/robowaifu/ - DIY Robot Wives

Advancing robotics to a point where anime catgrill meidos in tiny miniskirts are a reality.

Disappearing posts should be fixed. Let me know if the issue persists on irc.rizon.net @ #julayworld.


Warrant canary has FINALLY been updated.


Roadmap: file restoration script within a few days, Final Solution alpha in a couple weeks.

Sorry for not being around for so long, will start getting back to it soon.

Max message length: 6144

Drag files to upload or
click here to select them

Maximum 5 files / Maximum size: 20.00 MB

More

(used to delete files and postings)


Hand Development Robowaifu Technician 07/28/2020 (Tue) 04:43:19 No.4577
Since we have no thread for hands, I'm now opening one. Aside the AI, it might be the most difficult thing to archive. For now, we could at least collect and discuss some ideas about it. There's Will Cogleys channel: https://www.youtube.com/c/WillCogley - he's on his way to build a motor driven biomimetic hand. It's for humans eventually, so not much space for sensors right now, which can't be wired to humans anyways. He knows a lot about hands and we might be able to learn from it, and build something (even much smaller) for our waifus. Redesign: https://youtu.be/-zqZ-izx-7w More: https://youtu.be/3pmj-ESVuoU Finger prototype: https://youtu.be/MxbX9iKGd6w CMC joint: https://youtu.be/DqGq5mnd_n4 I think the thread about sensoric skin >>242 is closely related to this topic, because it will be difficult to build a hand which also has good sensory input. We'll have to come up with some very small GelSight-like sensors. F3 hand (pneumatic) https://youtu.be/JPTnVLJH4SY https://youtu.be/j_8Pvzj-HdQ Festo hand (pneumatic) https://youtu.be/5e0F14IRxVc Thread >>417 is about Prosthetics, especially Open Prosthetics. This can be relevant to some degree. However, the constraints are different. We might have more space in the forearms, but we want marvelous sensors in the hands and have to connect them to the body. The thread about actuators is related: >>406 and the discussion in R&D General starting here >>1627 is a lot about artificial muscles. My own concept in my mind for the most ambitious models so far is the following: We have no space to waste. We might try to use light canals and LEDs inside to indicate bending. We'll probably need to use PCBs for transport of current incl data while acting as part of the bone. We'll probably need connectors between the parts which transport current without bending cables, eg cylinders with layers out of different materials where some of them are conductive. We should try to use Will Cogleys files as foundation, but since we might want to only build the bones out of hard material we might need to do it in metal to go smaller. This metal might be something expensive, like Titanium. We'll need tools for fine mechanics and such skills. I also think air pressure might be usefull to help open the hands fast, of course in combo with strings. I haven't thought about the actuators in the forearms a lot, by now. That's it for now, I'm really curious about the ideas for the premium models, but even more so for the cheap and then maybe even smaller waifus.
>>4577 Thank you OP, I was starting to wonder when this thread would pop up. I would add that there is another interesting hand with fingertip sensors: Hand in action https://vimeo.com/154571244 Fingertip sensor example https://vimeo.com/155743044 Paper https://homes.cs.washington.edu/~todorov/papers/XuICRA16.pdf This paper has some good information about more human design of an artificial hand and the robot is, in my opinion, one of the best. what I dislike about this model is the bulky actuators it requires, but in reality most of the muscles controlling the hand are in the upper forearm connected to the fingers themselves with tendons. This can be made to look more natural with different actuators, like a hydraulic or pneumatic actuator, but the muscles in the hand which are anchored to the wrist or to other finger bones could prove challenging. Another pretty decent design is this one: https://www.youtube.com/watch?v=gd9d_BAXWvg which follows a similar idea but is completely hydraulic. A shame that neither of them made their designs public, but there is still a fair amount to be learned from observation.
Here a example of mistakes to avoid: https://youtu.be/s0LA48Sw62k
>>4577 Thanks for creating this thread OP. Hands are an important and complex topic actually.
New Will Cogley video on hands. I had the plan to look up some joints for different part of the Waifu, how they work, differences, ... Didn't know where to start and delayed it. Now he comes along with his new vid and gives me the right direction: https://youtu.be/dWUSH6DR4G8 Search for Condyloid joint, lead me to more, not all related to hands though: https://youtu.be/zHYEhDMV5Bg https://youtu.be/S7nUC__PRFU https://youtu.be/0cYal_hitz4 https://youtu.be/2H10SZHGdvE
The Making of a 3D-Printed, Cable-Driven, Single-Model, Lightweight Humanoid Robotic Hand >Dexterity robotic hands can (Cummings, 1996) greatly enhance the functionality of humanoid robots, but the making of such hands with not only human-like appearance but also the capability of performing the natural movement of social robots is a challenging problem. The first challenge is to create the hand’s articulated structure and the second challenge is to actuate it to move like a human hand. A robotic hand for humanoid robot should look and behave human like. At the same time, it also needs to be light and cheap for widely used purposes. We start with studying the biomechanical features of a human hand and propose a simplified mechanical model of robotic hands, which can achieve the important local motions of the hand. Then, we use 3D modeling techniques to create a single interlocked hand model that integrates pin and ball joints to our hand model. Compared to other robotic hands, our design saves the time required for assembling and adjusting, which makes our robotic hand ready-to-use right after the 3D printing is completed. Finally, the actuation of the hand is realized by cables and motors. Based on this approach, we have designed a cost-effective, 3D printable, compact, and lightweight robotic hand. Our robotic hand weighs 150 g, has 15 joints, which are similar to a real human hand, and 6 Degree of Freedom (DOFs). It is actuated by only six small size actuators. The wrist connecting part is also integrated into the hand model and could be customized for different robots such as Nadine robot (Magnenat Thalmann et al., 2017). The compact servo bed can be hidden inside the Nadine robot’s sleeve and the whole robotic hand platform will not cause extra load to her arm as the total weight (150 g robotic hand and 162 g artificial skin) is almost the same as her previous unarticulated robotic hand which is 348 g. The paper also shows our test results with and without silicon artificial hand skin, and on Nadine robot. https://www.frontiersin.org/articles/10.3389/frobt.2017.00065/full
Open file (226.48 KB 1192x341 frobt-04-00065-g004.jpg)
>>4859 Thanks, Nadine is a horrible looking fembot, but the article on her's and other hand designs is very interesting. I think we will need placeholder hands at the beginning and then we might try different development approaches with different priorities, before we will be able to join all of the traits later in one design: - Aesthetical hands for gestures and human like movement, but weak and without many sensors. - Sensory hands, but less good looking and still weak. - Hands for doing stuff, with some sensors, but probably not good looking, especially not human female alike. Since we all have different priorities, tastes, and financial ressources, the end goal is IMHO going to be that we need to have a lot of options by variety in the designs. For future reference, I'd like to link to a discussion on Sophias hands: >>4826
>>4863 YW. I agree on both points, and I too found the paper interesting in general. It seemed a decent overview if nothing else. >3 variations of general hand designs Seems like a good breakdown. >Lots of options That's one thing /robowaifu/ has tended to have up till now, numerous different ideas. I'd like artistic expression (eg, dancing etc) for my initial types of robowaifus, so grip strength and rugged utility is less important to me atm. One thing I think unifies the significant majority of us and others is expense. We need to find ways to make robowaifu components very inexpensively. Sophie is a very interesting robowaifu tbh. I wish Anon great prosperity in completing her quickly! :^)
>>4817 Thanks Anon, finally getting around to watching these joints videos now.
At the beginning and for cheap versions of the Robowaifu we might need to use the InMoov design. People improving that design a little bit, without going crazy, might then be particularly interesting. Especially if the designs are available: https://youtu.be/CIqzeBxkRws The Raptor Reloaded Prosthetic Hand might also be a cheap starting point: https://youtu.be/ULTyu4ppk-s The other extreme - A very complex design: https://youtu.be/K0RUxY3RKKo Site: https://www.wevolver.com/wevolver.staff/anthropomorphic.robotic.finger/master/blob/Overview.wevolver Ada Hand: https://www.wevolver.com/wevolver.staff/ada.hand/master/tree Humanoid Robotic Hand: https://www.wevolver.com/wevolver.staff/humanoid.robotic.hand/master/tree A page with overview of hand designs (link from the Sophie thread): https://www.frontiersin.org/articles/10.3389/frobt.2017.00065/
>>4876 Thanks. I quite like the integrated design of #2, and am intrigued with the apparent tensegrity-inspired approach of #3. I'm currently working on a very inexpensive tensegrity approach (thanks to /hover/-Anon!) for the long-bones and so my mind is particularly attuned to that area atm.
>>4877 should've posted this along with.
>>4878 BTW, please not the quite generous direct 'shaft' of open volume available with this 4-rod tensegrity design. I intend to mount a central tube down this space as a guideway for running Bowden cables out to the extremities all the way from the proximal major ball & socket joints located at the torso.
>>4879 >please note* (check the shadow directly below the framework)
The guy improving the InMoov design I mentioned here >>4876 moved on in a similar direction than Will Cogley with his new hand, embedding little motors in the hand, using SLA printed pulleys. https://youtu.be/3DNVadMs5tE Though, it needs to be noted that they are doing it because they might want to build a prothesis, which has other constraints than a robot arm. We should look at it, but also keep in mind that we have other priorities.
>>4883 That looks like a pretty promising approach actually.
>>4884 Thinks so too, though we might be able to combine it with Bowden cables and servos in the wrist later on. Just checked, I can turn my similar little motor in both directions if there's no torque applied.
>>4885 I think I missed description of the motors themselves. Can you fill us in Anon?
Open file (613.19 KB 1618x1365 IMG_20200830_211848~2.jpg)
>>4887 Minute 2:34 and min 4. Micro metal geared motors: I've got the one on the pic for idk 1.50-3.50€, I guess around 2€. But I only bought one for testing and don't have the u-shaped one. He talks about SLS printed pulleys, but the one he has, can also be bought on Ali in sets with gears for 2€, some of mine are on the other pic.
>>4876 >>4863 some guy posted your board on g i'm just driving by here but I'll drop you a hint hyperboloids for the metacarpals
>>4892 >hyperboloids for the metacarpals heh thanks for the hint, driveby-kun. :^) https://en.wikipedia.org/wiki/Hyperboloid https://en.wikipedia.org/wiki/Metacarpal_bones My presumption is that Driveby-Anon intends us to understand a geometric behavior of the metacarpals. The tendons connect both above and below, and the particular curve they exhibit gives them a leverage advantage during gripping motion, for example.
>>4892 >>4893 Someone was claiming in this thread on /g/ the human hand was mostly made out of those and this was patented till 2040. Patents need to have a certain amount of innovation, though. Also, I'd need to look into these vids about joints if this true, and then it might still not be the best way.
>>4902 I see. Hmm, I don't think it's particularly problematic. There's far, far too much prior art in this arena (literally going back for millennia) to be in any way enforceable.
> Will Cogley Progress intensifies: https://youtu.be/99RzNBQF6g8 It's of course quite to big for now, but I like where this goes. > Skeletal prototype with windows to look into it, to see how well the gears are moving. > MG996R Servos from Tower Pro bought on AliExpress > Ultrahigh molecular weight Polyethylene, PTFE cube Github: https://github.com/ikkalebob/NM-bionic-hand/
>>4945 Thank you for keeping us updated on his progress. The man is a real talent at this type of engineering and is obviously quite motivated as well. Good stuff, Anon.
>>4891 Someone having an opinion on which amount of RPM I should pick, when I order some of those motors? I guesstimated for the thumb it might move 2-4cm on the joint, the shaft of the motor has 3mm or so, which makes 30mm per second with 10 RPM, which is 3cm. So I think I'll really stay on the absolute low end of the scale. Any objections?
>>4974 >So I think I'll really stay on the absolute low end of the scale. Any objections? It's simply a trade-off like all engineering is between velocity and torque. If you need more power, go slower. If you need more speed, expect less rotational force.
>>4975 Sure, I was already aware of that. Don't want the fingers move to slow, however. Then again, I can replace the motors later anyways, so I probably shouldn't overthink this. I'll take something around 10-15 RPM, but not 5, and not 30.
>>4976 Sounds good. As always, Test, Iterate, Improve.
>>4883 There's a new vid of his hand with the microgear motors: https://youtu.be/Zi-s_RQNokw Sadly he emphasizes that he thinks it's only possible to print the parts for the hand with a SLS printer, which is out of reach for most people. He's getting the parts from some paid service bc also hasn't one himself. As inspiration the approach might still be very relevant. I'm saying for a while hands are going to be very difficult.
5018 >SLS printer A guy built his own? https://www.instructables.com/id/DIY-SLS-3D-Printer/ Also, that's a high-end goal to exactly reproduce a prosthetic-capable hand. We'll be able to make due with far less refined designs for our robowaifus for years. It's a good goal to work towards, but we'll be able to create good (enough) hands much sooner.
>>5026 Fascinating, but it seems to be abandoned. Even with 5x5 cm this would be great, but it would need at least to work with PLA or better Nylon. He used tea with 95% sugar or alternatively Stearin, which is some kind of fat that melts at 48-50C. However, I'm quite sure we'll get around needing some SLS printer for the hands. He used it for potentiometers, same as Will Cogley in his model. These seem to fail after a while anyways, so other sensors might be better. Also, with some little changes it might not be necessary, and resin printers might be good enough, or even FDM and standard parts which we can buy.
>>5043 Well, maybe someone with some resources can set up a good SLS printer and provide prints as a service bureau. Maybe someone already does this?
>>5018 Thanks Anon. I like his design in general. Quite simple and to the point. I particularly like how he thought ahead about ease-of-assembly in at least a couple of different points, and overall I think it seems a fairly good design approach for the fingers. Very utilitarian and not too expensive by the looks of it. Did he give any cost estimates for the out-sourced parts printing?
>>5045 > Cost estimate No, or I haven't seen it yet. Costs might not be the problem, if it's already finished. But for development it might be relevant, especially considering possible additional shipping costs and the delay.
>>5043 >Also, with some little changes it might not be necessary, and resin printers might be good enough, or even FDM and standard parts which we can buy. Sounds interesting. Have anything specific in mind, Anon?
>>5047 No, or I don't remember something specific, but we can print small with FDM: https://youtu.be/gN7QMhBzd4E https://youtu.be/LHg9phNSCEY
>>5048 >0.15mm wow i didn't know there were even nozzles that small, looks almost like resin printing. looks like it takes some mods needed to make it work.
A normal human hand is the ultimate multi-use tool. It isn't that hard to replicate one specific feature of a human hand, what seems impossible is to cram the feature list into the small space of one hand (plus forearm). How about this: Picture an android with two hands that look the same, but are different underneath. The wish list of features is split between the hands in some way (for example, one hand can be strong & the other one precise). As long as both hands working together can achieve what you want them to achieve, this is almost as good as each hand having the full feature set. Heck, when doing this you might find yourself suddenly with some extra space left and throw in a magnet or suction-cup function.
>>5592 Good idea. Noted, but I hope we can do better. Btw, if no one answers to a posting, it doesn't mean no one is interested in it or won't appreciate it, just that we shouldn't flood the forum with praises to each others ideas.
potentially-related xposts >>699 >>704 >>705

Report/Delete/Moderation Forms
Delete
Report

Captcha (required for reports)

no cookies?