... >>8502 (related to actuators >>12810) > electically activated polymers, dielectric elastomers >>8504 (related to skin >>242) > aluminized mylar film or space blankets for electrodes >>8522 (related to AI >>77 >>22 >>250 >>27 >>201) > VHDL, Hardware Description Language, FPGAs >>8523 > millimeter wave length for power and communication to the muscles > salt water solution in conduits to carry power. (energy systems >>5080) >>8546 > FPGA / computing in elastomers >>8547 (related to actuators >>12810) > mini-hydraulic systems / elastomer motors >>8548 > automating the hardware >>8728 > lists of mechanical solutions >>8729 (related to actuators >>12810) > double-wishbone universal joint ... for transferring rotary motion >>8730 > more human-like design vs more robotic >>8955 (related to AI >>77 >>22 >>250 >>27 >>201) > ReZero, AI, machine learning >>8975 (related to actuators >>12810) > Muscles, Reflexes, and Locomotion, library links >>8976 (related to AI >>77 >>22 >>250 >>27 >>201, SBCs >>16) > XNOR.ai >>8977 (related to programming >>128 >>86) > NIM - lang >>9080 (related to humanoid projects >>374) > Glados (animatronic or robot design) >>9184 > make our own circuit boards, voltera.io >>9265 (related to AI >>77 >>22 >>250 >>27 >>201) > SQLite, serverless databases >>9277 (related to AI >>77 >>22 >>250 >>27 >>201) > ACID transaction functionality > filesystems and (generally) text files for my 'database' >>9319 (related to AI >>77 >>22 >>250 >>27 >>201) > torch.from_file, which read and writes tensors directly to files and works extremely fast >>9325 (related to decentralized AI training >>8958) > BOINC system, decentralized ML training >>9332 (related to AI >>77 >>22 >>250 >>27 >>201) > external servers at home (for AI) >>9958 (related to actuators >>12810) > planetary roller screws >>9984 (related to actuators >>12810) > ball screws > solenoid-actuated 'shuttle' >>16707 > crush ribs for inserting ball bearings into plastic parts >>16750 > scents men prefer relating to women >>16766 (related to thermal management >>234) > elf/fox/cat ears or wings as cooling structures >>16766 > structures could be used to help with balance >>16767 > wings with wheels >>17204 (related to skin >>242) > multifunctional flexible and stretchable graphite-silicone rubber >>17468 (related to business >>3119 >>1642 >>18572) > early models ... geared towards the elderly or PTSD victims > business / profit >>18088 (related to skin >>242) > self-healing light guides for dynamic sensing (SHeaLDS) >>18276 > easy to print 20202 t-slot compatible bar >>18688 > Uncanny Valley >>18744 (related to actuators >>12810) > Open Emag - Make your own electronecks mechanisms (to look at and take inspiration from) >>19009 > functional tattoos ... visible under a special light >>19340 (related to actuators >>12810) > DaVinci's knight >>19390 > micro-robotics, biomimicry >>19392 (related to actuators >>12810) > Simple Field Oriented Control (FOC) >>19398 > rollers as axis, shaft or pivot. Or in some cases to join parts together without needing to add screws and nuts >>19408 (related to actuators >>12810) > printed clamps to ensure a tight fit >>19442 > links to the topics in the MaidCom Development thread >>19513 (related to actuators >>12810) > making gears by gluing timing belts onto 3D prints >>20119 (related to skin >>242) > flexible display films ... as skin >>20122 (related to skin >>242) > thermochromic powders in the skin >>20123 (related to skin >>242) > LEDs shining through the skin >>20327 (related to skin >>242) > impregnation of textiles with silicone (skin related) >>20344 (related to skin >>242 and armatures >>200) > silk paper shell designs >>20775 (related to skin >>242) > custom (open source) microphones >>20789 (related to actuators >>12810) > 'Parallel-Spherical' joint mechanism > torso >>21494 (related to actuators >>12810) (related to skin >>242) > sensing angular rotation >>22032 (related to vision >>97) > 3D printed optical sensor framework >>22129 (related to skin >>242) > fuzzy skin (3D printing) >>22284 (related to thermal management >>234) > heart that pumps coolant using a nechanical pace maker >>22287 (related to actuators >>12810) > creative gears >>22438 (related to actuators >>12810) -TOOLS- > self winding system Things which are left over and belong into this thread are mostly mechanics, the topic of skin (without sensors), tools, general design questions, and things which are really special.

Literally the most comprehensive OP I've ever seen NoidoDev. Nice work!

A channel around building tools to build other things: https://www.youtube.com/@MrInnovative e.g. DIY Arduino based Wire harness wrapping machine: https://www.youtube.com/watch?v=s923C0_tkjw >>24155 Yeah, thanks, took me some time. And I still forgot the pictures in the second posting. But okay.

>>24157 >e.g. DIY Arduino based Wire harness wrapping machine: If any anon finds any information about DIY actuator wire-wrapping machines, it would be highly-appreciated if you'd post it ITT! :^) >And I still forgot the pictures in the second posting. Oh, I think the OP is just right. And now you've got some add'l resources available for thread #3. BTW, R&D General has always been one of my favorite threads going all the way to back in the day. Thanks again for reviving it! Forward!

>>24158 Did you look into this? >>18744 > Open Emag - Make your own electronics mechanisms (to look at and take inspiration from) I wrote "Make your own electronecks", lol. That MrInnovative channel I just mentioned >>24157 also should have something like that, I didn't look through all the videos yet.

>>24160 >Did you look into this? I 'looked into it' insofar as reasoning my own OC regarding the processes involved, and I think I have settled on a rational basis for it. Unfortunately I don't have the time currently to begin exploring the practical prototyping of this work. If I receive the funding I'm going for r/n, this is near the very top of the stack -- since actuation is effectively fundamental to robotics entire! :^) >tl;dr We need to 'wind our own', if we're ever going to have thousands of customizable electrical actuator systems. Regardless, for our Model A Robowaifu designs, any such actuator-construction systems will be opensauce, and freely-given to the world, such as they are. Hopefully hundreds of competitors can soon be producing custom actuators and innovation will explode in this area.

Thoughts on 3d printing the skin with TPE filaments?

>>24163 Works well enough, be mindful that you must print very slowly with retraction settings of at least 2 mm. Internal details will be inaccurate. TPU is extremely forgiving as out bends and compresses, you still need to be careful and mindful of its properties. Pixel is printed too fast and at 1 meter scale.

>>24164 I think TPE is not the same as TPU. I'm hoping its less hard and more flexible than what you showed on the picture.

>>24165 I'm pretty sure that Grommet & other anons here have discussed TPE filament before (eg, >>23370, ...), so I'm confident you can get some good feedback ITT. Please be patient, we're a slow board at this time. In the meantime, can you share what ideas you have so far about your robowaifu's skin, Anon? For instance, I presume you want many areas that are soft since you want to print in TPE. But does everything need to be so, or would you be satisfied if she has other areas that are more firm or even hard? Regardless, looking forward to seeing your progress with your robowaifu Anon! Cheers. >=== -prose edit

>>24171 Yes the boobs and butt would probably have to be latex filled with silicon jelly probably. Also the fact that the 3d printer is limited to a certain size means everything needs to be put together in pieces and the skin would would have to be glued piece by piece if its 3d printed.

>>24172 On second thought never mind the latex. I'm of the idea of taking as many shortcuts as possible.

>>24172 >Yes the boobs and butt would probably have to be latex filled with silicon jelly probably. Makes sense. One of the ironic twists to all this is that in their pursuit of beauty, 3DPDs have financed the R&D of medical-grade, fake bewb prosthetics which we here will all benefit from the use of heh. :^) >Also the fact that the 3d printer is limited to a certain size means everything needs to be put together in pieces and the skin would would have to be glued piece by piece if its 3d printed. Yes, that's a complex problem. In my own case, I want to go for commercial production, so my primary thinking is to use something like big rolls of spandex textiles impregnated with some type of near-silicone-like material that can be a) washed, b) cut, stitched & fashioned together into a type of 'custom bodysuit', which can then be c) stretched over and onto the robowaifu's external exoskeleton/shell and snugged down tight to fit. Being forced to use only smol consumer-oriented 3D printers is a much more challenging design requirement than mine is, IMO. >=== -prose edit

>>24173 I just thought of something else. There are TPE dolls out there that have no exoskeleton really. The skin and the exo skeleton are one in the same if we go that route. >>24174 Yeah I thought about the commercial aspect as well but there's just so much friction. The patent itself is a huge hassle and you have to be careful about borrowing stuff and make everything from scratch so it can be patented. I'll just share the stuff and hope for some collabotion I guess. Like the guy who made reprap and didn't profit from makerbot someone is going to have to take one for the team. There is also the fact that I don't really have any connections. Might make a video about that.

>>24175 >The skin and the exo skeleton are one in the same if we go that route. As mentioned in the 5 principles thread (>>24047), economizing & combining things together is usually a good idea economically, when it's feasible.

>>24152 OMG, I tried to get ChatGPT to sort this stuff by topic and now I'm full of anger. What a piece of garbage.

>>24174 It would be simple to take a women's unitard and impregnate it with silicone. From there you could zip her skin on and off as needed. Just need to tailor the robot to fit. Fast, cheap, and you could make some parts touch sensitive by mixing in conductive powder.

>>24171 >TPE filament and here's TPU. I must say this plastic ranger guy is like the best 100% most excellent guy on the whole entire planet on general knowledge of plastics. He writes very informative general purpose articles on plastics. https://plasticranger.com/what-is-tpu-material/ and to have them together https://plasticranger.com/what-is-tpe-material/

>>24152 This really nice. Thanks.

Oops, my apologies. Somehow got on the end of this threads part 1 and commented. >>24214 It was about fiber optic sensors which is brilliant, here, >>22032

Using fiber optics for pressure sensors I "think" you could also use them for position sensors. Look at this article on gray's code. It's a way of making electrical signals and optical sensors more accurate. It allows low resolution and noisy signals to be determined accurately. https://en.wikipedia.org/wiki/Gray_code The picture you get through fiber is distorted but...we don't care. It may be crap but as long as you program whatever comes back to be whatever position it doesn't matter if it's not optically correct. Only that the pattern changes with movement. So one fiber carrying the light from a paper picture connected to the rotating appendage fed through fiber to a small section of the camera that reads it could do almost instant position sensor determination. This fiber optic sensor thing has got to be one of the best ideas ever that someone thought up.

I had an idea for a super cheap easy way to make electric coils. Especially for actuators. Use Low temperature metal alloys. This is after all what solders are basically. There's all sorts of them at all sorts of temperatures. You could use normal 3D printer filament to print the channels. Which would become wires when the metal is melted into the printed form. You could print PVA, melt the metal into it, dissolve the PVA with water, then mix ferrite with liquid plastic, pour into the dissolved PVA spaces and now you have a custom transformer and you're not stuck with traditional transformer shapes. The beauty of this is most of these metals you can melt over and over. If you get high enough temperature metal you could use them as actual conductors on finished products and have custom solenoid coils to fit where you need. Chobitsu was asking about coil winders but here's a way to avoid them altogether. After I had this idea I found someone was doing something sorta like it. They were making inductors in silicon, melting the metal and then stretching the silicon to change the inductance, then solidifying that value inductance by letting it solidify. Abstract https://www.semanticscholar.org/paper/Remoldable-inductors-based-on-self-heating-fusible-Lazarus-Bedair/b481fe3c8dd2e44b4f6a5e8b47fd57adef61a6e4 BTW here's a book on building a "Universal Coil Winding Machine" http://libgen.rs/book/index.php?md5=1DCAE7F26AA7CCF6D67810D54F3A619E

>>24216 This is actually a pretty fascinating idea Grommet. And yes, I think that as long as all the 'sensors' have the same basic dynamic range of signal, that the deflection off the registered baseline should be fully effective as a 'change-in-position' sensor array. There will be a lot of software maths involved, but nothing particularly difficult in general. As an added bonus, it should calculable on GPU or FPGA -style processors really quickly. >>24218 Can't say I'm really super-clear yet on your idea here Anon tbh. BTW, thanks for the book link. I also found a couple of other related texts that are applicable to our interests here. Cheers.

>>24188 I hope you feel a little better now NoidoDev. Thanks for all your hard work! :^) >>24197 >Fast, cheap, and you could make some parts touch sensitive by mixing in conductive powder. Thanks Kiwi, great idea!

>>24165 TPU is a TPE. I think no one in the printing channels I watch uses the term TPE.

>>24218 >>24221 >[low melting metal windings]Can't say I'm really super-clear yet on your idea here Anon I can more easily explain with a picture The first is a regular motor winding, the second a linear motor winding. I'm only using the pictures as a "general" illustrative purpose. Many different architectures are possible. As you can imagine winding these tiny ass coils if your actuator motor is say one centimeter long or around, [eye lid motor as an example], would be difficult. Just getting the wire to wind it with would likely be difficult. But you could 3d print the form to "hold" the winding, then melt in low temperature metal winding. The low temperature metal "is" the conductor, "is" your coil. The low temp. metals melting point is below the melting point of the filament you printed. So you heat the form and the metal and pour it in or just submerge it into the metal to fill. Searching...here is the first "iron" filament I came up with. It's ferromagnetic, so that's the field concentrating part of the motor. https://www.proto-pasta.com/products/magnetic-iron-pla Other methods to add the ferromagnetic part is to print in PLA, melt in metal conductors, wash out PLA then fill spaces left with ferromagnetic powder and then add plastic, like super glue, to hold the particles together. Even more important than these typical shapes are the odd shapes you could come up with. You could 3D print little stepper motors, all sorts of magnetically operated solenoids, switches, etc. in all sorts of shapes. Shapes that would be impossible with typical wound metal coil motors. https://en.wikipedia.org/wiki/Fusible_alloy https://www.xometry.com/resources/3d-printing/magnetic-iron-pla-3d-printing-filament/

>>24216 >Using fiber optics for pressure sensors I "think" you could also use them for position sensors. I remember now: My idea once was to communicate with light, through the body, but not using uninterrupted fibers. Think about the problem of cables in moving parts breaking over time. We should use a hinge out of metal to transmit electricity, then using light in different colors and frequencies to send the data from joint to joint.

>>24218 >super cheap easy way to make electric coils. Are regular cables too expensive? >There's all sorts of them at all sorts of temperatures. You could use normal 3D printer filament to print the channels. Which would become wires when the metal is melted into the printed form. Creativity has it's limits. >They were making inductors in silicon, melting the metal and then stretching the silicon to change the inductance, I hope I mentioned something like this before and linked the video somewhere. Wrapping a wire around a silicone part and stretching it, should do the same trick. Some female scientist invented or discovered that a few years ago.

>>24229 > fibers. Think about the problem of cables in moving parts breaking over time I'm not seeing that. Plastic fibers are really flexible. True some Chinese brands might be cheap junk so you would have to know what type plastic fiber optic cable you bought. What are we talking about? Breaks after 10 years??? Maybe longer. No. The utility and super cheap cost outweighs other solutions by more than a order of magnitude. I looked up the cost of plastic fiber optics and saw you could get a hundred meters for under $20 USD! Better fiber can't be too much more. Wires, well wires have plastic coating so...no better. Where the utility lies is in the interface. If you use capacitive, magnetic, any other sort of electrical type sensor then you have to add all sorts of parts to the input of a microcontroller. I've been looking at this and MC interfaces are complicated and they cost. You need another board, a bunch of resistors, capacitors (usually) and buffer IC circuits to make this work. You will have to have this for the actuators but adding this complication to the sensors is a whole mass of extra work. Just compensating for stray capacitance,[if you use capacitive sensors], in the pressure senors will be a total pain in the ass and will require constant calibration. Look it up you will see this is true. A lot of design work, more PC boards to make for the interface components. Lots of complication. Lots of cost. But fibers from all the joints, point it at a $9 ESP32 with camera chip and you are done. A thought experiment. Don't parse this too tight. It's just a back of the envelope calculation. Let's say you have 300 actuators of various sorts to be sensed for location. Most you could have 8 bits resolution or 256 units of movement. A few you would need 10 bits. Major limbs. To get 10 bits, I used arm movement. So call the arm movement 2' distance or 609.6mm. Now I say half that to 1218 units of distance so now you have a resolution of 1/2 mm over the whole travel of 609.6mm. Close enough is 10 bits resolution at 1024 units distance for the whole travel of the arm, leg, thigh, forearm. Since the waifu is likely to be smaller this would likely be good enough for 1/2 mm or less resolution in limb, appendage, whatever movement location. Now I speculate that you could use a gray code wheel picture. The one I uploaded has only three bits. You need more depending on which joint it is. One fiber lights up the disc, a lens pics up the picture, sends it through the fiber and then it's displayed onto the camera. A ESP32 CAM Camera Module has 1622×1200 pixels or sensor channels. That's 1,946,400 total. Even if you cut that by 4x for alignment troubles it's still 486,600. In fact to get 10-bit resolution on all 300 muscles for a full function waifu you only need 3,000 pixels. So you're way ahead here. There's a possible glitch in using one fiber to carry the whole picture. "If" moving the fiber screws up the picture, even if the ends are fixed, or changes it as the waifu moves then you can't use one fiber. But that's not really a problem, as we just use one fiber for light and ten fibers to carry the light reflection from the gray code wheel. We have plenty of pixels in the camera to pick this up. Even better. This would be so fast to put together. You make a standard bracket to hold the wheel and place the fibers. The brackets, wheel holders, etc. could all be the same. You place the fibers in and snap closed. Each limb will have different length fibers. So you do just like car manufacturers make wire looms for cars. They have a big board with channels. They thread the wires, in our case fibers, then cut them to length. Add on connectors and wrap and you have the whole sensor system in one nice package. And even better we don't care what order the fibers for each individual sensor are in. After the waifu is built you give it commands to move one actuator at a time. The various fibers will light up as it moves and the processor then makes a map of each location for each fiber. Saves you a lot of time individually tracing each fiber from one fixed location to another, No need. It calibrates itself in software. That guy that thought up fibergrids, pure genius. One of the best ideas I've seen in my life. https://hackaday.io/project/167317-fibergrid#menu-details

>>24221 >deflection off the registered baseline should be fully effective as a 'change-in-position' sensor array. There will be a lot of software maths involved I'm shit for programming but I think the math would be simple, and fast. For movement of the limbs. You have two arrays of data. You need,roughly, 300 actuator positions at 8 bits each. So 2,400 bytes. Next to nothing. Let's say you have 3,000 sensor points, 24,000 bytes. So you make one array that holds and one that is made of constant poling from the sensor camera. Each 37,000 bytes. Compare the one held to the newest and if it changed, it moved or a sensor was triggered. Such a small array of numbers compared has got to be near instantaneous even with a simple microcontroller. The actuator set is only 2,400 bytes. So it could monitor what it "thinks" the movement should be with another set of arrays that is compared to the actual close to real time read sensor value. By using only on-off type values for the camera and maybe less resolution surely even the ESP-32 could do 1/100 second tracking with a little computing time left over. Maybe a lot of computing time left over. https://lastminuteengineers.com/getting-started-with-esp32-cam/

>>24233 Brilliant concept. Thanks for the Hackaday link, Anon. Helps clarify the concepts. --- >update: I had a quick glance at the code, too. https://github.com/rand3289/FiberGrid/tree/master This is mostly good modern C++. This style code is usually more comprehensible (and therefore more maintainable), and in many cases also runs somewhat faster than older C++ or C counterparts do. GG. >=== -add 'update' cmnt

>>24228 I'm still having a hard time imagining how the molten alloy will remain as separate 'wires' in your concept. Won't they all just flow together into a single block of alloy? If so, then the idea of 'how many windings a coil has' disappears wouldn't it?

>>24234 Fair enough, it may be simpler than I imagine at first glance. Thanks Grommet! :^)

>>24230 >Are regular cables too expensive? No. The main point is the versatility of manufacturing complicated coils, which means motors of all types, solenoids, actuators, etc. , they all have coils, without building winding machines or not being able to build some complicated coil at all. Resistance will likely be higher in the low melt part but, so what. How much more than copper or Aluminium, I don't know but would surprised if it were over 10%. If 10% then you just make the coil 10% bigger to carry the same current. You would also likely have to make it bigger to handle heat so it doesn't melt. And even if it does, if it's contained, it will not drain out and will harden right back in it's channel when it cools. After I wrote the above I decided to look for the resistance. I find a chart of different low melt solder alloys and the highest compared to copper was 5%. https://concorde-electronics.com/blog/low-temperature-alloy-solder/

>>24236 >hard time imagining how the molten alloy will remain as separate 'wires' in your concept Imagine a solid block of 3D printed material. Now imagine a coil "negative", not printed inside the block. The negative, not printed, part is put in [merged in the CAD]the block and printed. So now you have a solid block with a "not printed", empty, tube in it. The NOT printed tube is where you pour the melted low temp. alloy in to, when it hardens, makes the coil which is a wire of hardened low temperature alloy. Remember most of these low melt alloys have a far lower melting point than the plastic so the plastic is just a mold. Note the metal hardens so it's a wire when cooled, if shaped like a wire. Going further to optimize it, this is assuming you do not print ferromagnetic packed plastic filament which would do this in one step. Let's say you do the above but print in dissolvable PVA, pour in the low melt metal coil, melt out the PVA with water then surround the coil with ferromagnetic material. Just so you understand ferromagnetic material. It's iron oxide [there are others but it's commonly iron oxide] and attracted to magnets or electromagnets. If you dig in the ground and separate the soil. Dirt. Then take a magnet and sift through the dirt you will get a bunch of iron oxide [ferromagnetic material] stuck to the magnet. So you can do this, wash it in water then use the magnet again and get something to work with from regular soil, dirt. Why use ferromagnetic material? The ferromagnetic material bends all the magnetic force towards the coils and so strengthens the magnetic field. It's why they have ferrites or steel sheets around solenoids, motor, transformer, etc. It confines the magnetic field and makes it stronger "where" you want it instead of it spreading out around the motor, solenoid, whatever. If you are channeling water you need a tube to hold it. Ferrites act somewhat like a tube channeling magnetic fields. Here's something that will blow your mind, or it should it does mine. Watch this short video. https://www.youtube.com/watch?v=eWSAcMoxITw So this business of ferrites and magnetic fields is NOT simple. It's complicated. I've seen no explanation for the video above. Maybe some physics guy will throw a bunch of equations at you but they really do not explain what is going on. It appears to me the field in the metal acts like an analogy to a superconducting coil except it's superconducting magnetic fields instead of electrons and that somehow holds the metal blocks together. And this is not fake it's real.

I printed this comment in the wrong place, Oops, but it belongs here so I'll repeat it. You could use a cheap ESP32 CAM Camera Module Kit and monitor fiber. Do them in a sort of X'Y' grid and then you have what part of a finger, or any other appendage, and how far up. You could use this for position sensors also. Have a rotating bump or wheel press on an array of fibers. You could have 10-bit resolution with 10 fibers that are pressed by bumps or wheels that press on the fibers as the actuator rotates or moves. Just like pressure changes the fiber light when used for pressure sensors. In fact this may be even easier and cheaper than making gray code wheels. So with a little work you could have every single touch sensor and position sensor in one $10 camera. Some ideas are so good they are just...stupendous and this is one. The guy that thought of this was really doing some top notch thinking.

>>24235 I was overenthusiastic about this. Great for position sensors, still very good, but not for pressure or touch sensors. I think I got carried away because it solved SO many problems that I became caught up in wishful thinking. Another reason I think the pressure part worked was I have actually heard of this and thought he had found some simple way to do it. He started out stating that he felt he could which I missed and read that he had. Sigh. The idea still has merit but it's not solved.

Maybe all is not lost. Maybe this guy could not make fiber pressure sensors work but others are sure working on it. Abstract Low Cost Plastic Optical Fiber Pressure Sensor Embedded in Mattress for Vital Signal Monitoring https://www.mdpi.com/1424-8220/17/12/2900

>>24246 Reading further,"...The POFs have been sensitized to pressure, cutting part of the jacket, the cladding and the core (Figure 1a). A fiber holder with a hemicylindrical groove with the same diameter of the fiber was used together with a razor blade to obtain cuts with approximately the same depth. As described in Figure 1b, any applied pressure on the cut point changes the light intensity at the fiber end..." So this can be done and without too much difficulty. The paper.

Found another paper. A much better one. Seems there's not much commercial fiber that has the right properties. These guys found a polymer, "Geniomer 200 (Wacker AG, Germany), a polysiloxane-urea-copolymer with a polysiloxane content of > 90%. The polymer was used as supplied with no further treatment. The melt flow index (MFI) has measured 17.78 ± 0.52 g/10min (n = 3) at 1 kg and 170°C (Wacker datasheet: 170°C, 2.16 kg, 110-150 g/10min), and a melting temperature (Tm) of 171.31..." So tentatively it looks very easy to 3D print with normal 3D printers after making into filament. They then dipped in a refractive index polymer, "...A low refractive index coating material (THV 220, n = 1.36; Dyneon, Germany) was later applied by dip coating the fibers in 5 wt% acetone solution (23°C). The coating thickness was calculated to 16 ± 3 µm by gravimetric analysis (n = 3). After the dip coating with THV polymer they lost their sticky behavior...." The response in these, if they are not lying, is very good. 0-40 Newtons pressure. So fiber optic sensors not outmore complicated than just weaving a few off of amazon. It may well be that you could still use some cheap amazon fibers but...likely too stiff.

>>24242 Thanks for the instruction Grommet, that's helpful! :^) >>24243 >>24245 >>24246 >>24247 >>24257 Fiber has a metric boatload of advantages within a robowaifu. I'm sure they will play many important roles for us in the future, including involvement with sensors.

>>24246 >but others are sure working on it. Yeah, it would be surprise if no one would. It's too obvious of a good idea. In theory it should be sufficient to have a little ball joint, where the ball can move in different directions, holding a fiber. Then find a way to add changing light into the fiber. Maybe the ball could be somewhat transparent, or some parts transparent others not. Then LEDs with different colors would shine onto it. Bending the fiber makes the ball move into another angle, changing the light spectrum. >>24247 >>24257 Thanks, I'll look deeper into it when I'm at a point actually needing something like this.

Imagine an entire robowaifu shell of this stuff, under a soft, transparent skin.

>>24227 I'm going by what bard says and it says that TPE is 20% softer than TPU and cheaper. I'm going with TPE because it seems like the path of least resistance to me.

>>24227 You are correct. TPU (thermoplastic polyurethane) is indeed a TPE (thermoplastic elastomer) a type of copolymer which mixes a plastic and rubber to achieve thermoplasticity and a rubbers softness/malleability. What really matters is the shore hardness of the filament. Picrel is a chart that can be helpful for getting the gist. Note how there's multiple scales, it's important to note the letter when checking the scale. I personally use 98A TPU which feels somwhere between skin and leather in terms of flexibility and apparent softness.

>>24268 >Shore-Hardness-Chart-01 Great chart, thanks.Very useful

I think the elephant in the room is the legs. We'd want to be able to spread them for the sexy time too. How to go about this.

>>24287 I think you need look no further than the ball-and-socket design that God created for the hip joint. Artists like Sukabu do a rather good job of rendering this basic approach into a robowaifu-centric set of studies.

I can't get my 3d printer to work and I've been thinking. Wouldn't slush casting be better anyways? Although 3d printing could aid slush casting by making a 3d printed mold though.

>>24297 >Although 3d printing could aid slush casting by making a 3d printed mold though. I think several anons would probably agree with you on this point Anon, good thinking!

>>24293 This thing is never going to get solved unless you guys get on krita or ms paint. Here is my attempt.

>>24300 Did you use one of those shuffling robots that pretend to walk on ratcheting wheels as inspiration? >Joints I'm still going to be using Revoltech inspired joint system. Currently experiencing with methods of implementation.

>>24298 For everyone who wants to get into casting, I highly recommend Smooth-On silicone to start out with. This channel is great for beginners, many tutorials that are clearly articulated. Always remember to use a release agent too. https://m.youtube.com/@RobertTolone

>>24302 what about the top though? What makes the sphere at the top move?

>>24304 An actuator of course. It's a sphere with two attachment points that could be rotated. I would recommend looking into articulated figures and robots walking on video sites. This website is a good resource for researching robotic motion in general. http://www.robot-welding.com/robots.htm

>>24287 Related (walking): >>243 I shared an partial sketch of my idea about this here: >>23962 That said, I didn't think rigorous enough about every situation. Which is really something we should be doing in regards to every movement of the body. Technically the problem is close to back-driveability. A outside force to the legs should make them move, but they also need to function with actuators. If it's just about security, as in getting out of the way, then some spring or rubber string mechanism that can overwrite the current state would be good enough. It would snap back after the outside force is gone. I don't know if my idea is going to work, but I want to use strings, springs, keep the servos back-driveable and also adjust automatically, if the situation indicates that the new position of the legs is meant to be the new state. General remark: Female walking is closer to how models walk, and we at least want that as an option. The legs do not just move forward from the side of the body. I also plan to take the hip moving while walking into account.

>>24300 While i drew that quickly I think it'd fit the design pretty well. The vertical gear with two side gear arrangment is called a bevel gear train. Each side would be offset by 180 degrees, so while one side is up the other one is down. The wheels could server for the robot to turn to different angles. So the waifu rotates say 45 degrees with the wheels and then walks without the wheels. A regular dc motor with high torque could be used for this. I'm not sure but I think regular dc motors are cheap than stepper or servo motors and never mind actuators. Slap in a couple of pneumatic cylinders to spread the legs and you got something that's good enough. The pneumatic cylinders would be used to limit the spreading and would not need to be powered by anything.

>>24300 >>24309 Thanks for the work Anon. Good luck. >>24302 >>24306 Neat. Please provide more Revoltech information if you will Anon. >>24303 > I highly recommend Smooth-On silicone to start out with. This. It's quite expensive, but I've seen it demo'd with flawless results multiple times. >>24308 >Female walking is closer to how models walk, and we at least want that as an option. The legs do not just move forward from the side of the body. I also plan to take the hip moving while walking into account. Great, biomemetic thinking NoidoDev. But this is a rather difficult set of kinematics problems for us to solve, heh. :^)

>>24309 >gear arrangment is called a bevel gear train Reminds me of this: https://youtu.be/W6xzBVVuml8 - I thought about finding a way to make this work without drilling into metal. >regular dc motor Might be a good idea if you don't plan to have an enclosure which you can't remove later. I think dc motors have coals which will need to be replaced after a while. >The pneumatic cylinders would be used to limit the spreading and would not need to be powered by anything. Excellent idea. We should also keep HRP-4 approach as option in mind (picrel 3).

>>24316 I found this toy laying around the house. I've opened it before to put a resistor in it to make it quieter. It also comes with a motor. It might be good enough to make a test for the legs on a small scale. I'll also be going to the hardware store tomorrow so I might be able to get some other stuff as well.

>>24319 Oh forgot name handle

>>24297 >3d printing could aid slush casting by making a 3d printed mold That's definitely my way of thinking also.

Stringing piece, 3D printed: https://youtube.com/shorts/EzQtoJlunwM

Hannah from @DaveMakes got additional emotions to change the skin around her eye cavity when she moves her eyes up and down. Never seen that anywhere before: https://www.youtube.com/shorts/lbZZIK33T1I

>>24749 Very pretty. I'd expect that Kiwi and others can utilize similar approaches for winding tensegrity suspensions, as well. :^) >>24750 He's made lots of progress over the year or so. Thanks for the update, Anon! Cheers. :^)

>What's the difference between a CPU and GPU? And what the heck is a TPU, DPU, or QPU? Learn the how computers actually compute things in this quick lesson. https://youtu.be/r5NQecwZs1A >Curious exactly what happens when you run a program on your computer? Read this article to learn how multiprocessing works, what system calls really are, how computers manage memory with hardware interrupts, and how Linux loads executables: cpu.land

>>24749 Does anyone know of a technique to automatically generate those strings? It should be possible, Cura does similar with infill patterns but, none that I know of can replicate this. If I could edit the "lines" infill pattern to have layer between, it would be close enough, Would be very useful for print in place tensegrity.

>>24950 Didn't he explain it in the video or in the comments?

>>24950 >Would be very useful for print in place tensegrity. Ahh, the old 'necessity is the mother of invention' driving-force of nature, ehh? :^) Very innovative thinking Anon. Godspeed.

>>24952 I looked, it actually isn't. I could try in OpenScad... <the rabbit hole teases you, some call it they Abyss>

>>24749 and >>24994 Related: >>24996

Wraith Madoka's true form has a fascinating design that looks similar to a tensegrity structure. This concept could be modified into a cute and functional robot

Four bar mechanisms can implement almost any desired motion in very simple and elegant designs. Hackaday has an insightful starter. https://hackaday.com/2017/03/29/marvelous-mechanisms-the-ubiquitous-four-bar-linkage/ Wiki; https://en.wikipedia.org/wiki/Four-bar_linkage Valuable deep dive; https://dynref.engr.illinois.edu/aml.html Here's a video; https://www.youtube.com/watch?v=XvYNqj4awM8

Are there any resources for robot spines?

>>25915 >>200

>>25824 Kiwi's link was very good and has a lot of links that link from it that are very comprehensive. One I found might be useful if not too academic.(most of it is not directly related to robowaifus but some good data points are in it relating to power needed, etc.) Animal as a Machine and a Prime Motor and the Laws of Energetics https://ecommons.cornell.edu/items/51558415-c151-49a3-8a0f-3dcc27a1e2e9 Large set of books on kinematics at the same site https://ecommons.cornell.edu/collections/78104a5e-7f84-41ea-929a-1fc694f75ba2

Hypoid vs. Worm Gears: A More Cost Effective Right-Angle Reducer: https://www.orientalmotor.com/ac-motors-gear-motors/technology/hypoid-vs-worm-gears.html

There are some STEMMA-QT connectors which make connecting a Raspi to I2P easier, there are even hubs for this. https://www.youtu.be/-d0rOHjntAo >Yes you CAN install STEMMA-QT / qwiic connectors on a Raspberry Pi! Several popular alternatives are shown, including hats, a shim, and directly wiring a port to the board. Skip breadboarding for I2C components, and build faster, more fool-proof projects.

Does she need a bellybutton? Making a more feminine design for clothing.

>>26330 Looking good Kiwi! No, I wouldn't think she particularly needs a bellybutton (particularly during this early prototype phase). I suggest doing things that simplify printing/assembly of the mesh flats at this stage tbh. BTW, it strikes me that her Lats are a bit too masculine, but admittedly it's hard to tell w/o a 360' turntable animation. Keep up the good work, Anon! :^) >=== -sp edit

>>26330 A question which comes to my mind is: Why are you still working on a human-like body design since now we know that Make Human works for that?

>>26331 I feel the same way, I favor a very feminine and petite frame. I'll try to make her lats softer. Optimizing for print-ability while having her able to wear clothes is a delicate balancing act. >>26332 I prefer a more stylized body than I can get from MakeHuman. I fully support anyone using it. I agree that most should use it instead of investing time to make their own mesh. I'm just a weirdo who needs things just right.

>>26337 >I'll try to make her lats softer. Good thinking IMO. Also, now that I can see her backside, again, I think her gluteus maximus are too masculine. I'd suggest giving her more of a rump. >I'm just a weirdo who needs things just right. Haha. Autism isn't being weird Kiwi. It just shows you have a real vision you're following! Cheers. :^)

>>26340 Smoothed things out. Her new lats are the same size, just much rounder. It should look better while still being appropriate for clothes. Made her shoulders smaller to be slightly girlier. Smoothed her booty over, same size, just rounder. More rump like :^) You can see the original and new as halves of the mesh. Added a slight bellybutton. I remembered I made a promise, long ago. Because raisins (I also have a thing for navels :^) (Her arms and legs are random kitbashed addons for the sake anatomy.)

>>26350 >More rump like :^) Heh. Looking forward to seeing the final outcome, Kiwi. :^)

>>26332 Make human is very limited in terms of possible scopes for body design. It can work for template meshes of course, but ultimately, you'll still need to refine the minute details if you're planning on making accurate human-like molds. In this case, meshing from scratch is the best option to have. >>26337 Nice mesh design.

>>26330 Why is this and the following discussion here in the general thread, btw? Not in modelling or prototyping? I did add links for all these more specific threads into the OP. >>26642 > you'll still need to refine the minute details if you're planning on making accurate human-like molds. This might be true, I have to look into that. But the thing is, I don't want visible muscles patterns showing up not to be in the mold of the skin, but these muscles should be created separately and then show up through the skin aka "outer (silicone rubber) shell". > In this case, meshing from scratch is the best option to have. Why? Is it hard to change the models from Make Human?

This video here, which is rather 3D print related: https://www.youtu.be/kkrkR1crNEc made me think that it might work to use some (conductive) paint, primer or some similar thin liquid to galvanize these parts. This has been used for 3D prints already for quite some time, but I never saw it mentioned for wood. I was thinking some wood, cardboard or paper might be able to take a liquid deeper into the material and then this would help to alter the part to a higher degree into a part consisting of metal.

>>26371 >Outcome That'll be awhile. I went back to making a mesh based on Aigis. My heart yearns for her specifically. >>26642 Thanks, of you want it, I can upload it to Ca box since I'm focusing on a different design now. Honestly, it would be nice to have those hours of work man something for someone. >>26762 I was researching into what design aspects mater to others. That was a design that I want even sure I'd use, but saw potential to develop into something. >Is it hard to change MakeHuman models? Yes, it's a huge PITA. It's easier to start from scratch if you know how. It would've taken longer to reduce it into what I want. Plus, MakeHuman models print out ugly. I speak from experience, maybe Chobitsu remembers my MakeHuman tribes from the old board? >>26766 Excellent idea, definitely could go somewhere. Please, try it out

>>26766 Very nice NoidoDev, thank you. :^) >>26773 >My heart yearns for her specifically. She is an amazing waifu. Patrician taste tbh. :^) >maybe Chobitsu remembers my MakeHuman tribes from the old board? Yep. OTOH, your OG printing skills lore has likely improved today? And IMO you're on to something ebin with the printed flats anyway, so I'd go with that if I were you Anon.

Doesn't really fit into 3D printing, but especially Kiwi uses 3D printers to make some kind of meshes and I just found out that this here exists: >8 Best CNC Vinyl Cutters (For Every Budget!) https://www.cncsourced.com/rankings/best-vinyl-cutter/ https://youtu.be/B0chtELuYJg Cutting Vinyl with a cutting machine, including cutting a pattern into it. I'm not entirely sure if it can cut Vinyl into a kind of mesh, though. No one seems to want to do that, but I think these devices could do it. After looking more into it, I think the right variant for it would be a "Pattern Plotter Cutter". Or just a laser cutter: https://youtu.be/zObGjvjmWDk (depends on the material) - knife cutters can cut multiple layers, not precise - lasers cutters are more precise, but might create very toxic fumes ChatGPT: >Regarding cutting vinyl into a mesh, it's important to note that vinyl cutters are generally designed to cut solid sheets or rolls of vinyl rather than mesh materials. The cutting process involves a blade that cuts through the material, and mesh might not provide the necessary support for clean and precise cutting. If you specifically need to cut a mesh material, other types of cutting tools, such as laser cutters or specialty fabric cutters, might be more suitable depending on the material's composition. >In summary, a vinyl cutter is not typically used for cutting vinyl into a mesh. Its primary use is for cutting solid vinyl materials to create various designs and shapes. If mesh cutting is a requirement, exploring alternative cutting tools may be necessary. Probably none of this is worth for prototyping, especially if someone already has a 3D printer, but maybe it will be useful for some local serial production of robowaifus in the future. Some of this might also be a way to make flexible circuits for electronics.

>>26846 Neat! There are also mid-tier vinyl/paper CNC-type paper cutters out there for not too expensive. Would be a good way to try shell prototyping on the cheap.

In a week or two, I’ll start looking into integral sensors (stuff that needs to be placed/accounted for before the design is “finalized” (rev1))— stuff like limit switches, thermometers, ammeters, voltmeters, maybe a couple strain/distance sensors. (many of these are only needed for the prototype, since I need to evaluate potential points of failure or get data for future design improvements) Anyway, the requirements I know for sure now are: (sensor) (interval or latency) 30+ thermometers 1.0s +- 1s 8 ammeters 100ms +- 50ms 4 voltmeters 1s +- 1s 10 limit switches <100ms, <50ms ideal 8 “rough”/RPM encoders ~ 500ms 4 “precise” encoders 100ms 2 2-axis angle sensor 100ms (probably a few more that I forgot, but hey, that’s enough to start) ‘Course, that’s a metric shitton of data with equally ridiculous wiring/IO requirements. Some of that can’t be avoided, but I’m hoping to mux and cyclically poll the thermometers, ammeters, and voltmeters: either in the literal sense (muxer upstream of ADC) or the conceptual sense (digital sensors sharing the same bus and being addressed/selected). I think the latter solution might be a tad expensive, but I don't have much experience here. Anyway, if anyone has experience with data acquisition on that scale, or ideas on how to implement it while keeping costs down, advice is greatly appreciated.

>>26846 Excellent reminder. CNC cutting tools are fantastic for everything you mentioned. With the right material and cutting tool, these machines could be great for production of panels in meshtegrity structures They're also useful for PCB fabrication at home. https://hackaday.com/2020/04/01/making-pcbs-with-a-vinyl-cutter/ As you mentioned, it can aid in flexible circuit production through cutting foil. https://fab.cba.mit.edu/content/archive/processes/PCB/vinylcut.html Many of us started with paper craft waifu. These make the prices faster and better. https://www.paperizedcrafts.com/2020/08/vocaloid-snow-hatsune-miku-paper-model.html?m=1

Bearings: >>28119 >Since silicon nitride ball bearings >>28200 >Rolamite >scroller wheels

>>28618 >placement of a speaker in the waifu? That depends, some would put it on the chest below the neck, others like myself would rather try to go with a internal system, trying to put several speakers in her neck and upper jaw, I guess. There are plenty of videos about 3D printed speakers, but I didn't deeply look into it. https://youtu.be/0tt3QHRiaRI https://youtu.be/4_4tu4rAtgI https://youtu.be/h-RSm97lkOU https://youtu.be/tZ7HznfA0YA

>>28621 https://youtu.be/lU4_E83RwUw (small, resin) .... https://www.youtube.com/playlist?list=PLh86lOr2S2w5WhsQLT9n65Olm4m0C3bq9

>>28621 >>28623 Neat stuff, Anon. Thanks! :^)

https://en.wikipedia.org/wiki/Anti-twister_mechanism - Avoid e.g. cables from getting twisted. >How can one supply a rotating lamp with energy WITHOUT a slip ring? As recently as 1971 Dale A. Adams found a mechanism although Indonesian candle dancers knew the solution for generations. We show a fischertechnik models of Adams' mechanism and another one that is easier to understand. We also briefly comment on Dirac's belt trick and the rotation group SO(3). https://youtu.be/gZvimEf6DFw >When you twist your arm or a belt by 360 degrees, the hand or endpoint is back to where it started but the rest of your arm or belt is still twisted. But if you do a 720 degree twist, you can manage to untwist your arm or belt! This is known as Dirac's Belt Trick or the Balinese Cup Trick. This crazy fact is even connected to physics with spin 1/2 particles, so let's try and figure out why! We will study rotations in 2 and 3 dimensions, and specifically study them topologically as opposed to algebraically as you might have seen before with rotation matrices. For a 2D rotation this is identified with points on a circle S^1. For a 3D rotation we need both an axis or rotation and an angle of rotation and we identify this with the solid ball of radius pi where a point in the ball gives a vector from the origin to the point that is our axis of rotation and the length of this vector is the angle. There is a catch: we have a double counting along the boundary so we have to identify antipodal points as the same. If you eliminate the origin (ie no rotation) this is sometimes called the Special Orthogonal Group SO(3) which is topologically the same as 3D Real Projective Space RP(3). A belt is then a path and I show an explicit way I can continuously deform the 4pi rotation path back to the identity. https://youtu.be/EgsUDby0X1M

>>28652 Neat! That could be useful.

Some fascinating idea about mechanics, don't know if this could be useful: https://www.youtube.com/shorts/NR75Ws54eNQ - Maybe as in inspiration to have a complex movement pattern, e.g. when walking.

Which do you prefer and why? I'm having trouble getting proportions right. Any and all comments are appreciated as part of research for development.

>>28772 It's hard to say which one I prefer definitively. If you put a gun to my head, I'd probably say four, but I have some misgivings. Do keep in mind that all of the following is just my own personal view. What matters most, of course, is what you want from your own waifu. While I like the stylized nature of the arms and legs in (1), I don't like that the hair is a part of the base model itself, the head itself is definitely too small, and the torso is too thin and underfeatured. Model (2) is right out. I could pick at little details, but the bottom line is that I don't have a realism fetish. I like the head from model (3) the most, for some reason. I can't really point to why, but it may be that it's appropriately stylized without being overdone, and it has a somewhat narrower jaw than (5). The torso has the same problem as (1), in that it's too thin and underfeatured, and the stylization of the arms and legs don't really fit to the body in my opinion. I like everything from the waist-down in (4), but things start looking weird to me in the upper portions. The shoulders seem a bit too wide, and the tits just look strange. This is hard to pin down, but they might be too low. The head is weirder to me. This is harder to pin down, but it's raising a red flag somehow. Lastly, (5) as a whole is too thin to the point of looking gangly. The neck is way too long, the head is too small, the torso as a whole looks underdeveloped, the hips don't curve enough, and the hole in the elbow just plain bothers me. Just for shits and gigs, I slapped together features from every model except (2) and resized them to get something I personally like a bit better. Of course, I fully realize that changing models around is much more involved than my efficiently lazy GIMP-fu, and I respect the work you're putting into this. Proportions become much less of a science and more of an art when you start stylizing things. It all comes down to what looks nice, and personal tastes can factor quite heavily into this.

>>28775 >I slapped together features from every model except (2) and resized them to get something I personally like a bit better. And that's really good.

>>28772 My apologies for not going in depth as you ask, but tbh I concur with every one of Greentext anon's points (>>28775). I'd recommend you try to re-create his model in a front ortho view. BTW, I'd give her a bit more of the booba than in GTA's image. :D Very good post, Kiwi. Thanks & cheers. :^)

>>28772 I just copy some of my comments from Discord over: - Two is the worst. I prefer four or five, I think. But I don't like the spheres. - I like >>28775 more as well. Fascinating how we have the same taste at the end of the day, plus minus some booba. - What I dislike with these realistic model like 2 is that curve between hip and upper tight. That dent. Drawn characters often don't have that. I minimized that in my Make Human model as much as possible. The other thing is, I think, that the thighs in such artificial models are longer, which looks better. I think I twisted that in Make Human as well and maybe I should do this more at some point.

>>28795 >I think, that the thighs in such artificial models are longer, which looks better. Yeah, that's one of the appealing things about designer waifus... we can plus things we want to exaggerate! :^)

>>28775 Perfect, I finally have the template I've been looking for. I've modeled so many designs that never felt right. This finally gives lights the path forward. I thank you for this tremendous help. It will take awhile to model the new design, I hope to collaborate further once it is done.

>>28795 >What I dislike with these realistic model like 2 is that curve between hip and upper tight. Those are called "trochanteric depressions" or "hip dips", and even some real wommins don't have 'em :D

>>28775 >>28785 >Booba Add your own silicone, it's faster and more effective to buy some fake jiggly's. Can confirm that printed booba ranges from soft but like honkin' a rubber ducky to cutting my hands from fondling too hard. :^) >>28795 Noted Newest body design using the advice given. Everything is scaled to a women of 150cm, will fit the xs women clothes. The clothes may be a tad loose, she's based on the smallest asian women. Thoughts? I'll make her arms, legs, and head after if this is goo enough.

>>28893 That looks really excellent, Kiwi. Really nice job on the cleanup as well. :^) BTW, Lin and I have a little discussion going that affects you too. Mind have a look? (>>28458, >>28874, >>28882)

I just discovered the Moiré Effect, which might be useful for making 3D printed parts changing their stiffness dependent of the orientation of the disk it is assembled of: https://youtu.be/jGoJObfYTtg https://www.youtube.com/shorts/D0nCNdl3-aA This is just a very early idea, I need to look more into this at some point or try something out. I found this while watching video in regards to magnetism, including kinetic magnetism: https://youtu.be/WPfcpHVlYZA - something about making materials magnetic or not with small amounts of electricity. Related (I didn't read this): https://www.nature.com/articles/s41586-023-06633-0 >>28824 Thanks, I'm pretty sure many slender white women don't have them. The model really wasn't a good pick in terms of looks. >>28893 Thanks.

>>28897 >I just discovered the Moiré Effect, which might be useful for making 3D printed parts changing their stiffness dependent of the orientation of the disk it is assembled of Wow, that's really interesting. I presume it will be as simple as printing different flat meshes, and then sandwiching them together appropriately?

Noise absorption, based on how materials are structured: https://youtu.be/y9-p4AkgVU8 - I have sound design and absorption on my radar for quite some time. I probably mentioned quartz sand for noise absorbtion already, which is of course additional weight. The other idea I looked into was printed music instruments e.g. https://youtu.be/QfWJbtmXOfQ - This would be more about sound design. Changing the noise that can't be avoided, e.g. to make her more sound like Ava from Ex Machina. But the idea behind the first link is about making cavities to absorb some noise. I think a combination of these will be useful, and we will experiment with this at some point. >>28878 >MyCera (printing structures for fungi) Hmm, not sure how this is going to be used. I guess we would need to print the bones with the hard material, or even leave a space for some metal there. Then the outside would need to be fungi. So maybe the hard shell would be removed, it would only be there to let the structure grow.

>>28907 >Hmm, not sure how this is going to be used. If we were to use something like MyCera it would probably be used for the bones. Cardboard bound with mycelium has similar strength to particle board so I imagine if used tougher materials like clay, sand, plastic, ect the end product would be even stronger. The advantages of mycomaterials are that they can be biowelded together, they are light, they are cheap, and they can be produced at the cottage scale. Disadvantages include they are biodegradable, they take a long time to grow, and they require a relatively sterile environment to grow.

A good explanation why learning how to program is still worth it: https://youtu.be/6CGtwF_5kzY - Even GPT-4 can not replace programmers completely, it rather helps to assist and to learn more. Related: >>128 (Programming Langs), >>159 (Python), >>12 (C++)

I think the most feasible milestone atm is a spaceframe, open-source doll. It takes care of the 3d aspect of the waifu while all the AI stuff is being developed by the greater open-source community. It can't do your dishes but I think most would be more concerned about the "waifu" than the "robo".

prototype here >>30545 and following link from >>30573 https://www.uea-inc.com/blog/what-is-a-slip-ring-and-how-does-it-work

Apptronik (Apollo Robot): https://www.youtube.com/watch?v=ab9QIzYkUgo (seems to have some CGI in it, without stating so) - Humanoid with (swapable) battery lasts 4 hours, so could be lighter with a smaller battery. - They tried wheels, and say legs work better and will win in the long run. - Dexterous hands seems to be one of the hardest things. Though, for simple tasks hands won't need to be as good as human ones. - Will cost 50k, for factory and logistic work.

>>30802 Neat! >and say legs work better and will win in the long run. The latter is probably true for general ambulatory locomotion around the home, etc. For longer-distance 'commute', however, wheels at the distal end of the limbs will clearly be faster & more-efficient. Thanks, NoidoDev. Cheers. :^)

Does anyone know where the patent for this is? Im just curious if there is anything unique or not in the lip moving mechanism. >product is remote kissing device >lead inventor Jiang Zhongli >patent holder Changzhou Vocational Institute of Mechatronic Technology >patent filed in 2019 >patent expired January 2023 I assume the patent would be listed here but I am having trouble getting past the login wall since registry isnt sending me confirmation email https://english.cnipa.gov.cn/col/col1362/index.html

>>30824 G*ogle has an effective patent search system, as well as a good index of scholarly works to boot. I hate recommending them, but for this area they are a valuable asset.

>In this tutorial we explain a method for bonding a silicone skin to a foam core. In this case the silicone skin is TC-5110 brushed in using Thixo. We embedded cotton fiber in the 2nd coat and used the exposed fibers to bond the silicone to the foam. TC-5110 is a soft, translucent skin material for casting medical simulators, dolls, masks, FX skins, and prototyping. https://youtu.be/hugF0RMyP9E

>>30824 Didn't I see this on an episode of "The Big Bang Theory"?

David Browne has designed+printed an excellent shoulder joint and he shows it in detail in this video. Those ASME high-torque servos seem to be a very popular choice for the larger joints. https://youtu.be/Vy17buWA_D0

>>31618 Neat! Thanks alot, SophieDev. His waifu design goal is impressive. Cheers. :^)

>>31369 Neat! Thanks, NoidoDev. Cheers. :^)

>>31618 >Gearing down a 180kg servo kek

>>31623 That might make some sense in a home-healthcare/medical-institution context Anon?

>>31623 It actually makes some sense. Consider that it's to actuate her shoulder. If her arm is 50cm, made of PLA reinforced silicone, with silicone skin, her arm likely will have a mass of a a kilo. 180cm.kg servos would max out at 3.6kg of load. Her arm would eat the .6kg of it. the 3kg for her to hold something. This is the breaking point though. Her servos will break right fast if pushed to its limit. In reality, she'd have to only lift 1kg objects to have the servo last. Gearing down with 3D printed plastic is indeed silly. There's no way they could handle 180kg.cm of torque. The teeth will sheer off. The ASME MR series are the most impressive servos in terms of torque to cost. Note, make sure you're getting the magnetic encoder variant, they are surprisingly accurate. https://www.youtube.com/watch?v=GwRxrp4nzDA

>>30824 Is this it? https://patents.google.com/patent/CN109821132A

2.5D robotics could be the answer to our problems. Make them light by making them almost completely flat. This way they can look like 2D waifus and still move around with us. They'd just look a tad wonky from off angles. I think we only really need her to look good from front and behind to be good enough. https://www.youtube.com/watch?v=qqOv69P3qiQ https://www.youtube.com/watch?v=YOXUDDPeYF0 https://www.youtube.com/watch?v=o-mjs-2UNps

>>31835 >2.5D robotics I think this fits into this thread: >>271

>>31848 Not inherently because you could easily use something like plywood or plastic sheets to make them less delicate and last longer as result so it would no longer fall under papercraft in such cases.

>>31860 >2.5D robotics True, it doesn't always fit into the papercraft topic. Maybe there needs to be a thread on it's own at some point, if there's enough to post about it.

I like this robot's eyebrows. I'm trying to figure out how they made the corners move. Just regular servos? And, I wonder why they made the corners go back that far, when shorter eyebrows would look more natural.

>>31835 >>31848 >>31860 >>31862 Glad to see this topic being brought back up here Anons. It was a good idea years ago, and it's still a good one today! Cheers. :^) >>31917 My general impression from his videos is that there are two simple servos per eyebrow. Both located at the furrow of the brow; one controlling the rotation, one controlling the vertical 'sliding'? (cf. : >>15287) . The other end(s) seems simply to be fixed in place. I hope you'll give it a shot Anon, and let us here know about your results. Cheers. :^) >=== -minor edit

Full color image transfer to the base of 3D prints. https://www.youtube.com/watch?v=eElO5aso8kY

>Wires and buses for sensors: Buses (I2C, CAN, DMX512) and multiplexing in the SPUD thread: >>32583 Light buses, light based sensors and related hardware: >>32874 >>32896 I meant the light could come from one side of a limb and go into the knuckle, then the rotating part in the connector would only let a certain amount through or change the color. The other side could pick that up. Though, I doubt we have such good sensors and small enough.

>>32900 You may find optical flex sensors interesting. https://hackaday.com/2011/10/21/building-optical-flex-sensors/

>>32909 Yeah, very interesting, but I prefer to not have anything bending in the knuckles or any joint connector. We should've something in there that rotates and keeps up the connection, while maybe also indicating the position.

As to my previous assertion that a fiber-like form factor for rotation sensing in knuckles : ( >>32882, et al ) : why not use electricity instead, and use thin, bi-metallic strips as the sensor? This should create a Piezo field that can be read at the MCU, wouldn't it? >muh grid o' light!! :^) In essence, isn't either multiplexing, or even just straight-up GPIO pinouts -- with the accompanying downstream software readouts -- just about the same? A sort of 'six-of-one, half-dozen-of-the-other' thought experiment, no? <---> In the end, I suspect that these three design criteria will be the most important (in no particular order ATM) : * energy efficiency * cost of materials for manufacture * ease-of-manufacture

>>32919 >What would really be slick is to put a reflector on the end of the fiber Yes! I had actually thought of that myself (and that sort of time-based reflectometry is already used for distance-measurements of fibre runs). But as you say, it seems this might increase complexity/cost ?

>>32909 That looks quite similar to my idea, actually : * but with one homerunned fiber per joint * the sensor element would be attached at the distal end of each, and extend through the rotation frame (in this case, a knuckle) * all 15 knuckles would thus feed into one, 15-element fibre grid camera (per hand) * optimally, the light source itself would also be located back at the camera assembly, feeding input light into the sensor. The thus 're-transmitted' sensor data (very-likely at a different frequency than the source light [and to which the cam would be tuned for]) would be fed back down the same lightpipe to the grid camera. Simple to assemble, simple to maintain. Simple as. :^) <---> I might add to this that the finger-actuation 'tendons' (feeding up through the Carpal Tunnel wrist port, up from individual actuators located down in the forearm) would be laid down right alongside the sensor fibres in the same grooves/channels. >=== -add 'tendons' cmnt -prose edit

>>32946 >rotation sensing in knuckles : ( >>32882, et al ) : why not use electricity instead, and use thin, bi-metallic strips as the sensor? No, no because then you have wires (possible electromagnetic interference) which cost more than cheap plastic fiber, you have to have AtoD converters to read them(slow and you need a pile of MC to have them, more chips, more power, more complexity) and possibly(likely) supporting circuitry for the AtoD. It's just a big mess where with fibergrid you send out light from a LED to a bundle of fibers going to all the joints, sensors, the light hits a varying level target, bounces back through another fiber to a into a bundle which goes to one CCD camera. You could send the value of each CCD pixel to RAM and then use these as values to compare. And the CCD will at likely operate at 60Hxz, maybe 100Hz. Fast, accurate,cheap. I don't think it can be beat and since I heard of fibergrid I haven't thought much about using anything else as I can think of nothing cheaper, more efficient and or faster. I talked about an idea for pressure touch senors here(in the wrong thread) >>32918 and here >>32919 and much earlier >>24233 >>30980

>>32946 >In essence, isn't either multiplexing, or even just straight-up GPIO pinouts -- with the accompanying downstream software readouts -- just about the same? I'm not trying to be rude but I'm not sure you understand fibergrid. The CCD "camera" is taking a continuous stream of data from a sensor and then you get a count of it where as any other sort of wire input would only allow a brief sample(without sample and hold or AtoD or other such time consuming expensive circuitry), not near as accurate. Maybe read the hackaday article or what I wrote again.

>>32959 >I'm not trying to be rude Ehh, it's fine Anon. Actually, I'm rather pleased that you care so much about this tech. I have worked professionally with fibre before, and it's amazing stuff! :D As to my position (6/12), I think that's still the correct view. I've worked with OpenCV before (again, professionally), and you still have to 'quantize' the operation within some computation device whenever you'd index into the imgmat to read the light splat's position. Not a lot different than simply reading a voltage value off of a GPIO pin, AFAICT ? The electrical readout may even be faster in a heavy-load scenario (since it's a single memory address read access, not an index+read). >=== -minor edit

This is related to magnetic amplifiers here, >>32646 >>32870 >Kiwi might be right, if there's something better and we work with DC then what's the point? The point is cost. "If", and this does not apply to everyone, but "if" we want a waifu with 100% human movement then you need roughly 300 muscles. That's a lot of transistors, Cost is not too bad but you also have to add circuits to drive those from micro-controllers and diodes for the feedback and boards and more boards, it starts getting complicated. Mag amps are seriously robust and while not as efficient overall as transistors they are not inefficient. That being said I'm not even close to 100% sure this the best way, but it does seem to be "one" way that might work.

BTW, I forgot, thanks NoidoDev for putting that into Llama 8B it did make things clearer for me.

>>32998 Thanks, but if I recall correctly in the response from Llama it told me that they are efficient anyways but the real problem was that these are harder to implement. Not sure if this is true, since Kiwi very likely knows better, LLMs can make errors in their responses it better to use them just for orientation and ideas.

>>33005 >harder to implement True, but that's because of the present "way" they implement them. Doesn't mean there's not another way. The present way is to wind a coil of wire around a armature but who says the conductors have to be copper wire or that they must be wound? Could they not be printed then sinistered or just printed or use low meting point soldier to replace copper lines. The AI doesn't know about these things only the way it's done now.

I wish I had a more powerful computer. Mine is not set up to do any sort of AI because I have no interest in gaming. So I've never needed a fast video card. They are anyways too damn expensive for me to buy one of any substance. Maybe if those guys who fiogured a way to refactor the neural net into a more effecient format, suing way less power start releasing models that are like 80b but only use the resources of a 8b then I could them.

>>33008 I don't think you need to run a big model. These might be more useful as an assistant anyways. You can always use some online service if you only need a bigger model from time to time. And for your card 7B models should work, especially quantized ones. This should be in the LLM and chatbot thread >>250 or just search for it.

I found a really good reference book called, "Mechanisms and Mechanical Devices Sourcebook" It's a good rough guide to how things are done. Has some effeciency references and some tps on wys thngs are "generally" done. I think tiis is an excellent resource to kind of browse through. It might giove you some ideas oir you might could use it to see if you're ideas could notbe improved with a few samll changes. I foud it on the excellent free Tor bookstore, Just Another Library http://libraryfyuybp7oyidyya3ah5xvwgyx6weauoini7zyz555litmmumad.onion/ The book http://libraryfyuybp7oyidyya3ah5xvwgyx6weauoini7zyz555litmmumad.onion/meta?fname=archives/books/nonfiction/9780071704410.McGraw-Hill.Mechanisms_and_Mechanical_Devices_Sourcebook%2C_5th_Edition.Aug.2011.pdf You owe it to yourself to look at this book depository. It's great and one of the really super good features is when you load a book below it will be a selection of books that are similar in subject, though this function is sometimes way off. I would say that you could easily find enough good books on the subject of making waifus, drawing, engineering, that this one site and a lot of downloading would cover most everything.

OMG sp? My fumble fingers can not hit the right keys, and I forgot to scroll up and correct.

>>33446 Thanks for the book. I did a skim read through the first quarter, and it seems to explain each concept pretty clearly and without too much industry esoterica or jargon. I'll give it a proper read once I dive back into concept drafting. I'm chaging Orchid's plan to include demultiplexers since I want to reduce the internal frame size and motor count, so I'll need all the help I can get figuring out how to make that happen. Bookmarked the library too, the 'most downloaded' section gave me a good chuckle.

>>33453 >Demux Picrel is a great conceptual reference for a mechanical demultiplexer. Would add sensors for indexing, also rounding the mating features would help. https://www.printables.com/model/163226-mechanical-multiplexer https://www.youtube.com/watch?v=8DWK3zm9SMs

An interesting idea that could be useful Making a 3D printer into a 2D plotter with G-code Generating Gcode from normal Images https://www.reddit.com/r/ender3/comments/1f6b75a/update_generating_gcode_from_normal_images/?chainedPosts=t3_15i2rr0%2Ct3_15xq6j7 https://github.com/IridiumIO/Polycut I could see this being useful for making stencils. I have this idea of making skin from NASA's flexible armor. To get the best results you need high resolution. You could print a stencil with ink, place over a plate coated with 3D resin and then put in the sun to cure the uncovered portions, building structures by layer. Slow to do but cheap because you don't need two machines.

I was looking into Joule thief circuits and came across a site where a guy claimed he somehow put out 8x as much power as the input with a few small modifications to the circuit: https://www.homemade-circuits.com/8x-overunity-circuit-using-joule-thief/ I haven't tested it yet myself, but I'd be happy if it was just a more efficient version of the regular circuit.

A website to help you develop your own cycloid actuators. https://mevirtuoso.com/cycloidal-drive/

>>33836 Very cool, thanks Light! BTW, if I haven't greeted you before, then welcome! Cheers. :^)

>>33836 Very nice. Thanks.

>>33836 Thanks, but the big deal for me was finding the formula for OpenScad by Dan Fekete on Youtube mentioned here >>24401 Maybe using something like this her and then importing the result into a program for further modification might also work, I could imagine.

Electromagnets work via the physics of electric fields having magnetic fields. (Measured in teslas) This magnetic field can be intuitively modeled through the right hand rule. A coiled conductor, often insulated copper wire, will have a magnetic field where if curled fingers match the direction of current flow, your up pointing thumb will point in the direction of the north pole. This magnet is likely to be weak due to the magnetic field freely circulated around the wire. To have an effectively stronger electromagnet, you need to constrain the field. This requires a permeable material, a ferromagnetic material, to act as a core. This core (usually an iron based alloy) has tiny magnetic regions that are disordered, cancelling themselves out. When a field interacts with these magnetic regions, they align to that field. More current will create a stronger field around the coil, causing more of these regions to become aligned with the field. Eventually, this will reach a saturation point, where further increases in current no longer meaningfully effect the usable magnetism. After current is removed, most of these magnetic domains will revert to their disordered state. Some of these domains will remain aligned, an effect called hysteresis, leaving the core a weak permanent magnet. This is unlikely to matter for us. If it does become a problem, you can degauss it by rapidly alternating current in the coil. When designing your electromagnet (or deciding which one to use), you need to think of the magnetic circuit you need to achieve your goal. Think of what you want, then consider how to use the flow of magnetic fields to achieve that goal. Consider a voice coil actuator, you want it to smoothly, quietly, and efficiently move a load linearly. To accomplish this, you could wrap a coil of wire onto a tube of electrical steel (used for its high permeability and low hysteresis) with a neodymium magnet (high magnetic force relative to cost) in the center to move relative to the field. A spring would be added so that it would return to a resting state without requiring power. (Voice coils can have either the magnet or coil move. Usually the coil moves in applications where moving the load fast matters most, ie; speakers. Where the load must be heavier and controlled with greater stability, the magnet is moved, ie; gas control lines.) It is important to always remember, the fields weaken with distance. You want your electromagnetic fields as close to the load as possible. This is why the space between a motors stator and rotor are tiny. Why the coil wraps around the magnet as closely as possible in speakers, etc..

>>35055 POTD Let's all do SCIENCE! to it!! :D

>>35055 Thank you kiwi im planning on using a mix of iron powder and polyester based on this video https://m.youtube.com/watch?v=YQNW4ZTsg2w he said polyeutherane but i assume polyester should work. Maybe i should finish the electromagnetic robot eyes. The steel screw magnet atleast was very innefficient however since it was very sensitive to resistance and drained the batteries in minutes. Okay yes ill finish the robot eyes might as well. i did this pretty rushed. not sure if i got the ratios right but i want to do this quickly.

>>35061 too quickly..

>>35062 What are we looking at here? Did something break, Anon? Regardless, just keep moving forward, peteblank. Persistence overcomes more obstacles in life than anything else!! More than talent, more than resources, more than genius. Cheers. :^)

I'd made this as a shitpost, but I was curious how viable it might be. The biggest engineering issue I think is the arms. Even the arm joints were very basic I'd be afraid of getting the pressure calibration correct. I joke about harm to the user if the safety settings are overridden but it probably wouldn't be a good look for the product if it regularly crushed their user to death.

>>35067 ive seen these on farms, humorously called crushers, doesnt look like a person but effectively functions the same, it just hugs/squeezes the cow for a while to reduce stress, they probably use a hydrophilic piston with a pressure valve so its not possible to go beyond the designed force

>>35067 >I'd made this as a shitpost, but I was curious how viable it might be. Great idea, Anon. I think it's entirely viable! It could be a big hit if you pull it off correctly. BTW, I'd suggest you have a look at our Kiwigril thread : ( >>104 ). Looking forward to seeing your project's progress here, Anon! Cheers. :^)

>>35067 sounds familiar ;) The engineer in me has solved the human squish problem: arms driven by a winch and use elastic cord as cable to pull the arms in. Similar to one of those hands made out of plastic drinking straws. Tighter the winch winds, the tighter the hug. But its bungee cord so easy to free if something goes wrong.

>>35063 tnak you chobitsu. its my 3rd attempt. i printed a mini spoon and made sure the ratios were right this time but i think itll get stuck on the mold like the second time. i hate casting. if you want to know what im trying to do watch the youtube link i posted.

>>35071 this is not what i had in mind. i could put some tin foil paper in between the mold but its very powdery. Hardly worth the hassle. why isnt there iron rods for sale ff. i already got the materials so i might as well keep trying it to do it this way i guess. He used polyutherane though im using polyester. Very strong odor.

>>35069 Stylistically, I can appreicate it but mechanically I don't like it. Kiwi project looks like a body pillow with legs. Also, I've seen what happens if you let your guard down around a kiwi girl...

>>35084 >Also, I've seen what happens if you let your guard down around a kiwi girl... Kek. Mean as. :D Anyway, good luck Anon! Keep us all up to date with your project's progress here please. Cheers.

>>35074 Im going to have to make a bigger mold i think but using tinfoil so it wont get stuck to the mold helped.

Using waste robowaifu body-heat to power a robowaifu walking light show? https://www.youtube.com/watch?v=uNGNP5EXNWs

New research into low power AI on the edge. Here's to hoping it goes somewhere that helps us. https://arxiv.org/pdf/2502.10089

>>37139 POTD Nice find, Anon! This paper's objectives really line up with our engineering design needs here on /robowaifu/ . Cheers, Licht. :^)

>>37139 Very nice stuff you've been digging up. Thanks.

Interesting gear train technique, relying on surface friction rather than gears meshing. I suppose you could call them atomic gears given the fundamental forces governing the behaviour. IMSystems uses this technique for planetary gear systems they refer to as Archimedes Drive. It's essentially turning a bearing into a drive. Relying on rolling friction is potentially efficient, with an added bonus of being as simple as possible. It's just cylinders rolling about each other. This method also benefits from slipping. The cylinders can slide along rather than breaking gear teeth, or worse, her owners teeth! https://www.youtube.com/watch?v=z_FdCGIzG-w

>>37277 Wow, your >pics-related are really interesting, Anon. Know of any thingiverse, etc., printable resources for this concept right now? Cheers.

Any'non know about these stair-climbing wheelchair designs? >related pic: https://trashchan.xyz/robowaifu/thread/26.html#52 >(cf. threads-related : >>2983, >>24744 )

3D printable animatronic eyes https://www.xda-developers.com/3d-printed-animatronic-eyes/

>>37277 Thanks, I think I remember a video from a few years ago about it. This strikes me as something that might hold up for a long time.

Tracing designs can be a fun way to learn and experiment.

>>38254 Interesting. I like where you're going with this, Kiwi. Cheers. :^)

Compiled anthropometric data on western woman. This can serve as useful data for designing a machine which is compatible with everything an adult woman is. Presented as a single sheet to print for reference while designing.

>>38450 POTD Nice work, lad. Cheers. :^)

How to make roller teeth gears. Potentially a viable way to produce a high efficiency gearbox at home. https://www.youtube.com/watch?v=VXcuryyRGbo&t=221s

>>38529 Wow, those look amazing, Kiwi. Nice find & thanks! Cheers. :^)

>>38450 Thanks, Really helpful.

>>38450 Nice! Thank you

>>38529 Another great link. Super interesting. I'll have to really lo at it to fully understand what he's done there. Complicated.

Another great video on how brushless motors work. https://www.youtube.com/watch?v=RUS7GGv-UJs

>>38547 THE LEGO FUN HAS BEEN DOUBLED!! This is great stuff, Kiwi. If we ever teach schoolboys these things *, I'd deffo recommend this type of approach. >11'000 RPM Lego Motors!111 Michael Faraday would be so proud! :D Excellent find Anon, thanks. Cheers. :^) --- * Maybe hold yearly "Robotics Week" competitions at the robowaifu factory(s); for private boy's schools during their Summer breaks? Something similar to what NASA does (or did...not sure now)? <---> If we ever actually do go this route, then a 'Robot Design & Engineering Class' curriculum suited to students should be developed by us prior-to. We should make multi-tiered (based on the number of parts, manuals, tools, instruments, &tc., included) classroom kits available for schools to purchase; which teachers can then use during the preceding school year -- so that the boys would have their team's robots all assembled & ready to compete during the big week at the factory. Interested & qualified factory staff could be given free time during the school year, to engage remotely & regularly with the teachers & classes ahead-of-time, as consultants to help guide their team's projects to success. Trophies & other prizes for the winning school's teams in various categories. Rando giveaways. Special rates & accommodations for qualified Boy's Homes. Big wrapup party+concert. Swag for everyone. In a word: treat these kids like kings during that week! :^) <---> I do not want 'Terminatorz-R-Us' -type sh*te to be within the yearly contest themes!! Rather, let's make it about navigating Anon's flat (or the outdoors); doing the dishes by hand properly (or laundry, &tc.); recognizing/responding properly to rapidly-changing (surprising) environments; timed obstacle-course events; &tc., &tc. >tl;dr Similar to all the nominal stuff we here are already aiming to accomplish. We can also throw in lots of AI-related and other things for their robots to demonstrate. Stuff like eloquent TTS<->STT, artistry, logic-puzzles, animal-care, gardening, monitoring & cataloging plantlife & wildlife, weaving & crafts, music-making (NAME MY BAND!! :DD, &tc., to flavor things up a bit. Kind of a Country Fair type scenario such as our own @Mechnomancer (and myself; possibly others here) might relish. Simply be creative ourselves, and always take recommendations from the schoolboys, the parents, & the schools for suggested improvements to the competitions. Everything recorded & edited together nicely after the fact -- with good soundtracks -- to serve as propaganda for the next year's big event! :D <---> We can entice the schools & parents towards the project by us thereafter helping particularly-talented young men network to get placements in realworld jobs (not least of which could be our own firm's jobs!) Ofc, other interested companies can come alongside as Platinum-tier, Gold-tier, &tc. corporate sponsors (no (((military))) horsesh*te, ofc)...once the event took off. We could set up a free jobs fair for the lads right outside the competition arena venue. Various qualified Unis, as well could engage there. >ttl;dr New blood is how you keep a movement alive. TWAGMI

>>38547 >>38558 > (lego -related : >>38416 )

>>39001 >NEEDS: >speech, robot movement, power supply and safety features <---> Let's address your primary concern here first. Premise: >A US$500Bn-taxpayer-money-theft-tier, Samyuel Copeman-tier, AI video generator setup is needed to: >dynamically run [robowaifu] movements Issues: >muh nuclear-powered datacenter is needed!! >it'll cost muh six-gorillion j*wbuxx!! >else its muh Baste Chinese >then its muh glowniggers, femsh*tes, and conservicucks!! >...and all that that implies!111 :DD My apologies for being so facetious, Anon. Just kidding around a bit. All due respect to you. :^) --- This is a fallacy. We don't need AI Movie Generation -scale capabilities to either a) control a robowaifu's movements, or b) to understand other humanoid's (such as her Master's) movements. This problem was solved long ago in the computer animation industry by skellingtons + (so-called) rigging. Both are highly-mathematical, highly-optimized (computationally-speaking) processes today that are relatively low-power. A) As to the first need, its simply a matter of animatronics-style animation of her skellington's rig. (This is similar to character animation in vidya & film, BTW.) B) As to the second need, its roughly-speaking been commoditized to the point that it will literally run on a few-bucks-cost microcontroller today. [1] So no, we won't need yuge-a*rse datacenters just to manage our robowaifu's bodily motions, nor for her to analyze properly the people/robowaifus in the scene around her. It can all be run onboard, fully self-contained, within the robowaifu's internal compute hardware of today (ie, tiny-PCs, SBCs, MCUs). --- The real issue is the systems control (ie, 'Executive', or 'C4 Systems') software to generally oversee & operate/send-commands to the robowaifu herself. This part is some real frontiersman, trailblazing stuff at this stage. Much has been done within industrial work, and humanoid systems are coming along now too. But we all still have a looong climb up that mountain to go! :^) >(cf. 'Safety' below) <---> Speech. As you mentioned, speech is largely a done deal today as well. The quality & hardware needs vary dramatically at this stage, but my instincts tell me that we can get working robowaifus at a reasonable level today with little other than consumer tier compute hardware -- all running locally there inside Anon's flat -- 100% disconnected & offline. <---> Ahh, batteries... Yes, it's a big issue still. There are ongoing research efforts (primarily in the EV auto industries), and incremental improvements are occurring. As you mentioned, qwik-replace/augmentation approaches are probably needed for our robowaifus today. Additionally, incremental-improvements are still happening with actuators research...an important point. <---> Safety. This is literally the.hardest.task. that we here on /robowaifu/ & elsewhere face, I deem. We're decades away from highly-solid solutions in this arena IMO -- and even then they will only be provisional! OTOH, we don't have perfect safety today with cars/planes/dishwashers/toasters. That fact doesn't stop multi-billion dollar industries from humming along surrounding all these, nor for millions & millions of consumers from using them. >tl;dr We'll muddle along somehow, I think. BUT we'll need some kind of legal-defense union [2] to keep the greedy lawyers, et al, from attempting to destroy the 'little guys' trying to participate in this nascent robowaifu industry. <---> I hope that touches the bases sufficiently to address your great post properly, Anon. Cheers! :^) --- 1. (cf. >>38841, et al) 2. (cf. >>38515, et al)

>>39001 >>39015 Even ignoring the societal complications, there are problems with having a non-local AI as discussed here; >>38090 >We'll muddle along somehow, I think. BUT we'll need some kind of legal-defense union [2] to keep the greedy lawyers, et al, from attempting to destroy the 'little guys' trying to participate in this nascent robowaifu industry. We should look into LLCs.

>>39021 The solutions for these issues are generally within the domain of the Executive/C4 Systems mentioned already (cf. >>39015 ). The robotic motion control is straightforward, and easily-achieved today -- whether pre-canned (MOCAP live-feeds/motion-libraries, etc.); or programmatic (PGC [1], et al); or (more-likely) a combination of the two. <---> The primary difficulty lies in properly devising the executive decision trees that will correctly mimic wholesome/right-minded (ie, non-kiked) human sentiments & activities with near 100% accuracy. Something I'd be comfortable allowing to watch-over/tutor young kids, for example. Certainly the LLMs of today are not up to the task!! :^) >tl;dr As mentioned, we still have a long way to go yet. Whether the robowaifu's motions derive from MOCAP or PGC sources is nearly-irrelevant to this much bigger question at this stage, IMO. --- 1. ala, MASSIVE, et al https://en.wikipedia.org/wiki/MASSIVE_(software)

Testing follow me eyes with 2 inch glass hemispheres.

>>39157 Neat! I think if you can A) Try lenses that are slightly less convex, and B) Try 'eyes' that are slightly concave+moved closer to the lens then you might get somewhat better results. Regardless, very cool Kiwi! Cheers. :^)

Tutorial for designing a cycloidal drive variant. https://www.youtube.com/watch?v=wijJdARkCBc

>>39158 The printed eye was directly on the glass. New eye is offset with flatter lens.

>>39157 >>39214 Nice work and good idea with the lenses. I tried follow me eyes before, but the sunken in eyes made it look weird. Fortunately, it looks like the lenses solves it. Where did you get the hemispheres?

>>39212 Another great entry here by this guy. Thanks, Kiwi! I may be able to integrate this component design into my smol'r hip actuators (and potentially the shoulders as well). What I really need ATM is an actuator system that is hollow through the center (as in a wheel or a ring)! That way I can 'stack' different actuators together axially into a compact volume. This will be especially helpful in the complex hip areas. If any'non knows of such actuators as opensauce designs (or even inspirational concepts/proprietary designs) that would be most helpful. :^) >>39214 Ahh, yes naicu! That's a good improvement IMO. Did you sketch those eyes? Those are quite good. I really like the idea that among the many, many things we can customize upon our robowaifus (much like changing her clothes), we can change her eyes! Cheers, Anon. :^)

Excellent tutorial on making stronk danboard constructs. Recipe 1 .5 Liters of alcohol 125 grams shellac flakes 15 grams beeswax All mixed on a well ventilated hotplate at low heat. Recipe 2 100 grams of hot glue stick 25 grams paraffin or bees wax 20 ml mineral oil https://www.youtube.com/watch?v=45JhacvmXV8

>>39331 POTD >danboard <uwotm8? who is this "Dan" you speak of? :D I've been playing around with the ideas of papercraft ( >>271 ), etc., type concepts for extremely-inexpensive approaches to fashioning robowaifu bodies. Cardboard is a great choice! I hope we can literally use just paper, glue (incl. wheatpaste), tape, cardboard, some comfy cloths & foams, wooden dowels, paint, now this shellac goop, screws washers & nuts, plus a few 3D-printed hardware components to make everyman robowaifu bodies. (Just for the skellingtons & shells though...they'll still need cheap electronics, batteries, motors, &tc. Software will be free.) Thanks Kiwi, great resource. Cheers. :^)

>>39242 >Another great entry here by this guy. Thanks, Kiwi! Can you please re-post this link, Kiwi. Looks like we lost that one. TIA.

>>39336 There's great potential for paper skellingtons! >>39337 I believe it was one of these https://www.youtube.com/watch?v=wijJdARkCBc&t=294s https://www.youtube.com/watch?v=pfvy5IcW-8U&t=71s

>>39344 >I believe it was one of these Yes! It was the first one. Thanks, Kiwi! Cheers. :^)

After failing to make a cute head too many times, I'm studying cute models. Starting with Rem

>>39445 Here's the head I heavily modified for Galatea https://sketchfab.com/3d-models/anime-head-pratice-free-download-3fd6a9ccdcaf444aaf01d55b58ea50d8

>>39445 Please just stay encouraged, Kiwi. Facial is literally the single hardest part of 3D Character modelling/animation...so its not just you. I'd say study the master's studies of the female face. This will take lots of time though, so there's that. And we're already about halfway through the year...only 6 months to go till Christmastide 2025/26...

>>39447 Seconded. Artistry is really hard. The fact that you can sculpt a female body from scratch is already impressive. Your previous model is very impressive, and I'm surprised you're disappointed with it. >cute head Keep one thing in mind, what you're seeing a monochrome material depiction of a face, often completely bald. It will make it not look as good as it would in real life with color. >facial "uhuhuhuh Beavis check it out, he said facial" "yeah hahahaha he did"

>>39447 >>39453 >le face is difficult. Which is why I humbly suggest prefab. The biggest hurdle is the "robo" part of a "robowaifu". Might have already been said by other folks but kigurumi mask (specifically animegao) would allow for the robowaifu developer to focus on other aspects rather than the face and make kits all the easier. And by using a prefab face you're using a product that has been more or less proven as something the market likes. (Picrel: Spamazon's selection)

>>39446 Thanks! Did some minor cleanup on the head and here it is as an .obj file. https://files.catbox.moe/ppgks0.obj >>39447 I'll keep at it. The rest should be easy, I'm rather comfortable with arms, bodies, and legs now. I hope to spend Christmas with a waifu this year. >>39453 >Artistry is hard True, learning that more and more over time lol >Bald Good point, adding a hair mesh makes heaps of difference. >>39461 >Those faces I mean no offense to you as I admire your work. God forgive me but, Jesus Christ how horrifying.

>>39453 "huhuhuhuhuhuhuhuhuhuhuhuhuhuhuhuhuhuhuhuhuh..." :D https://en.wikipedia.org/wiki/Computer_facial_animation >>39461 Thanks for the advice, Mechnomancer. Kinda creepy tbh tho, bro... :^) >>39463 Yeah, I think this looks fine. If you can perform a smoothing operation before sending to the printer it should be good IMO. >I'm rather comfortable with arms, bodies, and legs now. Yeah, you're doing very well at it now, IMO. >I hope to spend Christmas with a waifu this year. This.

>>39445 Rem or Megu inspired design, which would you prefer?

>>39504 Personally, I like them both! I'd say just go with the one in your pic-related for this post. Time is short. (And you can always go back in later with mods after you have a working waifu). Good luck with this project, fren Kiwi. Cheers. :^) FORWARD!!

>>39463 >I mean no offense to you as I admire your work. God forgive me but, Jesus Christ how horrifying. Lmao none taken. What the market demands can be quite horrifying. It is what it is |:o)

>>39504 I do prefer the megumin face.