@@GiuseppeGullo Not likely. It cannot do some characteristic 3 DOF real joint really did such flexible radius arm. But this polar joint has fixed radius arm.
@@clementyap1009 This is literally seen on earth. You just saw it there. What you meant to say is that there are people in the world who are far more clever than most of us are.
what's the most amazing thing about this is that it's eventually going to get even simpler. People think complex equals smart but it's the opposite. Beautiful design!
I would say that the one thing that I could see being made simpler is, to make the sphereacle gear on a CNC 5 axis mill, or to make it in such a way that you have 3 sets of gears cast them and a few thousands too big and have them run together to give them a perfect fit as well as work hardening them at the same time. ( i may be wrong about this ) But the design is already very simple, once proven economically viable for some there will be alot of patent avoiding, or patent disregarding chinese knockoffs that are more complex.
Some of the motion pairs between the links still have sliding/friction - not evolvent rolling. normal gear transmissions are invented to preven the sliding. So it isn't that great, but the complexity is mind boggling and the idea looks really cool.
My thought as well, that lubrication and keeping the system clean will be tedious. Also I suspect that there are "dead spots" with less strength, torque or otherwise, near poles or along the equator.
Not great how it reacts to singularities in the control inputs either, having to spin the drive gear 180 like that that quickly (and for no actual effect on the output) is terrible for the motor drivers.
@@fridje the motor drivers? the FETs? I suppose but you should spec your parts and designs accordingly. It's no dewalt drill which I'm still pissed about if you wanna talk bad motor drivers. I bet if the drive gear went directly above the pole it wouldn't spin given some pre determined dead zone for the poles(jamming may become more evident). I cant help but think the spinning could really be mitigated or completely cut out.
@@fridje Yeah I was wondering how frustrating it could get writing code for that thing then writing enhancements hoping they don't jam up the works. I wonder who and why tho? now a days a kid with a cnc bed can make these things.
@@Asdayasman The main sphere is easy, the monopole gears would be harder to model, and then the whole drive mechanism is only briefly shown. There is enough detail to replicate it, but obviously it was modeled at some point, so it would have been nice if they released the models along with the paper.
@@DrakeOola Current/classical robotic arms have only 1 DoF per join ....this one has 3 DoF, like our own human upper arm join ....i can imagine using it in prosthetics i.e. robotic arms....
@@juancarlosabad3298 I may be wrong but from watching the video, it seems like there are certain movements that this joint can’t replicate, although I’m not sure those are movements that a shoulder joint can perform either, so that may not be an issue.
@@babybirdhome Our shoulder joint can't rotate indefinitely. So, there is definitely motions our join't can't replicate that this machine can. And I suspect the tradeoff is our joint is (relatively speaking) capable of much higher torque than this (with less play in the motion). It's all a tradeoff. The more options we have to make joints like this the better an engineers life will be. This joint might not be great for a shoulder for a heavy load-bearing joint, but might work in other scenarios. We need more options so we can pick the best ones for the situation we need them.
You'r not stupid, they are not geniuses... had this idea for my J ENG graduation project, like 7 years ago, and my currator said - "put it aside, if you want ill help you to patent it, now focus on finding less complicated project". I found out that 90% of technicaly educated will give you 10^4 reasons why something will not work and only remaining 10% will try to work out with you those couple of ways the damn thing will work. If your idea generator find the 10%ters but be prepared to be the laugthing stock untill you make prototype and some money)))) monetary success shuts up ppl very fast)))
No, the real stupid people are those who are to stupid to know that they don't know that they don't know anything. D'you know what I'm saying? 😄 It's Dunning-Kruger all the way. You can only get better at things by accepting that no matter how much you know, somewhere somebody will know something you don't. Learning is not an ending thing, it's a constant work in progress. Like counting from 1 to infinity, you keep making progress but still you're not near infinity.
I'm happy that there are people in the world who can understand and design this. I had to stop watching mid-video because I was getting beyond confused and starting to ache at a fundamental level.
Why do I feel like a description of the driving software would sound like "The ball joint knows where it is, because it knows where it isn't." For those who don't know: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-bZe5J8SVCYQ.html
I love being alive right now and watching technology grow at a ridiculously fast rate. The stuff we are coming up with as quickly as we are is astonishing, and its only getting faster because of it.
Here's the Timetable for those who want it: 0:50 Graphical explanation of the mechanism 2:45 Components of the manufactured prototype 3:40 Motions os prototypes 4:20 Motions of prototypes with output link 5:05 Behavior of the Monopole Gear in the vicinity of the pole 5:25 Motions with extended output link and weight
@@OrangeC7 While potentially true, the original statement is also a classic example of false equivalence. Our musculoskeletal system is extremely complex, so much so that we don't have an actuator that fully replicates muscles in a 1:1 comparison. Whereas this uses fairly simple mechanics to actuate a ball joint. It's not an apples to apples comparison, it's damn near the furthest thing from an apples to apples comparison, the only comparable part is that a ball joint is actuated in literally any way, thus false equivalence.
@@DirtyRobot your ankle isn't driven by two sets of gears, its pulled in different directions by muscles, closer to a hydraulic system than a mechanical gear.
@@DirtyRobot As Battleoid said, this is very different from animal analogues because there are no tendons/muscles. It's like a moving skeleton, actuated from the joints. The leverage is really enormous.
A beautiful mechanism. Elegant design. Perfect application of two rotation axis symmetry and 3d printing...I see applications in multi axis machine tools, robotics, and medicine for starters.
I had something completely different in mind when I saw the title - basically 2 spherical burrs with their grooves meshing. This would allow one gear to be tilted off axis and, providing the contact is maintained at the same "latitude" on both burrs (requiring a slightly complicated hinge mechanism),, the rotational speed would be unchanged as they tilted. Chaining these together would allow torque to be transmitted from one end of a flexible chain to the other without the friction loss or binding due to buckling of a cable in sheathe. But yours is nice too.
If you put a soft ball at the end of the output link, and use 2 or 3 of them close together you could use it as a claw that can rotate the object without rotating the claw itself and without loosing contact. This is excellent for fruit collection robots. Of course it depends on the torque it can produce but this thing seams pretty sturdy for the freedom it gives
If you can make this strong enough to withstand the forces, this WILL revolutionize machining! Imagine all of the angles of approach on a part you can have if the part is mounted in place of the output link!! Incredible!
I wonder if a useful application for this might be able to avoid that situation through software. It obviously depends on how necessary this specific design would be for said application
I was wondering the same thing. But directly at pole it would be possible just to rotate the direction of the monopole gear, instead of pivoting it really fast. But this is only possible AT pole, not in the vicinity of the pole. I also wonder if mounting the motors opposite to each other instead of under an angle has influence on where the pole point is.
@sprock there's many autists on the planet, who are desperate to get every little bit of attention by vommiting memes where they are irrelevant. This is one of them. "BrurURUHuHUhHUHUH" and he jizzed over his keyboard. Probably.
This is immune to gimbal lock due curved gear joint teeth. This gear will be free to move from any steady position. Edit: 4 actuators required for 4 motors (two for pitch setting and two for drive).
It doesn’t seem like this is immune to gimbal lock to me. This seems to have a few inherent weaknesses and movements that it won’t be able to do, but I’m also not an engineer and am not certain my intuition is correct about that.
@@Inertia888 Worm gears are fairly simple to understand, think of them as a screw rather than the typical gear. A rotation of a screw will progress the teeth linearly by a certain distance. In actual screws, you can see this being done by the screw progressing into a material. But in this instance, you need to reverse which part is moving, here the screw remains stationary in space while the complimentary piece moves; and that complimentary piece is a gear that rotates on an offset, perpendicular axis. You can also think of it as a rack in pinion, where a rack would move linearly to rotate the pinion. Except this doesn't have the space for a linear rack, so you take the 2D profile of a rack and wrap it around an axes where one revolution aligns with the next tooth on the rack, thus creating a screw that can be rotated to rotate the pinion through rotation of the screw. Of course this is an extremely simplified explanation of worm gears, some worms do have much more complex designs than this. There's countless videos on explanations and animations on the internet if you still don't understand them.
@@xaytana I think your explanation is actually more confusing than the video... I understood it when it was on video but trying to understand your explanation completely ruined that xD.
This ball joint looks really interesting! I am curious what kinds of gear ratios could be applied to it and what maximum loads it can take. It doesn't look like the gear teeth could take large loads but maybe it could be used in prosthetics.
This is really interesting, but I am not sure what problem it is meant to solve, it is very complicated if all you want to do is make an actuated ball joint
This isn't just an actuated ball joint, it is a true 3dof actuated ball joint. Linkages limit your range of motion and require rotational limits. With the exception of the socket limitations, this has none of that.
@@pstrap1311 everything mentioned above for sure. However maybe a non-traditional use would be digging through soft soils. Having sharp points like that and being able to rotate them in any direction would aid in cutting through the soil and then removing the soil. If rocks were encountered it would help steer the vehicle around the rocks. There's really any number of applications. My favorite application for this though would be Omni directional steering. There's this thing called a magnetic gear, if This were re-engineered to use that then this is the perfect omnidirectional wheel.
@@DonnyDarko13 it appears to be quite precise, and being gear based, precision is a function of machining tolerance. I'm certainly no expert here, but it certainly looks applicable
@@zactron1997 i agree with you i am not an expert too but the Shape of these tooth on the Ball is a wedge when you compare it to an trapezoidal- thread there isat everytime a little bit of clearance and when its only a few tenth of an millimeter.
Very interesting idea! I'm going to have to do some reading now. What are the torque/power characteristics of this type of system? The wear characteristics? With 3DOF and 4 motors, it seems overconstrained, how does it deal with binding and accumulated/integration error? What about back-driving/holding on power-off?
This is mesmerizing to watch. Very cool. But what is its use case? Robotics is the obvious answer, but is there another application this is designed for that I'm missing?
interesting, though some comments on play and stiction would be required to assess the usefulness of the setup. e.g. towards instant 4'30" for instance, one can clearly see the lack os smoothness oof the motion, which is likely to result from non constant friction of a plasticly constraint setup in order to reduce play. then there is the difficulty of the computational inverse geometry with poles...
didnt watch the video yet just came here to quickly comment this is a really cool concept, but wouldnt it wear down fast after extended use because of the tooth grinding on the joint parts? that would be possible right? because under load it may have alot of friction but anyways i thought maybe you could invert the system so that the ball part has holes in it and that the others have the gears to reduce wear
I think you could swap its materials to stronger one like any kinds of tough metal, if you want it to be used as a joint for robots and be able to support weight when being used for lifting, I could think of attaching hydraulic piston support around each joints similar to vehicles that have one too
This looks like it would be fantastic for light-duty operations where maximum DoF is of primary importance. I'm less convinced that it is practical for anything with higher static or kinetic loading on the output link, since the "ball joint" (not really an accurate term here) seems to be held in place by nothing but the UHDPE socket. That is going to lead to some significant wear issues, and with them, a rapid degradation of accuracy and control. That may not be an insurmountable challenge, but it will need to be addressed if the plan is to subject the "ball" to greater loads.
Are there any good examples of the version with the single driving module? It feels like it should be easy to find, but I'm struggling to google the right things
What are the (dis)advantages of mounting the motors under an angle instead of opposite to each other? Does this give more flexibility in terms of rotating the ball? And does the ball have full 3 DoF from every point? Very impressive design!
