Monowheel with Stability Control
Vložit
- čas přidán 5. 12. 2021
- AD Analog Devices’ Active Learning Modules and Kits are available from Mouser Electronics, Inc. making the study of electronics accessible, intuitive, and fun: eu.mouser.com/new/analog-devi...
I've been watching a lot of Monowheel videos recently, and they look like a lot of fun. I decided I should build one and also try to add some active control so it's easier to drive.
CAD & Code: github.com/XRobots/Monowheel
You can support me on Patreon or buy my Merchandise:
***************************
Patreon: / xrobots
Merchandise: teespring.com/stores/james-br...
***************************
Affiliate links - I will get some money of you use them to sign up or buy something:
***************************
Matterhackers 3D printing supplies: www.matterhackers.com?aff=7500
Music for your CZcams videos: share.epidemicsound.com/xrobots
***************************
Other socials:
***************************
Instagram: / xrobotsuk
Facebook: / xrobotsuk
Twitter: / xrobotsuk
***************************
CAD and Code for my projects: github.com/XRobots
Huge thanks to my Patrons, without whom my standard of living would drastically decline. Like, inside out-Farm Foods bag decline. Plus a very special shoutout to Lulzbot, Inc who keep me in LulzBot 3D printers and support me via Patreon.
HARDWARE/SOFTWARE
Below you can also find a lot of the typical tools, equipment and supplies used in my projects:
Filament from: www.3dfuel.com/
Lulzbot 3D Printers: bit.ly/2Sj6nil
Bearings from: simplybearings.co.uk/
Lincoln Electric Welder: bit.ly/2Rqhqos
CNC Router: bit.ly/2QdsNjt
Ryobi Tools: bit.ly/2RhArcD
Axminster Micro Lathe: bit.ly/2Sj6eeN
3D Printer Filament: bit.ly/2PdcdUu
Soldering Iron: bit.ly/2DrNWDR
Vectric CNC Software: bit.ly/2zxpZqv
Why not join my community, who are mostly made up of actual geniuses. There’s a Facebook group and everything: / 287089964833488
XROBOTS
Former toy designer, current CZcams maker and general robotics, electrical and mechanical engineer, I’m a fan of doing it yourself and innovation by trial and error. My channel is where I share some of my useful and not-so-useful inventions, designs and maker advice. Iron Man is my go-to cosplay, and 3D printing can solve most issues - broken bolts, missing parts, world hunger, you name it.
XRobots is the community around my content where you can get in touch, share tips and advice, and more build FAQs, schematics and designs are also available. - Věda a technologie
Nice work James, this is something I need to incorporate into a future design it would definitely save a few bruises!
your channel is awesome, keep building :)
Ditto.
Why not add a gyroscope like James looked at with the balancing robot on the inline skate wheels, seems like the rotation might help balance and use that as the mechanism for turning/leaning, seems like at human scale you could pack one into a location that isnt in the way of the driver, such as under the seat to help keep the weight low to the ground. also have you given any thought to the idea your tire might need to be larger to allow it to counter steer like a motorbike?
Great idea!!! Beautiful implementation! Btw, double helical gear would work great here to keep big wheel centered without additional bearings.
Herringbone pattern?
That would be pretty cool
That is very smart.
@@jeffcarter4500 Double helical is the same as herringbone but with a gap in between the gears. The gap is useful for metal gear fabrication. For a 3D printed part, it is not required and therefore herringbone is more compact and suited.
Monowheels have been around for a LONG time and for sure their Achilles heel has been the problem of stability in both acceleration and braking. Glad to see someone really give this the analysis it needs finally.
Hello! I'm an EUC rider. I suggest you to replace the roll control with a yaw control moving the two masses in front and back in an alternating way. The stabilization software will use yaw to stabilize the device around a set point and the remote control will change the set point from the center to the sides in order to control steering. I'm pretty sure that you can achieve a more precise control at low speed because I do it all the time 😉
well, less steering and more leaning
@@Blox117 it is the opposite. You have to use steering to maintain balance and control leaning. Is the same thing that you do when you ride a bicycle. If you try to act directly on leaning moving a mass like in the video it is very difficult to control the device. Is the same to ride a bicycle with no hands
Hi, another EUC rider here. I agree,, a yaw control would be like us twisting our hips which we do a lot , to stay up at low and zero forward speed.
Veritasium had a video on this, and it seems rather convincing. Essentially, to initiate a turn, the rider steers slightly opposite to the direction of the turn, so that as a reaction he will lean into it and carry on the turn leaning on the correct side. This is a very small input and done istinctively, but can be tested by locking the handlebars so that they will only steer toward one side.
Also an EUC rider! Yes this would work. You steer left to go right and it's all about getting the right lean angle. @James Bruton: You should try an EUC. They are the best invention I have seen in my lifetime!
A herringbone or convex/concave gear pair would self center, removing the need for those side bearings. Some of the side to side issues could be because as it tilts the mass right, in order to steer right, the wheel will initially tilt left due to the reaction force. I'm not sure what the solution would be for this though. Maybe the mass ratio between the ballast and the wheel is important.
i've been thinking about two disks that are counter-rotating with masses on front of the one disk and back of the other. that should keep the torque at zero.
@@ChrisBigBad Like his previous one? Would that be classed as active where as the comment above would be passive?
@@andycrask3531 ah! The disks are not permanently skinning. And they are just weighted in a single place. Also i imagine them being both parallel to the ground. The disks only rotate to bring the weights into position. In neutral, the upper disk might have the weight in front and the lower has it in back. For maximum tilt, one wheel turns clockwise 90 degrees and the other 90 degrees withershins. The accelerations cancel out, but now the weight is it balance. Maybe. I cannot see any fault in the mechanism. Could still contain a problem that I cannot see
@@ChrisBigBad wouldn't it be a problem that the two disks and masses are not at the same height?
@@Jehty21 sure. But either make the disks of different diameters or just use different masses.
Actually, since the masses never need to meet, you can have the lower mass sticking up and the upper mass sticking down. And you only need one motor between to drive both disks.
Actually actually, make the disks have teeth and use the motors themselves as the mass. Mount them on bumps sticking out from the disks with berrings above and below the other disk for stability. Drive the motors with the same signal but wired opposite.
Actually actually actually, never mind the disks. Just mount a spoked ring with a gear track around the top edge. Clamp the two motors on with their drive gears on top and berrings underneath.
Hmm... can you use them for both balance and steering if they do not always move symmetrically? Or just use opposing reaction wheels for everything including gyroscopic?
Brilliant! The way you talk through your analysis and decisions of what to do next for each iteration is *invaluable*. Great work.
A note on the pull to one side: this can be caused by the pile of the carpet folding in one direction which drives the bias. Great work as usual, James!
Yeah. My first thought also was that it was probably the nap of the carpet.
Exactly this! 💯
It's so sweet watching it try desperately not to fall over!! 😆 & I LOVE when it knocks the light over too!
I love that you are covering all these different forms of robotic transportation
Awesome build and I like the thorough problem solving approach you took.
James is a master of always explaining the problems he is having whit his builds and never doing the oposite of showing how things should work so thats the same vid over and over again.
I was so happy I saw this in his community feed. I had three same videos in my feed as well as watched them. When he says he'll make a gyro to make it easier to ride. I was so excited
The Carpet Pile is also responsible for the side drift of your Monowheel. The fibres don't stand perfectly vertical. They all lean in the same direction, which let them bend more easily to one direction than the other. So your Monowheel is running "downwards" the bending fibres.
I would love to see you add gyro balance to this, also if you have the gyros spinning in the opposite direction you can apply a brake on them to use them to compensate the forward roll when braking.
Yes with 2 counter rotating gyros this should be doable
I think adding a gyro would be kinda lame.
He already proved in another video that a one-wheel device can balance with gyros.
Now doing it without a gyro is much more interesting.
@@WetDoggo should be fine with one, just accelerate or brake it to apply the torque. Could also use a control moment gyro
I think you might mean reaction wheels? i.e. wheels parallel to the main wheel that spin up in the opposite direction as the main wheel to compensate for that inertial buildup and pitch.
@@quinnobi42 A control moment gyro is like a reaction wheel but it spins at a set RPM and is rotated in a cage to produce a torque. Same idea as a reaction wheel, different implementation.
If you increase the mass of the outer tire, you can increase the angular momentum. Then it doesn't need so much active stabilisation.
Nice Project. Greetings from Germany.
Great idea James ! Like the concept.
James, great work. I’ve just added your channel to a list that I provide to my second year mechanical engineering students for inspiration. If may be so bold as to suggest a video topic, I would love to see a video in which you discuss the process of choosing a motor. I think that this is topic that does not get enough coverage but becomes a daunting issue when a new “maker” decides to purchase a motor and finds that there are literally thousands of different options.
The BB8 build! Thats what brought me here all those years ago!
Very cool, looks like fun to drive. I like how James blanks out the numberplates in one shot and in the very next shot they're all clearly visible.
That was awesome, as all your videos are!
Keep up with the good work!
Gyro stabilisation and steering is next level of this project!
Yes yes yes I’ve wanted this for so long. You stimulate my dummy brain enough and I just know if I can figure out how these work I can make something crazy. Love your work!
Watching at 9:00 gave me flashbacks of learning to ride my EUC ahah amazing work as always.
Your theory that the room you were testing in at first having an ever so slight slant to one side reminded me of the the room that one of the classes I took back when I was going to the local community collage was located in. The teacher would have the students look around the room for a few minutes and jot down what stood out to them. What most if not all the students would miss was that there a spot on the floor where there was a slight bump that couldn't be spotted with the naked eye or just walking around, one would have to slide there feet over the location to notice it and the teacher would have us do just that.
Awesome project! I've been fascinated with Monocycles ever since Revenge of the Sith.
You know i have recently watched one of veritassiums videos where he explains the way we ride bikes, in most cases to counter act the toppling over i would call it you need to momentarily steer the opposite direction to your intended one in order to initiate the proper slant for a balanced corner
In 1983 whist in church I sketched this on the bulletin. Realized a single wheel would be unstable so added another wheel to the side. The reason I was thinking about this was to have a vechicle with a large diameter wheel that could climb obstacles. Figured the powering unit would be lower, thus providing stability. Had no way of fabrication at that point, and was surprised to find the idea had existed back in the early 1900s.
This is really cool and I'd love to see a v2 of this.
Really cool start of a probably (hopefully) long process. I think that the delay in the compensation reaction is causing part of the problems as well. You would want the lowest latency for these kind of control setups, as when it compensates too late, the machine might have moved towards the undesirable side even more and it will then compensate too late again and it gets out of hand quickly. Ohwell, what am I thinking, you definitely already know this yourself. I am awaiting the next video patiently!
This reminds me of the veritasium video talking about how, in order to balance on a bicycle, you steer to keep the wheels under you, and when you turn, you have to turn the other way to make yourself lean into the turn beforehand.
I'm glad you mentioned active gyroscopes at the end, I'm looking at having ago at one of these myself and think gyroscopes is the way to go - I've been watching a lot of boat gyroscope videos lately :) , I was kinda surprised you didn't mount the main drive engine in the centre either vertically or horizontally to help with initial problems with balance and keeping the majority of the weight at the bottom...
I'm curious to know why you chose a straight tooth gear profile with bearing guides vs herringbone profile? Herringbones self align and typically run much quieter. Great channel and projects!!
James, an idea: you could get a bike wheel and 3d print the internal gear profile in a number of sections, which could then be fixed through the spoke holes with hardware. Why? Well, it's cool for a start, you can get a bigger wheel diameter than you could print, and you get to use real inflatable tyres which i feel would improve the handling characteristics of the vehicle quite a bit!
honestly that sounds like a great rc project
Thank for considering putting a gyro in the next version, I always speculated that it would help the center from spinning in upside down. Out of all the monowheels Ive seen noone had put a gyro in em
I have a script for a movie involving a robot that follows the main character around and this video just proves me that we can defenitly build it and not use CGI
Just sayin i cant believe the stuff you come up with and how fast you can make it ,,, awesome just awesome
I’m wondering if some of the instability you have sometimes is because the ballast that controls the roll is moving with the rest of the drive system. When the main drive system tilts forward or backward, it also lifts the ballast, giving it more mechanical advantage to roll side to side. This coupling could therefore be adding extra dynamism to your system that is hard for the electronics to account for reliably.
Not sure it would be possible to stabilize the ballast due to inertia, so perhaps modifying the code to somehow account for the pitch of the drive in how it rolls?
Its because they are trying to drive a balancing wheeled vehicle with leaning alone. Which doesnt work.
All wheeled vehicles the rely on balance to function need countersteer to balance properly. This is true of monocycles, bicycles, motorcycles.
So without a way to twist around the z axis and lean this isnt going to work. At least not as a stable under a wide variety of conditions design. Though I believe that with enough control authority and fast enough feedback it can brute force past this.
Regardless, the code should take that angle into account, to refine the calculation.
He's also fighting the inertia when the ballast initially swings over
@@TheGlitchyCorgi that’s true, he may actually want to swing the ballast the opposite way.
@@aDifferentJT That is the definition of counter steering. Start all turns by first turning slightly the other way. You are in violent agreement with @Nocare.
Another great vid.... nice work, mate! I wonder how much of an effect the "pendulum" angular acceleration counter-torque is having as it moves to steer. Its moment of inertia is pretty big so more aggressive steering could be induced some funky stuff.
All electric unicycles are self balancing in this way. Awesome project and effort!
Had a laugh at the licence plates blacked out in the first shot but clear as day in the second.
For the gyro steering you should use a high torque high speed servo like a heli tail servo. Awesome project.
Fantastic job 👍
5:18 You have answered an age-old question with a scientific answer!
what do we do with a drunken sailor?
measure the angle with an IMU, use a controller to drive the actual angle, drive the motor speed keeping the angle as near-constant angle as we can.
With Soft Ware Defined Radio you can also build a radio telescope! My father retired from the Green Bank Observatory and has been building his own Radio Telescopes for years and I have seen many an SDR box just like that turned into a Radio Telescope!
You need to do three nested monowheels, the outside one is the one for driving, the inside two would be simple weighted reaction wheels. Your instinct to put a lot of weight down low isn't necessarily correct, because it's still above the point of support. Better to have a uniformly distributed mass that you can spin and impart inertia to.
Like a flywheel that you can store inertia in, draw from it when accelerating and then use as a reaction mass when braking, nice! It would double up as a gyroscope to aid stability side to side as well.
Good job on censoring the car license plates - except for a couple of frames ;)
RK60-ZHG
FP57-CIL
Very nice project! I would be very curious to see if you could get active reaction wheel control could keep it balanced.
You discovered an EUC. :) The future of city commuting.
veritasium just did a video on how bicycles don't steer the way we think, I would recommend that video but the gist of it is that we need to countersteer the bicycle to keep the center of balance correct, so I wonder if you would have better control by steering in the opposite direction first, than steering into the turn, love the idea it's really cool!
That's so cool! Next time a case to protect the electronics could help
Interesting, nice work
Guess im not going to bed yet
Adding big concentric gyro ring inside the tire would look incredibly futuristic. Very very cool.
Great work
Cool project! However, I see a flaw in the latteral stabilization. When the monocycle moves the counterweight to one side, the rest of the vehicle moves to the opposite side because the center of mass wants to stay in the middle. Since the wheel is so thin, you won't get any reaction force to actually tip the centre of mass over to one side. What happens is that the accelerating counter weight creates a moment on the monocycle in the opposite direction, meaning that you'd expect a latteral force on the wheel contact point, which you probably can take advantage of. Just a reaction wheel could potentially work better than a counter weight. But then you also have the gyroscopic effect of the wheels together, which could be tricky to handle
it gets pushed to the opposite side really
Hi James
I have a question. Why don't you turn all your previous robots into one that includes everything? For example a BB8 or R2D2 that seeks its own way and explores its surroundings. At the same time, he can react and interact with conditions, colors and people. His arms could grab and when the battery is empty he can find his way back to the charging station. With a neural program he could learn.
I really love these experiments. :) You should get yourself a larger workshop though so that you can test all your creations properly. There should be plenty of locations available for cheap now due to the pandemic.
I would still say nothing beats the attic... just that I never feel like getting out of my room 🤪 as much as my family would want me to😬😬
Awesome, always something to learn... had been waiting for it. Can adding a horizontal spinning wheel help it to go slower? Giving it some more stability... you did mention it at the end but those were vertically mounted? How does it differ to mount horizontally?
As I was thinking about the steering on this, I think I can foresee a problem as you increase the balance mass. In order to move the mass, you have to accelerate it - impart some angular momentum to it. This, in turn, imparts negative angular momentum into the entire system and this "negative" momentum will tend to tip the whole wheel even further, accentuating (momentarily) the whole lean.
This requires that the pendulum be moved further in order to compensate, which pushes it further over which...
Of course, as you decelerate the pendulum, that will now try to roll the system in the other direction.
Could this be the source of some of your instability problems? And the problem will get worse as you add mass to the pendulum.
This same issue is also the reason that the steering seems to be delayed.
This recent video by Veritasium explains reasonably well how this applies to bicycles:
czcams.com/video/9cNmUNHSBac/video.html
I think the solution would be the same, both regarding stability and steering, as with bicycles, to countersteer, or "counter-balance" the monowheel. This makes the programming a hard problem, though, because it has to know if it should steer or countersteer at every single perturbation, or really every single instant.
I suppose the calculation would be whether or not it already has enough angular momentum in the correct direction to right itself if it were to begin steering, in which case it steers, otherwise it countersteers until it creates enough angular momentum.
Actually, I just thought of a much more robust solution to this: two steering mechanisms, one to apply countersteering, the other to apply steering. The countersteering operates a flywheel-like ring, which means it doesn't actually contribute to the balancing of the vehicle, but spinning it can generate angular momentum. (This is of course a reaction wheel system.) You then have the steering mechanism as on the vehicle, which you program to have the sole job of spinning down the reaction wheel. (Particularly by unbalancing the vehicle so that the reaction wheel system will then accelerate the reaction wheel opposite to how it's already spinning, of course.)
This vehicle is called an intracycle. To prevent somersaults, you need to place the center of gravity as low as possible. Heading control is carried out by shifting the center of gravity to the left-right. (Google translation)
Mr garrison on south Park had something similar. Im glad you left the mouth control off
I love this project! Maybe you can make adjustable gain from the remote controller? Please continue to develop more mono-wheel stabilization techniques. (gyro) Its such a futuristic looking bike, and its a pity to look at makers driving them with stability issues. I hope others will learn from your experiments and will be able to make more stable mono-wheel bikes.
Love your channel, you should mount the batteries and the electronics on the swinging device this way you will use the weight of the electronics instead of adding weight
12:18 Looks like the wheel hit a rock and got some air. The way it tilts causes a significate amount of torque when it lands back down.
1m bro. I found you when you had 10k and I was like why doesn't this guy have a million yet?
Loved this video it's Amazing!!!! you Should make it a little bigger and Close it on the sides making "A ball" But it's a Monowheel that's enclosed and Then I could turn it with Two Giant gyros!!
Heads up on the pull to one side. That could be the carpet piling alinement.
I’ve noticed myself that the direction of carpet pilings makes a big difference to RC cars driving over it. Cars will pull heavily to one side when driving at a right angle to the pile direction. (In certain types of carpets.)
Amazing work as always.
Maybe you could help me understand what would happen if there were a pair of counter rotating gyros mounted on vertical axis in the slot between the two idler gears.
I saw you were using the off axis movement of the gyro to move the other one.
Could this be used to dampen the accel decel and to give tilt?
I think you should look into counter steering in motorcycles. Tipping the wheel in the opposite direction of a desired turn to initiate the turn then shift the weight into the turn to maintain the direction.
The biggest question here is if your house is sinking. Doesn't appear to be much, but that's something that happens with old houses. Assuming your house is the same as others i've seen in some of the neighborhood shots, then it's quite possible. In my town, on one of the bigger roads, there's a two level house that's so sunk on one side, you can visually see it from the outside.
It would be very awesome to see a kinda working droideka model in the future o.o!
I love your videos. I always struggle watching you use the wrong Phillips or Posidrive screwdriver for the screw. That's just me. Still an awesome video as usual.
Well done.
Cool video as always. What software do you use for your printed gear designs?
That's a wheelie good RC vehicle, I bet it'll run rings around others.
As I'm tired I'm going to motor off and dream of donuts. 😋👍
If you add a bezel in the center of the Wheel (so like through the center of the cogs), you could replace the idles with two just two sticks (or normal wheels), which should help reduce the friction from the top two idle gears
Love your projects, James! You might want to take a look at 12:24 again though, you forgot the blurs in that shot!
Very smart and cool.
I think the random crash at czcams.com/video/13JNW_H3DSg/video.html was because it got air and the drive mechanism caused a massive tilt, which only affected the monowheel because there was no longer traction on the ground. In other words, your system relies on the monowheel maintaining constant contact with the ground.
Also, the carpet actually has a grain to it. Like an angle to the bristles.
Have you ever thought of using a non-linear gain? This way you'd have more control over how the vehicle reacts - for example, if the thing is experiencing a lot of random low value wobble, trying to compensate too violently for this may just make the matters worse, but at the same time if it tilts too much a violent reaction would certainly be beneficial. Also, I'm pretty sure that if you'd graph force needed to put the vechicle upright vs the angle of tilt, the result wouldn't be linear ...Does that make any sense?
Try different tire profiles. The curvature of the tire patch will determine the sensitivity of the steering. And a softer tire will be easier to control too.
God damn you're bloody brilliant!!!!!!
I would be interested, how the stability of the side to side control, is affected by speed. And wether adapting the control scheme to this does significantly improve the controller. I.e. smaller gain, when driving slower, or (manually) speeding up, when near maximum tilt to stabilize again.
curious as to why you didn't just do a herringbone gear, instead of the idler alignment bearings
Beautiful
Nice project. For the roll compensator, did you tried to put all the mass in the middle instead of equal weight on the side?
Could you match the gain inversely to the speed? So that as you get going faster it mellows out it's compensation, but at slower speeds it is a little more aggressive?
well the solution to monowheels is actually having an inner and outer wheel. The inner one keeps inertia while the outer one starts and stops as a normal one does. I think of a planetary gear system that is lockable as a viable solution so perhaps 3 wheels an inner one with high mass, and 2 outer ones with very low mass so hey can accelerate and decelerate quickly. Turning can likely be handled with a leaning mass.
Take a shot every time James says "angle."
Also a reaction wheel (or inertia wheel) can be a solution for the roll control.
Could you hybridize a monowheel with a quadcopter design using the quad for stability control on the ground and jumping/flying over obstacles?
This man made a remote control wheel.
I feel as though instead of attempting to compensate for the roll with balancing a weight. The underlying problem was that the yaw and roll needed to be fed into the equation for the pitch. This is essentially a reverse gyro...
Braking is always going to be limited by the mass of the stator. A smarter design would be to have some deployable wheels (think wheelie bar) which can drop to the ground or contact after a certain degree of rotation to raise the brake force limit to available traction. I'd never ride one that didn't have this feature at more than a walking pace.
or even some completely/ mostly passive auxiliary wheels that in normal operation never touch the ground but touch the ground when braking over the limit mainly so you dont have to rely on a mechanical system to deploy a wheel
Was looking for a reliable EU electronics shop. I think I'll give Mouser a try.
If you use herringbone gears, not only will it not slide out, but it also will be much quieter. Especially if you notch out the joint in the middle (small gap)
Nice project. How do you get angular velocity from mpu6050( i mean how do you calculate it from the angle) to complete your contol loop?
YES PLEASE DO THE GYROSCOPE VERSION
Ayyy I love Mouser