Author Topic: Do you think this is possible? (Read 5604 times)

jamesjim · « **on:** July 16, 2009, 05:25:48 AM »

Hi

I am thinking of making a robot to replace the human in this[url+http://www.youtube.com/watch?v=1oyWMusaDTI]hydrofoil windsurfer[/url]

I want to use an IMU to shift weight along the pitch and roll axis.

I haven't really done anything with robotics before and am looking for some insight into bidirectional motors.

I am just looking for honest opinions about if it could be possible?

SmAsH · « **Reply #1 on:** July 16, 2009, 05:42:06 AM »

when adding youtube links, just paste the plain old link... Hydrofoil Windsurfing
it can be done, but quite a bit of knowledge must first be gained...
im guessing what you want to be able to do is shift weights along an axis to keep the board stable and going?

Razor Concepts · « **Reply #2 on:** July 16, 2009, 06:31:55 AM »

It is possible but this is a very very advanced project... you will probably be spending at least half a year on this.

ArcMan · « **Reply #3 on:** July 16, 2009, 05:50:56 PM »

If you could replace the very skilled human in that video, you would be decades ahead of any advanced robotics institution. MIT would definitely want to talk to you

So - No. It's not possible.

If you're interested in robotics, start out with something simple and work your way up. There are lot's of great project ideas on this site.

jamesjim · « **Reply #4 on:** July 17, 2009, 02:05:10 AM »

Score so far

YES NO

2

jamesjim · « **Reply #5 on:** July 17, 2009, 02:06:39 AM »

Score so far

YES NO

2 1

thankyou and keep it coming

SmAsH · « **Reply #6 on:** July 17, 2009, 03:17:17 AM »

im with arcman, the amount of skill and knowledge for this project is way huge...
i would say you would be spending at least a year on this... mines probably a no...
not trying to offend you but i think this is a bit much for your first project.

guru · « **Reply #7 on:** July 17, 2009, 12:37:25 PM »

My best guess is 6-10 years. I'm with ArcMan. "If you could replace the very skilled human in that video, you would be decades ahead of any advanced robotics institution." Think of the honda robot, about as advanced as you get, and it would fall on its ass and short circuit. So not possible yet.

You might be able to replace it with some sort of non-human looking box like contraption with lots of accellerometers, wind, lidar and other sensors and be able to get it to surf. It would need to be a pretty fast sensor loop and actuators (motors), and fast processors. It wouldnt tack or gibe, you'd have to have one unit of each side.

Joker94 · « **Reply #8 on:** July 18, 2009, 05:27:33 AM »

i must say, even if there was the slightest chance you could make a bowx all the elctronics and sensors it would take would costa fortune, i man thousands upon thousand of bucks to get that going, $:-\$

but if you have the money go right ahead and prove every person on this forum thread wrong.But start of with something much,much,much simpile than that to build up suitabe knowlage.

goodluck

jamesjim · « **Reply #9 on:** July 20, 2009, 12:38:25 AM »

http://www.youtube.com/watch?v=g-65Q3BpwH4

just trying to paste a video

jamesjim · « **Reply #10 on:** July 20, 2009, 12:39:18 AM »

it worked !!!!

jamesjim · « **Reply #11 on:** July 20, 2009, 01:15:57 AM »

thanks to every one who has posted so far.

i will try to outline exactly what i want to do to clear up any misinterpretation.

i want to use this imu from diydrones http://www.diydrones.com/page/uav-devboard

the best visual representation of what it does is below
http://www.youtube.com/watch?v=g-65Q3BpwH4

it will give me accurate pitch and roll angles in relation to the earths Z axis (straight down)

If we imagine the X axis running through the bow to stern

i want to slide weights along the X axis to compensate for pitch rotation

and i want to slide weights along the Y axis to compensate for roll rotation

so each of the two axis will have a machine like this running along it
http://www.youtube.com/watch?v=mqLI1d6R-Kc

so the robot would have nothing to do with the sailing of the rig

its sole purpose would be to keep pitch and roll horizontal.

so i would first be looking to build a single axis (out of the water) that could be controlled by the imu and the sliding of weights.

is the motor in the above video of the inverted pendulum called a stepper motor? how do other electric motors compare to it insofar as speed and precision?

Thankyou for your help.

SmAsH · « **Reply #12 on:** July 20, 2009, 01:30:30 AM »

i think i understand what you want...
the best approach for this that i can see is to have a threaded rod going along both axis and have weights that could move up and down the rods.

most other motors do not compare to steppers while they are on their own, steppers have much better control than any other motor. although, steppers require more to control them. a simple dc motor with an encoder could also be used but it would need something to read the encoder.
i think steppers would be the best choice for this.

Joker94 · « **Reply #13 on:** July 20, 2009, 05:28:03 AM »

di you know that 85% of the humans thought process goes into balancing alone, there for replacing humans i a hard task to replicate

wil.hamilton · « **Reply #14 on:** July 25, 2009, 10:40:30 AM »

Would this be possible to do? Yes.
Would it be possible for an individual to do? Most likely not.

Trying to mimic human behavior is rather hard. To try to control a windsurfer like in that video you need to consider a few things. First, how does a person control a wind surfer? By a combination of shifting their weight and by positioning the sale properly against the wind. The person also needs to be able to correct and predict for any water conditions (waves, obstacles, etc). While creating a robot that would be able to shift its weight would be possible, having it do it the proper amount to control a windsurfer would be an almost impossible task. In addition adding the sensing required to properly position the sale in the wind would be very difficult. Thirdly, being able to read the water conditions with current video recognition software would be almost impossible.

Really, i think attempting to do this would be an almost impossible task. it would require years of work ,experience and a lot of cash.

airman00 · « **Reply #15 on:** July 25, 2009, 10:56:37 AM »

Quote from: wil.hamilton on July 25, 2009, 10:40:30 AM

By a combination of shifting their weight and by positioning the sale properly against the wind. The person also needs to be able to correct and predict for any water conditions (waves, obstacles, etc).

Why make it function like a human? Do submarines go underwater like humans? No they dont.
Make a machine to do the functions necessary, maybe a rudder would help out too? Sail can be moved with a servo. Its definitely possible, it would be a R/C sailboat with a balancing feature. Use PID for balancing.

By the way, how big are you planning to make this thing?

Joker94 · « **Reply #16 on:** July 25, 2009, 06:52:10 PM »

i had one though. I read a thread where a guy wrote some code and hooked up a web cam and could identify red blobs and woud place certain coloured rings around the ascending sized blobs. this may work to icentify the waves, I'll try find it.

jamesjim · « **Reply #17 on:** July 27, 2009, 12:38:01 AM »

Thanks for you replys
Much confusion has been created due to my inadequate explaination of the project.
Firstly, a human will be on board the vessel.
Secondly, the human will do everything required to sail the vessel except balance.
Balancing the pitch and roll axis will be the job of the robot.
Steering will be the robots job(using banked turns) but i don't care about working on that at the moment

The purpose of the project is to make a single foil boat (which is what the windsurfer is) but with a hull, and possibly to be a larger vessel that could carry more people and or stuff.
I have never seen a single foil sailing boat before. There is only one sailing hydrofoil called the hydroptere and it is limited by the size of the swell. I hope to build a boat that will glide above the chop and balance itself from above, as with the windsurfer, rather than having an extra 2 foils to balance the boat under the water. I want to have robotic counter weights do that job for me. Using an IMU to measure pitch and roll and adjust the weights accordingly along the x and y axis.
The windsurfing hydrofoil is limited in size due to the mans size.
The windsurfing hydrofoil is limited in range due to the mans endurance and the size(storage) of the vessel.
Using the robot to balance pitch and roll would remove these limitations, so that a larger vessel could travel greater distances. The advantage of an ocean going hydrofoil is speed the windsurfer is about twice as fast as a yacht.

Money and time are issues I believe i can work my way through at this point. What is important for me is my enthusiasm for the project.

Thankyou Smash for your threaded rod suggestion.

Thankyou everyone else for your contribution.

Joker94 · « **Reply #18 on:** July 27, 2009, 05:50:01 AM »

so in a nut shell you want a small hull (a boat) with a single foil sail to balance on the paddle thing.

Soeren · « **Reply #19 on:** July 28, 2009, 05:30:45 AM »

Hi,

Quote from: jamesjim

I haven't really done anything with robotics before [...]

Quote from: jamesjim on July 17, 2009, 02:06:39 AM

Score so far

YES NO

2 1

thankyou and keep it coming

Sorry to put you down, but if this was a betting match, I'd put my money on a big NO (and win I'm sure), but if you have half a lifetime to dedicate...

When you count the opinions, perhaps a valid weighing factor would probably be, whether it came from a slightly teenaged kid dabbling in "hot glue robots", or a seasoned electronics engineer winning bread by working with cyborgs - making general robotics and a diverse spread of electronics in his spare time.
(Don't have to tell you which category I represent I'm sure ).

You won't be able to just control roll and pitch and Bob's your uncle - the surfer use a lot of skill in controlling the sail as well and using a lot of muscle groups and probably loading both his afferent (sensory) and efferent (motor) neurons pretty hard in the process.

But... if you reduce your anticipations, you might get something exciting to spend your time on - one should never stop just because something is very unlikely to complete - We're in it 'cause we can't leave it be - it's the ride that makes us tick and when we reach our destination, we get bored and take off on the next ride - so "happy trails", I cross my thumbs for you.

Best way to analyze what control you need would be getting a talented surfer to carry a lot of sensors and send him on a board with lots of sensors (like a gyro, an accel.meter etc.) and use a datalogger to see how much is going on.

wil.hamilton · « **Reply #20 on:** July 28, 2009, 06:18:03 AM »

Quote from: jamesjim on July 27, 2009, 12:38:01 AM

The purpose of the project is to make a single foil boat (which is what the windsurfer is) but with a hull, and possibly to be a larger vessel that could carry more people and or stuff.

If you're just looking to build a single hulled hydrofoil, it might be easiest to have it motorized instead of using wind power, then you only need to work on balancing it.

Joker94 · « **Reply #21 on:** July 29, 2009, 05:26:30 AM »

that sounds like the way to go, you can add the foil on later but for the first rig just go with an out board.

wil.hamilton · « **Reply #22 on:** July 29, 2009, 08:31:57 AM »

i wouldn't use an outboard, i would keep it smaller scale and use a motor assembly from an RC boat or something similar. That would keep it cheap as its only a concept. making it big (say, if you used an outboard) would make it rather expensive.

Joker94 · « **Reply #23 on:** July 29, 2009, 07:23:24 PM »

good point start off with an rc boat. If you go big an out board would be the easyest and most time saving. If you get a small one you can get the really cheap, remember they don't need to be the best of the best for this application.

News:

Author Topic: Do you think this is possible? (Read 5604 times)