WALL CLIMBING ROBOT... HELP!!!

[katherina]

hi everybody...
we need help   I'm developing a biped window climbing robot for my final year project. me and my partner have never develop a robot before this and we're lost. we plan to use PIC16F877 for our climbing robot and a DC motor with a motor driver IC. wwhat d u guys think of dat?
our lecturer insist us to use DC motor as other motors are much more expensive, besides that we only have a budget of $100 only.
Is the microcontroller that we chose is suitable? and is motor driver IC better than H-bridge?
By the way, at the moment we have 3 inputs for the robot; swith1,switch2 and a potentiometer.  4 outputs;4 motors.
the window climbing robot is using 2 suction cups to hold on to the window and motor to rotate it and to stick the suction cup to the window.
is our idea ok?
please advice.... we're still in our proposal stage now...
please help us.... thank you...

[ed1380]

wall climbing as in climbing up a wall, or rolling over a bump(they've been called wall climbing)
what kind of wall will it be? plaster? plywood? concrete? etc?

[katherina]

sorry.. wrong info... WINDOW climbing robot. the surface is WINDOW only.... tq

[katherina]

just climbing up a smooth vertical window surface....

[Admin]

does it need to be a biped?

ive seen wheeled robots, frame walking robots, as well as 4 legged robots all climb on glass . . .

[nanob0t]

There actually is a commercially available robot that climbs smooth surfaces in the exact same manner.  Here it is:

http://scientificsonline.com/product.asp_Q_pn_E_3036976

It's $10.  Maybe you can use cash out of your own wallet and take it apart and see how it works mechanically.  I actually owned one and turned it into other things.  It should be helpful creating your own.


The biggest factor you will need is getting suction cups working with motors.  The drive in that robot is quite interesting, and worth a look at. 


We have your idea, but how big are you expecting to build it???   I need an idea before I start throwing out my suggestions.   

[ed1380]

breemen has one.
http://www.personalrobotics.nl/mambo/index.php?option=com_content&task=view&id=94&Itemid=1

[Somchaya]

There actually is a commercially available robot that climbs smooth surfaces in the exact same manner.  Here it is:

http://scientificsonline.com/product.asp_Q_pn_E_3036976

It's $10.  Maybe you can use cash out of your own wallet and take it apart and see how it works mechanically.  I actually owned one and turned it into other things.  It should be helpful creating your own.

Ooh that thing looks pretty neat. How well does it work?

[nanob0t]

It works pretty well suprisingly.  If I could find it I would post the thing scraped, so you could get an idea of how it works.

I figure, the largest problem is getting the correct amount of pressure or the right suction cups.  If you get gianormous suction cups, sure it could hold up 100 pounds, but you're going to need a 50 pound motor to get it off *sarcasm*.   The suction cups on this robot are very small, I was suprised when it worked.  They are very shallow, not being very dome-like used on some applications.  I would say it would require math, but I cannot think of anything to use for it. 

Then again, suction cups are quite cheap.  You can probably get one for $ .10.  I would develop a system, then try out different ones after getting programming down.  You may want to jumble that order up though.  Hmmm, I'm tired today

[Admin]

If you get gianormous suction cups, sure it could hold up 100 pounds, but you're going to need a 50 pound motor to get it off *sarcasm*.

it has a lot to do with how you pull the suction cup off, too. peeling takes much less force, for example. its kinda like gecko hands . . .

[Brandon121233]

and is motor driver IC better than H-bridge?

Motor driver IC's are H-bridges, they can be either half, full or double H-bridge

[katherina]

I didnt know that motor driver IC are H-bridge... 
huhu.... i'm VERY CONFUSED now...
 
we've already develop the mechanical design using solid works and we've chosen the simplest biped design...
and we've to submit the proposal in 2 weeks time from now... 11/05/07
 huhu... we're short of time now to change anything by now...
By the way, we're from Malaysia 

[Admin]

I would recommend using servos as they are cheap and all the electronics are built in.

[Hal9000]

From my experience the commercial window climbing robots would have trouble carrying extra weight. Bear that in mind.

You can also get suction cups which can stick to glass at automotive stores I would think, since these are used for people carrying large panes of glass around.

[nanob0t]

it has a lot to do with how you pull the suction cup off, too. peeling takes much less force, for example. its kinda like gecko hands . . .

First, I'm pretty sure geckos have a biological design built into their hands that simulate tape.  They have thousands of tiny strands that, when mositened, stick to objects, this is how they can climb anything.  A type of tape is based off of the gecko. A piece the size of the palm of your hand is enough to support yourself in the air.  The only biological creature that uses suction cups is the octopus I believe, though there are probably others. 

Next, in order to peel off a suction cup, you need to have a lever of some kind pulling from the edge of the cup, this would need another motor or something in order to function.  If you were thinking in terms of a human biped walking and they can exert force by pulling with the upper body weight, sort of like getting stuck in mud and lunging yourself foreward to pull your back shoe out, that suction cup is still there and in the exact same medium.  The suction cups are an amorphous material, so they most likely will not change durastically pulling from different angles.


I believe the best way to approach it is like a plunger, like in cartoon shows.  An industrial suction cup that you may find at an automotive shop is used to support something hanging off of it.  I believe if you have something pumping, you'll get the suction you want and the ease of the project.  With a plunger, you can adhere to the surface, but the complete surface area of the cup is not touching the object.  There still is air within the cup, but the pressure is still aquired.  Then the lip can be lifted and the pressure equalized.


Then again, material can be accounted for.  I would just experiment with different suction cups.

[Admin]

Actually geckos use van der Waals forces
Its an intermolecular force that works at the micro/nano scale . . . it is also dry adhesion, meaning no water is used.

As for peeling . . . its sort of like the difference of pulling the plunger straight up, or by pushing the stick sideways. The latter makes removing the plunger much easier.

Geckos remove their feet like you remove tape. You peel it from an end, and not try to remove it all at once . . .

The suction cups are an amorphous material, so they most likely will not change durastically pulling from different angles.

But you did make a good point here. It depends on the shape, size, and flexibility of the material used to make the suction cup. Will probably require experimentation to find one that works best for your bot . . .

[ed1380]

any thought of using some valve to let air in and just get rid of the vacuum to unstick it

[Brandon121233]

This looks pretty neat http://www.newscientisttech.com/channel/tech/mg19025526.500.html, the pentagon seem very intrested in it, you could be the next Q for James Bond!

[JonHylands]

Here's another one, with a very novel foot/walking design:

http://www.technologyreview.com/Infotech/18602/

- Jon

[Admin]

Here's another one, with a very novel foot/walking design:

http://www.technologyreview.com/Infotech/18602/

ha! i know the guys that worked on that back at the Nanorobotics Lab at CMU . . . I learned sooo much from that lab . . . Metin Setti was my advisor, too.

[katherina]

thax a lot for the opnion, advices and recommendations...
we're using 2 motors for the moveent and 2 motors for the suction cup now...