485Views101Replies

Author Options:

iRobot Scholarship idea : Pool Cleaner Answered

I've seen mods for the roombas before, where the robots are water sealed, and then the wheels are turned into propellers. My idea is that I could water proof the robot in plexi glass, or something, and then a net on it, to collect all the leaves, dead bugs, and other stuff that collects on the surface of a pool. normal pool cleaners only vacuum the floor of a pool, and it takes days for floating leaves and other grime to sink to the bottom. With a robot like this, I could just set it in the pool, and let it clean everything up. What's nice is, when company is coming over, it's embarrassing to have a dirty pool...this modification would allow you to quickly clean up your pool-even on short notice. Update (for some clarification): What I would want to do, is make a case for it, so that you can simply remove the wheels, and connect the propellers/steering device to where the wheels were, then just place the roomba in the watertight case, and it would be able to swim around and clean! Also, i'm working on some sketches of how it would work. I'll post those when they are done.

Comments

The forums are retiring in 2021 and are now closed for new topics and comments.
0
ewilhelm
ewilhelm

13 years ago

This might be a little late, but one of the tricks I know for building underwater robots is to pressurize the inside of the robot; so, if there are any leaks, air leaks out rather than water in.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

That's a really neat idea, and it makes a huge amount of sense. Unfortunately, I am past that point...we're ready to add the bushings, axles, bearings, etc. today. The entire box is waterproof, except for the lid, which will still have weather stripping. Thanks for the idea though, if I ever one another one, i'll definitly try that.

0
Weissensteinburg
Weissensteinburg

13 years ago

Ok guys, so my latest design works like this:

_
| |
X---|---o|o----|---X
||

If that stays how it should...then the X's are water wheels, the -'s are axels, the o's are bearings. The robot will be places on top, so each wheel will power one propeller. The bearings will be used to keep the axels in place. Bushings will help keep the axles water tight...but I need help finding rubber bushings...does anyone know where to look? I'm using 1/2" pipe as the axle, which needs to fit through the bushing. Also, i'm seeing measurments on bearings such as 5x11..what does that mean?

Thanks =]

0
NachoMahma
NachoMahma

Reply 13 years ago

> I need help finding rubber bushings . O-rings epoxied to the case? Would need to be a relatively loose fit or you'll lose too much power from the drag. . cooblades "grease seal" might work. . > Also, i'm seeing measurments on bearings such as 5x11..what does that mean? . I used to know that, but can't recall it to save my life. It's hell, getting old. Check out the web site of a bearing manufacturer (eg, Timken) and see if they have a page that explains it.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

apparently the drawing didn't hold up.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

I've seen that, the thing is though, their cleaners don't clean the entire surface, like this will.

0
W_world
W_world

13 years ago

Really neat idea. Have you considered making it submersible? you could use a "swim-bladder" type system to ballast it. I think it would be cool if it could seek out debris in a 3D setting; you could use a camera to find the debris against the solid color of the pool walls.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

I think that if I were to step up to that level, it would make it infeasible for me to do in less than two months..i'd need: Totally water proof casing cameras and face recognition like software motors..and a way to control them in relation to what the camera sees automatic water pumps etc.. Good idea, It would just take too long, and would be very expensive.

0
loubard
loubard

13 years ago

Isn't this similar to the Scooba product iRobot already available? Great CAD btw.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

No, the Scooba mops tile floors. iRobot does have a pool cleaning robot, but not one that cleans the entire surface of the pool. Thanks..I had to sit around for a while (long story) yesterday and the day before, so I spent a while working in sketchup.

0
Weissensteinburg
Weissensteinburg

13 years ago

I've made a semi decent drawing of it on sketchup...now what would be the best method to export it?

0
NachoMahma
NachoMahma

Reply 13 years ago

. That looks great - both the design and the drawing. I really like the funnel/scoop design. I think the prop needs to be closer to the centerline or it will go in circles - some bias on the rudder may handle it.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

Thanks! I'll be able to fiddle with the alignment when/if I make it.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

..figured it out. The video shows an animation of the it from different views.

0
NachoMahma
NachoMahma

13 years ago

> How big/how many blades do you think a propeller for something like this would need? . If noone else has an answer, I have a friend that has a speedboat (and seems to know quite a bit about prop design) that I can ask for some tips. Not sure how well what he knows would scale down. I'm betting that the speed you will be turning the prop and the amount of torque/HP available will be the major factors.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

It would be great if you could ask him. I've been trying to guesstimate (can't believe i just used that word) how fast and how strong that robot will move. Do you know where I can find those kinds of specs about the create?

0
NachoMahma
NachoMahma

Reply 13 years ago

. I talked to my friend and (after much discussion) he recommends getting a prop with ~18" pitch and a diameter of ~4". This will probably be too much prop, but he thought you should be able to chuck the prop to a drill and use a file to make the prop smaller and smaller until you get the speed you want - if it looks too ugly get a new prop of that size. He said you may want to start with a 3" prop, too many variables to be sure.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

Thanks. I called our local hobby shop (like I told you about before) and they do have a selection of props. Also, do you think there's a way to water proof the bearing I posted a link to?

0
NachoMahma
NachoMahma

Reply 13 years ago

> do you think there's a way to water proof the bearing ... . Not that I know of. Most of the brgs I've seen that needed to work underwater either used a pressurized lube system or were designed to use water as the lube. I'd try it and see what happens - it'll probably last long enough to get your project built and a video shot. As cheap as they are, even a poor student can afford to ruin one or two. Silicone spray may help, but it's as expensive as the bearings. . . That's great having the hobby store nearby. I'd try browsing the aisles for ideas and inspiration. They probably have all sorts of gadgets, widgets, and gizmos that will come in handy.

0
NachoMahma
NachoMahma

Reply 13 years ago

. Have you tried seeing how the R/C boaters pierce the hull? Remember that my background puts me at a disadvantage at the scale you are using (when you're using a 100HP motor, frictional losses from a bearing shield just don't matter). This kind of thing is what keeps boaters up at night experimenting and I bet they have much better ideas than I.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

I think that I might just put the bearing inside, and then have a tight hold drilled in the plastic, and use petroleum jelly to lube it/water seal it.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

I don't know if i'd call it nearby...but i think i could annoy the rents into getting me there...

0
NachoMahma
NachoMahma

Reply 13 years ago

. No idea on the specs of a Create. If you can't find the specs on the web site, try contacting someone at the company that might know.
.
. I'll talk to my friend and see about getting y'all together. I do remember him mentioning a web site called Scream and Fly. Just checked and sure enough, http://www.screamandfly.com/ is a site for "hot rod" boaters - try posting to one of the forums. They may even be able to tell you how to guesstimate. LOL

0
NachoMahma
NachoMahma

Reply 13 years ago

. Well... I went to the iRobot site, but everytime I'd get close to what I thought would be helpful I'd get a Flash not installed msg. The info seems to be on the web site, but I can't tell. If not, it appears that they have a few good ppl in the forums - maybe one of them knows (or can give you some ballpark figures). Wouldn't be surprised to discover one (or more) of the engineers hangs out in the forums.

0
cooblades
cooblades

13 years ago

Oh btw just to warn you water robots will always veer off course. Also the propellers will create a counter torque which will make your robot veer into the wrong direction so when you mount them, make sure they are not straight but at an angle to correct this. My guess if the propellers are anything like RC planes it should face the bottom right. Sensors will be a must if you want anything that does more than get stuck at the edges. Especially for a skimmer. The level of difficultly programming a water based robot goes from Crawler, Skimmer and Hovering being the hardest. Good luck!

0
TheMadScientist
TheMadScientist

Reply 13 years ago

oh really? i always found that skimmer was the hardest, followed by hovering then crawler... always found that all of the challenges of actually being partially submerged in water far outweighed the challenges of keeping the hovercraft above water...

0
cooblades
cooblades

Reply 13 years ago

I meant hovering underwater in the definition "wavering between conflicting positions". In the video I posted the first robot is a hovering robot the second was a skimmer and the third was a crawler.

0
TheMadScientist
TheMadScientist

Reply 13 years ago

ah, sorry, I mistook hovering as my favorite type of land robot!

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

Yeah, this will definitly be a skimmer

0
NachoMahma
NachoMahma

13 years ago

. Have you done any prototyping of a mag coupling yet? The more I think about it, the more problems I see for a DIYer on a tight budget and short schedule.
. It will need some type of thrust bearings to keep the magnets apart. The bearing mounts will need to be sturdy, which may present space or aesthetic problems. For your use, plastic/plastic or plastic/steel may work, which would make them easier to fabricate, ie, no "real" bearings, just pick low-friction materials.
. Alignment may present problems.
. The distance between the magnets (hull + 2 airspaces) may require larger and/or more expensive magnets that don't fit your space or budget.
. These may not be a problem for you, but I think you need to do some experimenting before you get too far with your design.
.
. While I think the Cool Factor of a magnetic coupling is orders of magnitude greater than cooblade's grease seal, the grease seal should be much more practical. Very easy to seal the hole in the hull and it shouldn't be too hard to find a "grease" that won't leave a slick.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

I haven't..my plan was to find some, like you said, low-friction materials..the magnets would keep each other in place, and separated by the Plexiglas. The only distance between them would be about a fourth of an inch of plastic.

0
NachoMahma
NachoMahma

Reply 13 years ago

. But they will be strongly attracted to each other. If you don't force them apart, they will slam into the plexiglas and not rotate. . There will have to be an air gap between the hull and magnets (or they can't turn, see above) with a little extra thrown in for manufacturing tolerances. A pretty big gap considering the scale. If you were using the 4-6" magnets I'm used to, the extra distance wouldn't be a big deal.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

4-6" diameter? They're going to be smaller than that...

0
NachoMahma
NachoMahma

Reply 13 years ago

. Exactly. With the magnets I'm used to, 1/2" gaps (magnet coating, barrier, etc) were the norm, but this was with much stronger magnets.
. I'm not trying to talk you out of it - I think you can do it - just that I'd run some proof-of-concept tests before trying to design around it. It's shape/config may change drasticly before you nail it down. Maybe hold a couple of good disc magnets about 1/4" apart and see if it feels like enough torque can be transmitted.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

I've got some plexiglas from my old 8th grade science fair experiment, somewhere. Could I just use bar magnets for a test?

0
NachoMahma
NachoMahma

Reply 13 years ago

. I'd try to make the test as close as possible to the design in your head. . Not sure how bar and disc magnets compare/constrast (other than the shape). I'm just assuming that since the ones I've seen use discs, then discs are better.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

I just don't have any disk magnets...If this way doesn't work, I don't think it would be too hard to just drill a hole where the coupling would be, extend and plug the pipe, then seal the edges with grease and lube. I could then put the prop on the pipe. Come to think of it, I think that might work better than a coupling would.

0
NachoMahma
NachoMahma

Reply 13 years ago

. LOL So many options, so little time. . If your plexiglas is in short supply, use plywood (or something similar) to make your "first drafts." If an idea doesn't work out, you haven't ruined the plexiglas.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

Well i'd be buying more if i actually get the create. I think I would use card board and hot glue for my proof of concept and design finishing..*really hopes to win*

0
NachoMahma
NachoMahma

Reply 13 years ago

> I think I would use card board and hot glue...
. Great idea! That should float long enough for testing and the price is right. A good coat of paint/varnish/wax/etc and it might float for days (and would probably make it more rigid, ie, more like plexiglas, also).
.
> *really hopes to win*
. Your's is the most practical idea I've seen.
. The thiefbot comes close. Very high Cool Factor (especially with the right appearance and sound effects), but finding/acquiring random objects may be difficult to implement even if he's not picky about what gets picked up.

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

Even so..they're giving out 15...so... =D

0
cooblades
cooblades

13 years ago

I've done marine robots before. If you are worried about water proofing, try a motor shaft in a long narrow tube with some grease in the middle. That should prevent a lot of water from backing into the motor. For your other connections try using Marine Goop, it is used to seal boats connections and its pretty affordable. Magnetic coupling is going to cost you some energy due to friction if you plan to use it so be careful what you choose. For floatation you can use home insulation foam boards (approx 1-2" thick). They are sold in large sheets and very cheap at Home Depot. A hot wire cutter can help you cut that to some pretty nice shapes.

0
NachoMahma
NachoMahma

Reply 13 years ago

. I like your grease seal, but would it have less drag than a mag coupling? The only drag to a mag coupling is the bearings. Grease can be pretty sticky. . Would the grease stay in place well enough not to leave an oil slick on the pool? If it will, your idea would probably be easier to implement.

0
cooblades
cooblades

Reply 13 years ago

Primarily we add a small amount grease or oil as a lubricant so it reduces friction. Whether or not it is less friction than a magnetic coupling depends on design features out of my control so I wont give you a definite answer on which is more frictionless. Regarding contamination I wouldn't worry too much about it. Oil and greases don't dissolve or mix well with water. Plus you're only using a drop and there are many safe and clear greases you can use. In the past I have used petroleum jelly from the store cause it was easy and cheap. The greased axle design is really simple and cheap. Go to a R/C hobby store and you can buy the tube, axles and grease there. Its a pretty basic technique in RC boating.

0
NachoMahma
NachoMahma

Reply 13 years ago

. OIC. You weren't comparing the two, just something to be aware of. . Since this project is for a swimming pool, I'd think contamination would be pretty important - who wants to swim in an oil slick? But, as you point out, it should be easy to manage. . It'll be interesting to see what he does. Wouldn't be surprised if he finds another (better) way to do it. ;)

0
Weissensteinburg
Weissensteinburg

Reply 13 years ago

The home insulation being the fiberglass? I'd like to stay away from any of that "try not to breathe it in" stuff. I think that if I just use a rubber tube around the edge (for stability), combined with the boat like flotation, it will be enough, what do you think?

0
cooblades
cooblades

Reply 13 years ago

Sounds like it would be pretty good. My family owns a fish pond and we grow lily's. To ring in the lily's so they look nice we cut a rubber tube and use a extending joint to join the two ends for an airtight seal. The trapped air is enough for it to float.

However I think you are confused with what I was talking about earlier. Here is a link to clear things up
http://www.owenscorning.com/around/insulation/products/insulpink.asp
These boards are NOT fiber glass. They are Polystyrene and pretty safe to use for hobby work such as this. Here is a link of someone who used this to make a game icon. You can carve very nice models using a Hot Wire Knife

http://oz.games-workshop.com/games/40k/daemonhunters/terrain/icon/default.htm

0
NachoMahma
NachoMahma

13 years ago

. By using a magnetic coupling, you won't have to breach the integrity of the waterproof case.