Let's Make Robots!

Yellow Fun


I built this model out of stainless steel scraps from my welding shop workplace several years ago. I had it working with a Radio control transmitter and ran it around San Francisco Bay in several spots before storing it for awhile.  Well a couple of months ago I finished chopping it shorter a little bit & rewelded it back together to make it lighter. Now I want to replace all the RC stuff with something a little more advanced so I could have it run autonomously in a pattern which brought me to this site. I am reading the storys on this site as a way of becoming somewhat knowledgeable of all the hardware used in all sorts of robotic systems.  I have a larger 30 ft sailboat I would use at times use in conjuction with this robot boat. Anyone with ideas on how to fit together a package electronics to make the brains of such a system? In the meantime I will keep studing the projects show on these pages.




Alameda CA

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
OllyR's picture

It certainly looks cool! strikes me that a sub is the perfect platform for semi autonomous motion. Chug around by RC on the surface to whatever spot, and then dive dive dive! As you would loose RC comms, you could develop a standard autonomous mission plan for a dive.

Diver vertically, using an ultrasonic sensor to find the bottom, go down, take a picture, log the depth or map the area in the immediate vicinity perhaps, then return to surface and check in!

You might want to build in an emergency blow recovery float though. How about a gas blown balloon which is kept deactivated by the battery. if the battery is allowed to run flat while underwater, a pin is released and it returns to surface on a big yellow balloon. Just so you dont lose all your hard work!

solarbay's picture






solarbay's picture


Here is a picture of it from yesterday, notice how its a bit shorter plus I have this trailer I built for it so I can trailer it behind my bicycle

to some hard to reach location. Its now 5 feet long as opposed to 6.5 feet long as it was in that in the water picture I first posted. I do not have plans to make it submerge because I would likely loose it that way. 



birdmun's picture

this forum post.


You can read both pages, but, the solid info is in that post. It would seem the waterproof ultrasonic sensors are about 5 times more expensive than the "cheapies" we use above water. :) The poster explains a bit more in a few later posts in that forum topic.

Another choice, because I couldn't quit searching. :)


The only thing is, I don't know how it would fair being submerged at any depth. You would have more learning to do with the second option too.

bdk6's picture

I'm with ossippee, gotta play "Yellow Submarine".  

This is really awesome.  That thing must be pretty heavy and a pain to tote around.  Five feet of steel?  Wow.  In Hawaii in the mid 90's I rode on a yellow tourist submarine.  That was cool.  I also agree with lumi.  We need more details. Pictures, data, etc.  This is too good for just one photo.

Anyway, if it's already wired for RC control you are halfway there.  You will need sensors of some sort to detect what's around it.  I'm not sure what is available for water/underwater use, but I'm sure there are some.  Since that is outside normal hobby robotics it is likely to be rather expensive.  You will probably want to know where you (the sub) are and GPS is probably the best way to go.  GPS radio signals don't penetrate water very far at all, so you will probably need a surface bouy for at least the antenna. (BTW, the higher the radio frequency the less it penetrates water.  Radio communication to/from subs is normally done at a few kiloHertz rather than MegaHertz or GigaHertz.  The US Navy has huge antenna farms for the huge, miles long, antennas needed connected to very high power transmitters.  Google submarine radio communication if you are interested.)  Once you have sensors of whatever type, you need something to control it all.  If you are new to this sort of thing, Arduino is probably the best/easiest way to get started.  There is a lot of Arduino knowledge here on LMR and people always willing to help, especially with a cool project like this.

Nice build and welcome to LMR.

Does GPS work under water?  I thought that you would not get a signal so would need the receiver above waterlevel.  Otherwise use an inertial navigation system.

lumi's picture

Hoho, that sub looks like 2 meters long. Hard to tell without a reference. 

Why not let us know more details combined with photos. I am (and I am sure all others as well) are greedy to see more of that boat coz we do not have much marine robots here. With more details from your side we could give you much more tips regarding your questions. So it's a win-win situation :-)

I had that idea of an autonomous boat as well but got stucked with the time in some issues with the communication vs price.

solarbay's picture

Hi Lumi,

Yes at the point that picture was taken it was a bit over 6 feet long and a bit of a pain to move it around from its storage to where I would want to launch it since it is after all made out of stainless steel. A few months ago I chopped it and now its only 5 ft long and now I am gunning for some new electronics.  After studing the issue I see that I might be able to use a quad copter auto pilot that I can buy off the shelf and modify it to my needs.  I will post a new picture here later today showing it as it appears now with its new shorter look with a white undercoat.

Passero's picture

I'm building my own little robot that moves around autonomously: http://letsmakerobots.com/node/39087

The hardware and electronics should be easy for you if you have build your boat yourself. 

The tricky part here is the brain. It will require advanced programming and mathematic skills. I am currently reading up on some robotics and probability books as these topics are really important. 

Your robot does not only need to be aware of his surrounding, he also need to take into account the errors from the sensor and how to interpret these.

Just google SLAM (Simultaneous localization and mapping) as this is what your robot requires. 

solarbay's picture

Hi Passero,

Thanks, this is just the kind of feedback I am looking for.  I am continuing to read many of the stories here to try & decide what to buy first.



Alameda CA