersible pumps mounted on the frame of the craft?

There are pros and cons to each approach:
PUMPS: On the PRO side, pumps require an easily sealed hull-hole for wires only. They are ready made, off the shelf, plug and play. Simply send power to a unit and it goes, moving the craft accordingly. However, we'd likely require 5 pumps, which adds up to around $100. The only way to reduce the number of pumps is to articulate the nozzle, or rotate the pump body. Both these approaches require an additional servo or motor in the hulll, which adds to the complexity.

MOTORS: On the PRO side, we'd only need 3, assuming reversibility. Enclosed in the hull, we can use cheapo motors which keeps costs down. However, this requires 2 or 3 shafts that pass through the hull. Each breach of the hull presents sealing problems. And if the motors are sealed inside the hull there may be overheating issues.

Right now these are the major things we're trying to work through. As each day passes I'm thinking more and more seriously about just buying that kit - it takes some of the fun out of the 'figuring it out for ourselves' part, but at least we'll be assured of having something that works to drop in the water!

 

[Perfection]

Everything costs more and takes longer then you like. Welcome to engineering (or life, for that matter)!

I would be very hesitant on trusting my own sealing. Sealing is tricky with moving parts, especially at significant depths.

One thing that's cool about the kit is it gives instructions on how to make it, so you could use those ideas to make a different one in the future or modify the current design to do different things (I mean, you have to have a manipulator arm, right? )

 

[peter grimes]

I'm very accustomed to engineering in my day job, bit not at depth, pressure, and with my own money

I am leaning strongly towards the kit, but the funds will be tougher to sneak past my resident bursar. It's much easier to get a pass on 5 $30 items over a few weeks than a one-off of the dame amount.

I can source off the shelf shaft seals that can more than handle the loads, but they are sized for much larger shaft diameters. So then I consider a shaft coupler that would mate the motor shaft (3mm or so), to the drive shaft, but I don't have a metal lathe or milling machine, so there's more complication.

My other concern with internal motors is heat buildup. I have experience with venting and cooling in the past, but not in a sealed tube underwater. The ambient temperature the ROV will operate in will be in the range of 45 to 70f (~7 - 19C). But I'm not at all prepared to estimate how much heat to sink over the required surface area at the required rate.

I'm not interested in reinventing the wheel, thermodynamics, or scada!

 

[Perfection]

I'd suspect that being in water would be pretty beneficial from the heat dissipation standpoint.

Dunno about much else.