/robowaifu/ - DIY Robot Wives

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

Build Back Better

More updates on the way. -r

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)


Have a nice day, Anon!


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.
>>8700 Please keep us updated if you find this man makes a nice breakthrough that might help us out here. I like the lateral motion he's able to achieve thus far. Very life-like.
This hand design by a prosthetic company seems pretty useful: https://www.youtube.com/watch?v=B-f0egEgKms They changed the shape of the hand to make it easier to grip objects.
>>9690 Yes, looks like a decent, reasonably rugged design Anon. Thanks for bringing it to our attention. BTW, anon started a prosthetics thread here as well. Seems on-topic. >>417
Open file (1.79 MB 480x360 ohayou.webm)
I've been looking for attempts at playing piano with bionic hands. Someone made a rudimentary robot 11 years ago but they barely count as hands since they can only press the white keys. https://www.youtube.com/watch?v=IwuiOwZrzCM The closest thing I could find was this military prosthetic: https://www.youtube.com/watch?v=DP677lA_DEA And this robot typing on a keyboard which is pretty cool: https://www.youtube.com/watch?v=Nek_yvvLRdQ Are servos not fast or strong enough to make quick movements or is it more of a control problem? Could larger servos be stored in the chest and operate the hands with cables using pulleys?
>>10072 Bionic prosthetics normally don't have sensors. For building a robot playing piano you would probably want that. Btw, I just got my heat sensitve plastics for making hands skeletons like therobotstudio did (16€/kg). Here: https://youtu.be/bVDneXZ_YfQ - related (various vids): >>9362 >>9471 Some of the hands related vids: https://youtu.be/kWeJyduvhwA https://youtu.be/CfB8Mrh1dSI https://youtu.be/1PYMT1U8qiM https://youtu.be/tO3B_OyyqmQ https://youtu.be/tZ-ouSmt5gw
>>10072 >Could larger servos be stored in the chest and operate the hands with cables using pulleys? Not that anon, but you have a sound design idea there Anon. If we can figure out a way to entirely remove the weight and volume of servos out on the arms, then that will help everything move along better for us, physics-wise (and therefore, agility & energy-consumption wise). It will take some creativity to figure out how to do this, but using some form of cable-only approach in all the extremities will pay off in spades for us all.
>related (>>10179)
A big problem I am having with robotic hands is that the fingers can grip quite well, but they don't open back up again. When the hand is the 'relaxed' position, the fingers will remain partially closed unless I physically push each one back. One solution to this maybe to run taught elastic bands or small springs over the top of each finger, so they are pulled to an open position when the finger servos are unpowered. I believe many makers solve this issue simply by stringing their fishing line tendons REALLY tight (maybe with the help of self-locking forceps or a small clamp), so that the fingers are forced back into position when not gripping. However, this makes each hand very difficult to string. Which is why I want to give external elastic a try.(I have used an image of a paper and string hand here because I don't have access to my robot at the moment but also the visible tendons illustrate exactly the same problem).
Open file (24.72 KB 673x650 Elastic Robot Finger.png)
>>10999 Like so?
>>10999 This might not look very human, though. We have tendons on the back of our hand to open it. In your design approach the robot would need to spend energy to have the hand closed or keep it open.
>>10999 >>11000 I think many designers opt for exactly what you're describing, and simply weave a thin bungee-like or elastic band along the dorsal ridge of each finger. While this is a minor source of energy consumption during the finger contraction phase, it's obviously a passive benefit during the relaxation phase. And in fact it's probably fairly efficient energy-wise, as the normal state of hands tends to be in the relaxed pose. nice digits
>>11002 LOL nice digits for a post about digits. I have some experimenting to do!
>>11003 Thanks! >I have some experimenting to do! I'd suggest you have a look at the Nano Hand by The Robot Studio guy. He has an arrangement using elastic bands. Nano Hand - Finger Assembly https://youtu.be/kWeJyduvhwA Some other links anon posted here (>>10075). There's another quite sophisticated hand design video posted here somewhere. Can't locate it ATM for you SophieDev, but I'll plan to link it back here for you in the future when I stumble across it. Cheers.
>>11002 Streched out is not relaxed, though. >>11007 Here are some videos on hands by therobotstudio. Also look into the following postings. We probably should also link them in the other thread for the corresponding body parts.
>>11015 >Streched out is not relaxed, though. Agreed, ofc. By 'relaxed' I mean the passive position of the fingers/hands/etc. where the actuators are not actively driving the part. So, for the fingers that would mean 'not in the gripping position'. So, elastics can help 'passively' return fingers to this position, which is what I think SophieDev was talking about. >We probably should also link them in the other thread for the corresponding body parts. Yes good idea Anon. I'd say that having some kind of 'Anatomy Map' in the library thread might be good too.
Open file (57.05 KB 680x587 E2oTyDkVUAADm8v.jpg)
>>11016 Maybe the elastic band or spring could be adjusted by a very small motor. In case of a band it could even be twisted to make it shorter and pull back stronger or the other way around. But mainly I was thinking about ajusting the length, to define the baseline of where the fingers should be if releaxed. So in other words, if the hand would be supposed to be rather closed, then the little motor would give some extra 'rope' to prevent it from flipping back.
>>11017 I like the idea of having an actuator-controlled re-tensioning system Anon. Good thinking. >..if the hand would be supposed to be rather closed.. Ahh, that's the source of my confusion: we were operating on different definitions of 'relaxed'. In the interests of unambiguous & clear communications I suggest we adopt the related term used in the animation industry, namely Neutral Pose. 'Modified T-Pose' is the closest example to my meaning, and roughly-speaking represents the middle-position of a joint (as in animation-rig joint), roughly halfway between it's extreme positions. Make sense?
Open file (306.79 KB 450x482 ClipboardImage.png)
>>4863 What about just something like this for a placeholder 'doing stuff' hand? It's just an elastic balloon filled with coffee grounds that serves as a useful gripper. Doesn't look very humanoid, but it does a good job of doing 'hand tasks' I think having hands that serve different goals is a good point to develop for. Maybe even have a hot-swappable slot for the to switch to the right tool for the job. If you really wanted to polish the system you could even have a rack of hands somewhere in the house for the robot to go change its own equipment.
>>11146 Yep I like that idea Anon. Letting our robowaifus have multiple 'hands' to use would be pretty useful (and interesting). I don't think the majority of men would be particularly averse to the idea itself either.
>>11146 Also, breasts full of coffee grounds? Go to squeeze booba - booba squeezes you.
>>11153 Lel. In Soviet Russia...
Open file (402.63 KB 1152x2048 IMG_20210624_175649.jpg)
>>11029 >Modified T-Pose Okay, but humans don't act like that. If possible I'd like to avoid very awkward behaviors. That said, it could be a fallback position. >>11146 All of us have slightly different goals and priorities. I want something quite human like, so this isn't really for me.
>>11167 Nice. Alita looks great with her long hair Anon. I liked her before, but I like her even better now.
>>8651 Cheap but reasonably powerful (like MG996R that can produce up to 15kg starting torque at 6V) analog servos weight around 55g each. There are 15 servos on that arm (3 for each finger it seems). So servos alone would weight 825g. And that's only for fingers actuation! There supposed to be more moving parts in that part of the arm but as you can see on the vid there's barely any space left and construction itself doesn't look like it was designed for anything more than this one demo. There are smaller and lighter servos. And they cost a lot. I think they can weight something like 15g. That would make 15x15=225(g). Better but super expensive.Imagine paying 750-1500$ just for servos that move only fingers on only one hand. It seems like they are using something like that in their vid because servos are tiny.
>>11482 The mass of servos is exactly why I've been lobbying for relocating them into the 'central core' of the robowaifus, and transmitting force out to the extremities via Bowden cables instead. >tl;dr Reducing thrown-weight in extremities is crucial to both agility/performance, and (much more importantly) reducing energy consumption in our robowaifus.
Open file (307.67 KB 1024x1024 gallery-5-1.jpg)
Open file (299.27 KB 1024x1024 gallery-4.jpg)
Open file (288.84 KB 1024x1024 gallery-2.jpg)
>>11007 >>10075 therobotstudio has a new videos and a website about his Nano Hand online. Intro: https://youtu.be/uOeS_jklU2Y Website: www.robotnanohand.com Playlist with assembly: https://youtube.com/playlist?list=PLy7gxZH9jzfSGinQz8W42F5HdiTkT0Xm8
>During RSS 2021, my colleagues and I published a surprising new finding. It turns out, the softness of hands lets us manipulate some objects completely blind: without any visual/tactile/force feedback. We studied the nature of these open-loop skills, and identified three key design principles for robust in-hand manipulation. ... >Yet here, with this soft hand, a simple step adjustment to two actuators' inflations triggers this cool twirl. We do not compute anything. Instead, we just hitch a ride on the physics. This approach is super low-tech, and does something that advanced robot hands have a hard time doing. We published this in 2021. But this work could as well have been done 30 years ago. Air pumps and inflatable rubber are old things. https://aditya.bhatts.org/sensorless-in-hand-manipulation https://youtu.be/2vwdP4WjGoQ Paper, same as uploaded: http://www.roboticsproceedings.org/rss17/p089.pdf
Open file (137.64 KB 350x350 carlos.png)
>>13977 Neat, that guy is quite a talent. >>16898 This is interesting. I'm a firm believer in 'holistic-systems-by-design'. Clearly our hands in particular exhibit this kind of thing. It will be extraordinarily-gratifying to me personally once /robowaifu/'s anons can produce appealing & effective hands for our robowaifus that don't cost an arm & a leg.
https://www.youtube.com/watch?v=lGnSQbEzI9s Latest version of Clone's electro-hydraulic hand with a small form factor.
Open file (36.15 KB 336x188 mech_hand_play.jpg)
>Building a Robot Hand from Scratch >Why is it so hard? https://www.youtube.com/watch?v=OZ9g3O3X7dk I generally like the approach with putting small motors into the hand, though my thinking was to add additional one in the arm which could hold more than one finger but are stronger. The one in the hand are for fast and weak movements, the few motors in the arm for holding and more powerful moves. Also, our versions have to be much smaller than all these man hands, of course. Btw, Proto1 - Sophie's arm and fingers thread got never linked into this thread: >>6323 - now it is. Updates from therobotstudio on YouTube (focusing on hands and arms only): https://www.youtube.com/watch?v=hmhOAQ2yPsE https://www.youtube.com/watch?v=saTkGAdT8h8 https://www.youtube.com/watch?v=4bgoUk-Dcl8 It's called Dexhand: https://github.com/TheRobotStudio/Dexhand I'll probably wanted to go with his former variant (Nano Hand), if not something like the one from the first video above: https://github.com/TheRobotStudio/robot-nano-hand - but it's gone... Never mind, it's on Thingiverse: https://www.thingiverse.com/thing:3648120 - I got it. BTW, I'm not endorsing the guy in a general sense. He's as nuts as many nowadays. Fitting into the culture of madness. His main project is a full humanoid, but it's supposed to look exactly like something between man and female. Brad Pitt and Scarlett Johansson in one, dependent on the angle you look at. Good reminder why this board here exists. I would use the clown emoji if it was allowed here. Look for yourself if you can take it, I mean it's still somewhat interesting and I like that he's making a great open source humanoid robot for work: https://www.youtube.com/watch?v=S38Fnp4m6W0
>>7702 >Shadow Hand >What makes the human hand special, and why is it worth replicating in mechanical form? https://www.youtube.com/watch?v=fm7QUiRUAJA >mainly using it for remote control with human operators TL;DW Hands are important and difficult to build
>>19431 Thanks Anon! Once we master hands, we'll be very far down the path to success IMO.
The MERO Hand: A Mechanically Robust Anthropomorphic Hand using Rolling Contact Joint - doesn't really look like a new concept to me, maybe this was the first one, or I'm overlooking something: https://youtu.be/EA9mRS_-SC0 >This video presents the MERO hand, which is a MEchanically RObust anthropomorphic hand using novel COmpliant Rolling-contact Element (CORE) joints. The proposed CORE joint, which has a simple structure, exhibits compliance in multiple directions. Its structural parameters were designed, to form underactuated finger designs that can perform adaptive finger motion during grasping. Experiments showed that the hand could withstand severe disarticulation and violent impact. The hand could perform various adaptive grasps and also in-hand manipulation, suggesting that the proposed design might be a viable solution for robust prosthetic hand.
You guys seen this yet? Think this has a future? I've seen loads of similar projects both by college students, researchers in universities and companies but they all come to naught. I think this tech has atleast a decade more to mature. The power consumption of just moving a single hand with these artificial muscles is eye-watering.
Open file (9.02 MB 720x1280 clone_hand_demo.webm)
>>20643 >>20644 Thanks Anon. Looks quite interesting! >LOL those silly Polish Bronies :^) https ://www.clonerobotics.com/clone-hand >=== -add funpost
Edited last time by Chobitsu on 02/23/2023 (Thu) 14:25:50.
>>22710 it better be able jack me off too.
>>20643 Yes many or all of us have seen this. We have two whole threads, one on humanoid robot videos https://alogs.space/robowaifu/res/374.html .... and another on waifu development projects https://alogs.space/robowaifu/res/366.html, it has at least been mentioned in the first one since it's not a gynoid. Here their YouTube: https://youtube.com/@CloneRobotics - they had a different name a while ago (Automaton, I think) >The power consumption of just moving a single hand with these artificial muscles is eye-watering. Okay, I don't remember that part, tbh
>>20643 I disagree I think the tech is here right now and its a race against time to see who gets there first which is why I'm kind of semi panicky.
okay so I definetly want to start with the robot hand but which robot hand tutorial should I follow? which robot hand do we want on the waifu? Should I follow the tutorial or should someone engineer it from scratch? I don't think I can engineer it from scratch...
>>22785 I don't know what you mean by doing it from the scratch. Of course you would look into tutorials. Big problem with many hands is that they're not about bones plus soft material. But if you go for that, you will most likely need to make some elements out of metal.
Open file (40.23 KB 640x480 1703971123127274.webm)
Well i did make a hand so i'll chime in. I think the joins ought to be cylinders for more seamless movement. I also think most hands out there are too masculine. So I'll be modding/making a new hand i think.
>>29428 Does anyone know of a printable hand design that has an open channel on the interior of each finger joint for mounting sensors? I'm thinking of a removable pc board with contact/pressure sensors that could be swapped out as technology improves and better sensors become available. That way we could start with something as simple as circuit board push-buttons or contact switches (like are used in copiers and printers) and upgrade later without having to build a new hand design. The hand and sensors would be covered by a silicone glove. This way the "skin" doesn't have to be made with built in sensors. Separate subassemblies mean cheaper upgrades and repairs. Basically I want to future-proof, as much as possible, the work I will be doing soon.
>>27988 That's fine, I agree, but I think Chobitsu will ban you anyways. >>29719 I wanted to do this, but didn't get to it so far. I think you will have to go with a simple design first and replace it later. Making a really good hand will be really difficult and not everyone will need the same complexity. I'm convinced that it has to be made out of many small machined metal parts (maybe even Titanium, lol). Otherwise there's not enough space, and plastics would be too fragile if being so small. Also, ideally the lines for data and energy should not be cables, but the hinges should be made of metal parts insulated from each other. To picture it: Imagine the middle layer of each finger bone to be a PCB with lines, then an insulation on top and bottom, covered with metal, then maybe a TPU before the silicone rubber. The sensors would be in the TPU or silicone, but in the tips it might be little cameras ("gel sight") as part of the PCB and a little hole would allow to look at the soft material in the finger tip. The hinges more like very small "piano hinges" (aka "continuous hinges"), but with insulation between different parts, so different signals can be transmitted and received. Even if this would work, there would still be a problem with making sure the sensors wont be harmed or messed up by little cuts or interactions with the environments (needles, thorns).
>>18745 >Dexhand There is a site for it that didnt get listed https://www.dexhand.org/ I wonder if there is a way to make the hand smaller to a more feminine palm width. Also how to make it quieter aside from an outer casing.
>>30267 I recognized the use of mediapipe immediately: https://mediapipe-studio.webapps.google.com/home I encourage folks to check out the demos. Among other things it can do hand/body tracking, which means Real Steel -erm... I mean... robowaifus doing a synchronized dance with a human ain't so far fetched.
Not sure where to post this. When I was looking at Punyo which I had posted into the video thread >>30748 I had came across that Toyota has the gripper mechanism used is available for people to build. Not entirely humanoid just having soft pads with sensors instead of full on hands but it's workable and I think the same thing could be modified into a hand form either using it only in the palm then add fingers maybe even smaller versions in the pads into the fingers or something for more sensory ability. Just a thought. https://punyo.tech/
>>30750 Thanks, it's the right place, I'd say. Btw, if people aren't responding more with some encouraging comments, it's just to not "spam" the board with that. Doesn't mean it's not appreciated.

Report/Delete/Moderation Forms
Delete
Report