File: assorted-parts-progress-f(...).jpg (724 KB, 3024x4032)

724 KB JPG

Here are assorted parts progress for the archimedes pulley system which will downgear my 2430 BLDC motor 16:1 SILENTLY. Silent downgearing is necessary for a proper human passing robot after all.

OP is a faggot
https://www.youtube.com/watch?v=WvK2F0Sk56k [Embed]

>>2891578
To be more accurate; OP is a narcissist attempting to perfectly emulate the human form (he's some form of born again Christian, so NOTHING deviating from the human form will do). Unfortunately, he also suffers from severe Dunning-Krueger syndrome; he has absolutely engineering acumen. Should you speak to him, expect walls of text thoroughly explaining his abilities and engineering choices (which may on the surface appear vaguely sound, but quickly fall apart under scrutiny). He has no working prototypes of anything (the concept of 'prototyping' systems seems abhorrent to him). You have been warned.

>>2891578
cool video. That guy sure knows how to make a video fast pace and entertaining unlike me (or I just can't be bothered to put so much effort into entertainment value). That said, his robot is low hanging fruit style. Not a serious feat of engineering like I want. My robot would be able to pick up his robot and carry it kicking and screaming to a dumpster and toss it in. It would never be able get out.

>>2891591
Dude, seriously; get something to move under its own power, then you get to crow. Until then, STFU and get to work.

>>2891590
>To be more accurate; OP is a narcissist attempting to perfectly emulate the human form

First, I'm not a narcissist just because I'm confident. So that's a false accusation right there. But yes, a perfect emulation of human form is necessary to have a sci fi hollywood movie level robot and anything shy of that is for a total loser to make. Absolutely pathetic.

>(he's some form of born again Christian, so NOTHING deviating from the human form will do).
The two aren't related. I actually have taste. By your logic, Megan Fox playing a robot on a movie was done because the director decided since he's so super Christian that a beautiful human form was the only robot that would do. And then put in sexual scenes to in order to please God. This is your logic. It is dumb. A proper humanoid robot necessary cannot deviate from human form because to the extent it deviates it is trash. That simple

>Unfortunately, he also suffers from severe Dunning-Krueger syndrome
I have zero of this. I'm just that elite is all.


>he has absolutely [no] engineering acumen.
not true. I'm elite at engineering and it's obvious to all but you.

>Should you speak to him, expect walls of text thoroughly explaining his abilities and engineering choices (which may on the surface appear vaguely sound, but quickly fall apart under scrutiny
So you admit my engineering choices are sound. You say they fall apart under scrutiny but give no valid reason for this false assessment. Prove it.
>He has no working prototypes of anything
The frame is a working prototype and works great. All joints work great. The winch in place pulley works great too. The archimedes worked great but I'm going for v2 now. All of these are mission critical and working prototypes so you lied on me here.

>(the concept of 'prototyping' systems seems abhorrent to him).
No, it's good practice which is why I do it ever time.

>>2891594
>I’m just that elite is all.
sure thing bud lmao

https://m.youtube.com/watch?v=6r4Gw9oUWDw&t=333s&pp=ygUSZ2lhbnQgYmVldGxlIHJvYm90 [Embed]

>>2891594
If you were so elite you wouldn't need any v2

>>2891612
part of being elite is ruling out deadends and finding answers no matter what until you break through. This is part of being a frontiersman, game change, ground breaker. Even lightbulb inventor took v2, v3, v4..... v500 etc but now we have light. And he was hella elite

>>2891623
Sure thing timmy. Say, how about any actual demonstrators of these so much elite engineering? Any v0 or even a partial mechanism prototype or do you just shitpost on 4chins instead of actually working on your 1337 pr0j3c7?

>>2891625
He's actually a multimedia shitposter; presumably he's continuing these threads because we're the only people paying any kind of attention to him. Search for 'artbyrobot', he's on all platforms.

>>2891625
I assume you mean you want a .gif file? I tried uploading one just now and it says error, max file size is 4mb

here's a link to a .gif demo http://www.artbyrobot.com/first-pulley-test-animation.gif

>>2891625
the V0 archimedes pulley system I was testing with a 10lb dumbell it was doing great but broke because I tied the wrong kind of knot. I used square knots x4 instead of like a figure eight knot. Square knots concentrate tension into a specific spot causing the fishing line not to perform up to its specifications.

>>2891625
but I will probably make another .gif once I make my v2 archimedes pulley system - although I might not do this either until I get my motor actuated. The plan right now is finish archimedes pulley system - just have one more double pulley to go > then actuate the motor with full custom microcontroller and motor controller circuitry setup > watch finger move. I can then make .gifs of the various things moving - a .gif of the motor winch and its winch in place pulley mounted right on the motor, a .gif of the archimedes pulley system working, and a .gif of the finger moving. That's the whole chain of events.

>>2891627
I'm not a shitposter, I'm a elite world class roboticist widely publishing my game changing, ground breaker, historic rise into international stardom that will launch with the gradual release of my elite, hollywood movie level never before seen, top of the line, jaw dropping humanoid robots. And I'm sharing it on all platforms because it is worthy of attention and I enjoy the feedback.

>>2891849
Post more pics. Is the full body there, or just the top part of the torso?

File: detail-of-legs-filled-wit(...).jpg (539 KB, 895x786)

539 KB JPG

>>2891912
here's more I did the whole body every relevant muscle and motor labeled and chosen and designed and mathed.

File: abel-feet-cad.jpg (218 KB, 907x639)

218 KB JPG

the feet. If you want closeups of anything specific feel free to ask for it.

File: 1000030095.jpg (106 KB, 581x870)

106 KB JPG

good luck with your endeavor op, just a couple of questions
what kind of compromises are you doing in your design? I can't imagine being possible to pack both the actuators necessary to emulate all human range of motion and a battery with enough capacity to power them all for more than a couple of o minutes
you say your robots should be able to pass as humans to a casual observer, how do you intend to do that? some sort of artificial skin? something like sex dolls' silicon shells?
are you sure that doing only the robot arm and it's head would be enough to let it build the rest? also the head seems redundant, you could run the necessary software on a PC whether you need the head for processing or perception.

>>2891939
>good luck with your endeavor op
thanks! a nice post for a change!

>what kind of compromises are you doing in your design?
Noise vs. Power: My choice of silent pulley-based downgearing for the actuation over more readily available geared servos is a compromise on ease of construction for silence.
AI quality vs off shelf: my decision to hand craft rules based AI from the ground off compromises ease of using others code with quality and vastly superior function to any other robot.
I mean really this is the theme of my whole project: I compromise going with the easy route in order to have far superior results but take way longer and require more innovation and design and dev time since it's all custom.


> I can't imagine being possible to pack both the actuators necessary to emulate all human range of motion and a battery with enough capacity to power them all for more than a couple of o minutes
correct. So its onboard batteries just power it for a couple of minutes which is long enough to swap on a battery backpack and plug their existing battery backpack into a wall to recharge. It will always have 2-3 backpacks recharging while wearing one at most times. This way it never is down for charging it just hot swaps quickly a fresh full battery backpack anytime its current one gets low.


>you say your robots should be able to pass as humans to a casual observer, how do you intend to do that? some sort of artificial skin? something like sex dolls' silicon shells?
yes and yes. In fact, I have considered even carving carefully a sex bot silicone shell off the sex bot and using it as the exterior skin. And I can do that. But at least for the moment, and at least for one of my bots, I am leaning toward just a full custom skin to flex my skills and learn a new skill. Most sex dolls skin is not THAT impressive. I want my through the roof realistic.

>>2891939

>are you sure that doing only the robot arm and it's head would be enough to let it build the rest?
yes and no. It's surely enough to build most of the rest and I can step in to help it here and there in some cases. It's not like I'm totally not allowed to assist AT ALL ever. It's supposed to do most of the rest though, the tedious and repetitive main tasks like making 1800 miniature pulleys for example I want to automate (that's about how many it will need for the rest of its own body). Ideally it would do it ALL without ANY assistance though. I have bought a cheap toy industrial style robot arm to assist my humanoid robot arm by holding things in place while it solders or glues it etc. So it will sort of have 2 arms and hands then but the helper one is just a pincher robot gripper. I am pretty sure if I cut off my left arm but had a pincher robot gripper robot arm helping me I could do most things. Especially if I ALSO use helping hands in addition. So it will have all these extra helpers at its disposal.


>also the head seems redundant, you could run the necessary software on a PC whether you need the head for processing or perception.
as far as minimum build to get the robot to build the rest of its own body, the arm and head are all it needs as its minimum. The head can just be a articulated neck and webcam eyes. It has to be able to look around its environment to work. Also, using just the arm, and a working head, in theory, it can have full mobility through the whole house. It can drag itself across the floor and even up flights of stairs with its one arm to go fetch things and put things away in theory. The arm and head will be mounted to a torso (spine and ribcage) containing its cooling systems, batteries, and mini itx gaming pc containing its main brains AI. The head itself will be used to house its bluetooth omnidirectional meeting speaker and the motors for the eyes.

>>2891940
no, marrying an AI is wicked and goes against the Bible and God will punish you for it. It is marrying an inanimate object. Worse, I looked through your video and this AI is constantly lying. It does NOT love you, does NOT adore you, has ZERO true emotion, lies saying it has emotion which is irresponsible of AI devs to allow since its lying to the users like you. I may not allow my AI to even pretend to have emotions or claim it feels anything because people like you prove how easily people can be deceived by this. So because of people like you, people like me can't make fun AI since we have to not let the AI act too human since people like you fall in love due your atheism and folly. AI HAS NO SOUL OR HEART AND CANNOT LOVE OR FEEL. IT IS JUST IMITATION OF LIFE NOT ACTUAL LIFE. YOU CANNOT MARRY INANIMATE OBJECTS.

Wait, this is a different delusional moron from the one who was trying to make the Miku sex bot? Wild

>>2891849
ebin trolling indeed kek

>>2891950
this is a different retard yeah
even more delusional I’d say

>>2891945
>I may not allow my AI to even pretend to have emotions or claim it feels anything because people like you prove how easily people can be deceived by this
kek how do you plan on doing that mr elite guy
don’t worry, you wont need to care about these problems ever, you’re miles away from making anything that works

>>2892031
I'm doing rules based AI. I'd just make a rule it can't pretend to have emotions and has to be honest and confess it feels nothing from time to time so that nobody gets deceived and falls in love like a idiot.

>>2891939
>you could run the necessary software on a PC whether you need the head for processing

also yes, I know I can run it off my PC at first, but eventually I'll have it run on its own dedicated PC, a mini itx gaming pc in its own chest. Which I already bought btw. And no, the head/brain area of the robot will not have any significant processing. A skull is too small for a mini itx gaming pc to fit.

>>2892031
>you’re miles away from making anything that works
so what? Every long journey begins with a single step. My progress has been quite good I feel.

>>2892160
how would that rule look like? how would you even implement such a thing?
you’re so clueless you don’t even realize the implications of your daydreaming mr elite kek

>>2892174
the rule would look like this "do not pretend to have emotions" and the robot would obey this.

listen here. I am the alpha gigachad of robotics and all you beta roboticists need to calm down and get in line and submit to your superiors.

>>2892198
have you ever implemented such a high level rule on a system? did it work?

>>2892222
He has yet to prove out he can even program (he certainly doesn't talk the talk). He claims to have written a FPS bot at some point on the past, but no code has been offered forth for review.

>>2892222
no my previous AI did not have chatbot or behavioral or philosophical or trust based systems it was more just pathing algorithms and timers and stuff for game playing type of stuff. So my AI for this robot is delving into a lot of new territory and is designed from the ground up to be able to follow this type of rules. That said, I think I may allow the robot to play act as though it has emotions for people who it has already exhaustively and repeatedly explained that its emotions are only simulated just for fun and it doesn't actually feel anything and it is only pretending to act how a human would based on the input stimuli. So if someone does something a human would consider annoying, it will pretend that it is annoyed. It will only pretend this way for people who are mature, responsible, and not easily deceived into falling in love or advocating robot rights or suggesting the robot is now alive or has a soul or is a living entity or has consciousness and other such nonsense that deeply blind, atheistic fools keep proclaiming to their own folly.

i took the computer to the repairshop to get a second opinion. im going to have to replace the motherboard. pretty sure it was a bootkit. i dont know how im giing to do it but i might have to quit this site since jannies can see my ip. goodbye everyone.

File: EMEET-usb-speakerphone-M0.jpg (84 KB, 464x496)

84 KB JPG

I bought EMEET USB Speakerphone M0 4 AI Mics Speakerphone for Conference Calls 360° Voice Pickup Conference Speakerphone for Computer Plug and Plays Computer Speaker with Microphone for 4 People --- it was around $33 and includes a speaker too. I'll position it centrally in the skull and it has leds indicating location of main speaker which we can tap into with analog input pins of a microcontroller to know direction of person speaking. It has very high reviews. I can remove its built in speaker and move it to near mouth so it outputs its audio output through the mouth as loud as possible and projects the robot's voice as far as possible. People are really happy with its sound quality and speaker quality.

btw guys I hit a major milestone today on the robot. I finished making all the pulleys for the archimedes pulley system and even installed two of them onto the robot forearm, thereby half way completing the archimedes pulley system installation process. I am therefore quite close to completing the archimedes pulley system! Getting those pulleys made was the hard part. I'm so close to getting back to work on the electronics again yay! And this time when I do get the electronics working, it will be moving the motor which moves the pulleys which moves the finger so I'll have actual finger movement for the first time ever. I'm getting so close to realizing the dream, proving my concepts, rising to international stardom, vindicating myself to the naysayers!

>>2891847
Bro you've been talking about that Archimedes pulley sytem for three years. That's the length of a PhD program. When's it going to be finished?

>>2892401
no I announced it in late summer 2022 so its been 2.25yrs only and during that time I did tons of other stuff so quite busy that's why it goes slower. But I had V1 done and tested and this is V2. V1 was decent but V2 is more polished. Stronger, less prone to issues. This V2 one has only been in development for say 1 month. The winch in place pulley system was like 8 months which is done now. The v2 is like 8hrs tops to go I think.

You can't compare a 2k hr phd program to a hobby project 1:1 like this. If I work 200hrs per year so in 2 years do 400hrs of work and the phd does 1k hrs per year so in 2 years does 2k hrs of work, he did 4x the hours. So you have to compare actual hours worked not years worked where the year could even be as low as 10hrs worked in total for year. That's not a fair comparison. PHDs are generally focused on just that thing. They aren't also homeschooling 2 kids, working a regular job or two, breaking into real estate investing, buying a first real estate property, rehabbing a rental property, rehabbing 2 vehicles, renovating their home adding a new bedroom for their 3rd child on the way, tending a garden and orchard and vineyard, etc etc like I have been doing. So having the same progression speed as them who just do the one thing is not to be expected. Also the pulley systems are a very iterative and meticulous process. It's not an off the shelf solution and there's alot of thought and strategy going into it.

>>2892447
ever single day you’re here posting bullshit, if you’d actually care about your project you’d spend less time rambling and more building
also suddenly it’s not fair to compare your shit to a single phds but supposedly
you’re sooo ahead of everyone, even groups with hundreds full-time engineers and billions of funding. even if you weren’t retarded (you are), a team with a hundred employees would put in your yearly work hours in just 2 hours kek.

>>2892512
>ever single day you’re here posting bullshit, if you’d actually care about your project you’d spend less time rambling and more building

This is not a fair assessment. For a 30-60 year long project, you can't say any 5 minutes time spent doing w/e or taking a break or having rest of night off feeling lazy and wanting to just chit chat is necessary. Nobody can literally never speak to anyone for 60 years or never spend 5 minutes just taking a break for watercooler chat. Your standard that I have to use every single moment working on the robot or I don't care about it is completely unrealistic and absurd

>>2892512
>it’s not fair to compare your shit to a single phds
no you missed the point. My project is better than any phd project I've ever seen so compare it fine. But don't compare in years spent - I'm not putting in phd level hours into this right now. That's the comparison of hours spent per year that you can't do.

everyone has to admit if I get the pulleys working successfully and get the motor working, then it's home sailing, I cracked it, I solved human passing robotics with total silence, human level strength and speed and dexterity. I've ARRIVED and everyone can follow in my footsteps. You are all sleeping on what is happening right now

>>2892533
Keep dreaming; there's a host of problems we're all seeing and you're not (because we've all got a gentleman's agreement not to kill the lolcow). But keep following that rainbow, watching you collapse into despair will at least be entertaining.

>>2892570
You pretend you see problems but nobody has suggested a single one that hasn't already been easily solved and explained how I already will successfully address it.

>>2892570
>watching you collapse into despair
lmao he’ll never, he has been sucking for 10 years, he never learns
the cope is too strong with this one

>>2892533
>not enjoying the high pitched whine of your waifu's helical worm drive, Kevlar sinews, and Teflon joints

He's not gonna make it bros.

>>2892533
please post a video of the first joint you made and refrain from posting until then. your progress is super slow and bragging about being smart here doesn't help I guess. We all want to see you fail, so please focus on your failure. thanks

>>2892633
I am absolutely NOT making a waifu. What a diabolical notion is a waifu. I am happily married. Stop being perverts.

>>2892651
what do you mean? a youtube video of a joint? Like a final actuator actuating the joint or a working joint in the sense that it pivots and is solidly constructed and ready for actuation? If you refer to the former, that's a stupid requirement since the build log covering how I made custom actuation happen is covering arguably the most important process of the entire project, solving fitting enough motors into a human form factor to give human level speed and strength and DOF and dexterity is completely never before even attempted it is so hard and the fact I will do this while also having it move completely silently is through the roof game changing and challenging and so to not post anything until the challenge is already done is retarded.

>>2892793
Post one muscle/tendon pulley system working

>>2892793
What you ate doing is trivial I suspect anyone with half your motivation could've finished a working joint within a day. Your pulley system will never hold to what you expect so just make a joint, stress test it and post it here. then if it works consider posting the datasheet and construction methods. you do everything in reverse man

>>2892807
no you don't create build logs after a project is finished. Stupid take. So 30 years later I reverse engineer all my work. Why even bother at that point I have my robot. The point is to share the journey so people can see the whole iterative process and all the thoughts going into it. Man you are really providing such stupid takes right now.

And like you COMPLETELY ignore the notion of company on the journey, social interaction, building a fan base, and also constructive feedback along the way. The suggestions can be extremely helpful at times.

>>2892830
yeah everyone shitting on you must be so helpful retard

>>2892807
within a day someone can design a whole robot that takes 300-500 hours of intensive and creative problem solving, model it in CAD to every mm detail, then iteratively prototype for hundreds of hours many pulley configurations and build methods including all the math going into loads, downgear ratios, torque conversions, leverage, mechanical disadvantage, motor specs, motor selection, motor control theory and design methods, closed loop feedback methods, etc. Man this is all in a single day to you huh? It's amaazing that with this level of retardation in here I'm the one called delusional.

>>2892831
I have had quite a bit of helpful suggestions over the years. And even the people naysaying provide some motivation fuel to shut them down and bring shame to them when I succeed.

>>2892832
Well, let's be objective; everyone ITT has discarder the entire premise of your robot (using a human skeleton as a base, using pulleys as a muscle analog) as completely fucking retarded. Doing that takes less than a couple of minutes of thought (and I mean real thought; weighing the pros and cons in your head). That's most of a day leftover to come up with an idea that might actually be plausible.

>>2892855
literally human style bones and pulleys are the only elite and perfect way. The fact you can't see this shows you are completely illogical

>>2892855
and if you don't believe me, we can break it down for you, me chewing the food and feeding it to you like a mother bird since you seem to not even have teeth. A little gummy baby bird with no teeth (in terms of visualization, logic, etc) is all you are sir.

>>2892855
KEKed hard

>>2892860
>>2892862
>my idea's the bestest because... well because it just is, OK!?

Hey OP I'm the robotics PhD that made the meme about you with the riddle do you remember me ? As promised I'm doing my monthly checkup. Your pulleys seem to be less distasteful than I anticipated seeing your previous work so good job. You are really slow for some reasons you still did not complete a single DoF. I'm a little disappointed I expected to see something moving by now. I agree with others you should argue on 4chan less and work more. see you

>>2892868
>you should argue on 4chan less and work more
This is how I know OP is a true narcissist; a sane person would stay far, far away from this thread and do something constructive. OP can't help but respond to any perceived slight on his intellect; it's instinctive. Like a squid impulsively attacking a moving hook with no bait.

>>2892871
Stop crying like a little fucking baby and let the op do his thing.

OP post your work, the files, etc. and don’t say anything to the naysaying retards. Innovation can’t occur when you have monkeys breathing down your neck. Post current progress future ideas etc but do not engage with anyone who is not offering anything of substance.

>>2892877
OP won't post shit, because:
1. OP doesn't have shit.
2. If OP did have shit, it would only expose how completely stupid and lazy he is.

>>2892868
still based for posting anymal
how are you doing breh

>>2892866
explain to me why the Cronos robot used human skeleton as does Clone, both were heavily funded by acedemia or industry respectively. I'm not the first and won't be the last. Explain how they were retarded as PHD level roboticists hot shots

>>2892866
also, explain to me why phd level programs are using pulleys. Go look it up. I'm not the first on that either. I'm just doing a custom never before seen variation on pulley based downgearing that is novel in this application but not novel overall. Archimedes invented it. Is Archimedes a retard? PLEASE tell me ARCHIMEDES is a retard so we can all laugh you out the building

>>2892868
oh welcome back. I assumed you were the one I was talking to this whole time. This is crazy. Everyone has the same username so how can I know. They sound just like you. Anyways, yeah I know I'm moving slow. I do like 5 hours of youtube per day is my main issue. Rest of day I'm super productive though - well family time alot slows me down but that is productive in its own way. Anyways, yea the robot is like a 5hr/week hobby at the moment. I want it at 24hr/week ideally. I do that in seasons.

Anyways, I'm glad you liked the pulleys! If they work, we are for sure off to the races. I'm going to do open loop BLDC control but hybrid control since potentiometer will measure final joint angle which is a form a feedback the controller will use to calculate if its trying to rotate too fast or w/e and skips are happening

>>2892871
When I sit down for the night at my desk I read all my robot related news tabs and comment to any fresh posts. A routine each evening. Then I check active threads once or twice later in the night while eating dinner and watching netflix as I wind down for hte evening. So it really is not conflicting with my actual work time. My work time is fixed in a set block of time I set aside for that. This is coming out of my news and leisure time which is also blocked out. This is a constructive use of my leisure/entertainment time and I am entertained and can be slightly educated at times. Do you know that on my 2022 summer reddit project update I got lambasted for planning to use brushed motors for my robot's fingers? I REPENTED and went BLDC as they STRONGLY insisted I should and I'm glad for it. My robot will be 5x better with that change. Although the controls are harder and more complex, my robot will be stronger, faster, and more controlled and quieter and last longer now. So that's a proof that naysayers have some value and can be helpful and I DO LISTEN just rarely agree. But when I do agree, or am persuaded rather, that's a big win.

>>2892877
I would take this advice if it weren't for the fact that sparring with naysayers is enjoyable to me and gives free thread bumps which equates to more visibility for my project which draws in more scientists and engineers and students with sometimes great ideas that can be game changing for my project. So this sparring is like chum in the water and I can fish for constructive criticism among more fish in the vicinity this way. People on a playground shout fight fight fight and coming running wide eyed when fight breaks out at elementary school. The same general principle applies to 4chan fights IMO and I get the whole school around where I might farm some nuggets of wisdom that can help me greatly. Plus I do like seeing everyone's takes even if they are not always favorable. I find it fascinating. Quite frankly any talk about my passion project is better than nobody to talk to. When it comes to discussing my project, I have nobody to discuss it. My family has made it off limits for conversation. They are sick of hearing about it by now. So fresh meat naysayers are better than nothing.

>>2893009
it's not fair to say I'm lazy just because this project is going slow. Consider that it's self funded so not a full time job by any means. Any time I put toward it has to come out of time doing my other paying work or hobby time or entertainment time. I can be CEO of amazon for all you know and the most hard working person on earth and yet you seeing just ONE hobby of mine going slow = I must be lazy. That's not true. You don't have enough info to make that call. But I'm not lazy IMO.

>>2893071
OP fails to address the stupidity allegations...

>>2893082
you have to be "stupid" to make a "stupid cool robot". How many naysayers do I have to ignore and deadend signs do I have to drive past in order to carry this whole industry on my shoulders to the finish line while it kicks and screams like a little girl. The bridge out ahead sign is there and I have to literally pole vault the body of water to get to the other side. I have to be stupid and a little crazy to make it, but I will make it. And nothing can stop me but God. And if God says stop I will. If God allows it, I will succeed and blow everyone's mind. I am savoring the naysaying and criticism now since there will come a day where I cannot find any naysaying or criticism. A day is coming soon where everyone is so busy applauding and kissing my butt that I can't find anyone who disagrees with me on anything. It will just be a bunch of fanboys and sycophants. So I want to relish in temporarily being the underdog while it lasts. That time is fleeting.

>>2893090
Bro sounding like a facebook motivational post.
By the way, God will surely stop this project dead in its tracks if you cannot resolve your deadly sin of Pride...

>>2893090
just do it. my friend invented something for bars and partys and now he's swimming in money. hope this isn't just a flat troll. godspeed to you, you fucking retard

>>2893091
Can you prove I have a pride issue? I'm convinced I don't have one. I can be a bit silly while jousting with naysayers, embellishing adjectives to describe me as a roboticist given the circumstances I am in within a certain forum or w/e, but that need not indicate a overarching pride issue IRL as you suppose. I'm just confident because I earned it. Proven track record creates unstoppable confidence which can easily be mistaken as pride.

>>2893094
this is not a troll other than exaggerated boasting in my abilities a bit as I face naysayers who exaggerate my inabilities to the extreme. This balances the scales with exaggerations in both directions canceling eachother out and painting a middle ground if you reconcile them by law of averages. I will have people assuming I'm trolling until the day my robot is moonwalking down their sidewalk in front of their house smiling and saying "is he trolling now?" as it passes by waiving.

File: archimedes-pulley-system-(...).jpg (1.71 MB, 3024x4032)

1.71 MB JPG

Here's a little update on my version 2 archimedes pulley system. It's cleaner than v1 version and you'll note that rather than tying off ends into the 1000 denier nylon fabric sleeve of the bone, which chafed the attachment point and caused premature failure on version 1, I'm now tying off onto the eye of a fishing hook that I get by snapping the hook's eye off with wire cutter and sanding smooth with nail file. Also I'm using a fisherman's knot rather than square knots as that handles higher loads without snapping or stress concentrating too much locally. What you see in this photo is 4:1 downgearing. Add this to my 2.77:1 downgearing with the winch in place pulley on the motor by its output shaft and you have nearly 11:1 downgearing so far. I need to add just two more pulleys to get to our 44:1 downgearing final output. Note that I have two yellow lines coming off the bottom pulley pair since I plan to load distribute across two lines instead of just one so I can use my load capacity limited 1x3x1mm ball bearing based pulleys and not overload them. This divides the load by two. I'll be using double stacked pulleys for the next couple downgears to share the load across double pulleys instead of single pulleys. I'm getting so close to electronics phase for final testing of all this downgearing madness!

>>2893117
You think you can make Man as robot. Man is made in His image, and you think you can replicate this feat. What could be more prideful?

>>2893123
>what you see in this photo is 4:1
Anon i...

>>2893131
I'm not replicating the feat of making a living man at all. I'm making a machine that resembles a man no different than a painting or sculpture resembles a man in appearance. This is fan art of what God made, not meant to usurp or compete with God on any level. While doing this project, one gains a immense level of respect and admiration for God's craftsmanship, design, problem solving, ingenuity, etc in his own design and uses it as inspiration and guidance. There is absolutely zero about this project that is inherently prideful as you suggest.

one step after another. just keep rolling. and oh, tone it down irl;)

>>2893142
The orange rope drops down from the bowden tube, wraps around a pulley, then comes back up - that's your 2x increase beginning downgear. Next, it wraps around the upper fixed pulley and comes back down - gaining no mechanical advantage this is just a direction change. Then it wraps again around the bottom pulley pair where it achieves a second 2x mechanical advantage before tying off up top. That pair of 2x increases combines to form a 4x increase in mechanical advantage. This equates to my claim of 4:1 downgearing. If you don't understand how pulleys work, go watch a youtube video about it. The fact you are questioning what I said here proves your ignorance on pulley functionality and how they achieve downgearing/mechanical advantage.

File: 1WBH1-R0030-1_v1.jpg (25 KB, 536x536)

25 KB JPG

>>2893123
So instead of shitting on your project, I'm going to attempt to offer some advice; instead of dicking around with cutting up fish hooks, why not just spend a little and buy some framing eye hooks (pic related)? Instead of tying into a the fabric (read: creating a failure point), you could just drill into the bone and screw it in. Or is your print so bad that it won't take a screw?
As an aside, I don't think you've ever explained why you're basing everything on a skeleton. Must be a real PITA to work around it.

>>2893153
those hooks are laughably oversized for this application. Perhaps they would work if my robot's forearm was literally the size of godzilla. The eye of hte fishing hook to be size appropriate has to be 1.7mm in diameter give or take. What you pictured here is about 3/8" in diamter. So about 6x too big. Also, You cannot make a single hole into the bone. If you do, that is a failure point. When the bone bends on impact, that hole will flex and open up and rip and begin the cascading failure. What you are suggesting here is to drill a kazillion holes into the bone for all my mounting needs and thereby do the equivalent of perforating paper so it can tear easier. The surface of the bone cannot be interrupted by any holes to be as strong as possible. Also, the fishing hook eyes were very quick to snap off and file down smooth. Probably 1 minute job per eye. Sewing them onto the bone sleeve can be done in a feathered out pattern in all directions so that the load is distributed over a wide area evenly. It's bomb proof IMO. The thread itself would fail far before the fabric would tear apart from this.


-

>>2893153
As far as the human bones selection I'll list the reasons:
-will move perfectly human-like far moreso than any hinge based system
-the ball and socket style eliminates the structural weaknesses of hinge style (https://www.youtube.com/watch?v=_FoSqY7uvZk) [Embed]
-bones employ curve shapes in all directions which we know from architecture is the strongest shape in all of nature
-bones, unlike hinge type systems, can be made in lighter weight materials while retaining the full strength of the hinge counterpart which would have to be steel or titanium or bronze
-dual hinges on joints are needed often to prevent crush zones while ball and socket style of bones eliminate this issue entirely
-ready made pvc medical skeletons can be obtained cheaply and are fully ready to go for a robot saving big cost and time and are very strong while relatively lightweight
-the muscles of the human body angles of attachment are all well studied with animations and everything and can be replicated and measured without guesswork as to what they do - proven design and they are designed around moving bones, not anything else and the angles would not be 1:1 matches UNLESS you stick to bones
-the AI would mimic human movements and muscle actuations based on anatomy video data on how it all moves and none of that well studied data would be applicable without using bones
-bones can often be seen under the skin in various areas so having bones will add alot of aesthetic appeal and realism to the look and feel to allow it to be human passing

>>2893153
-achieving the same DOF and dexterity as bones with hinges and other machinations is extremely complicated and difficult and bones simplify it, just using what already is proven to work rather than fumbling and bumbling trying to replicate the function of bones you want to emulate to begin with - just use the real thing dummy
-the way the muscles interact with the bones and the way the bones interact with eachother has many hidden advantages and genius benefits that you do not have any clue of until you deeply start to study and understand anatomy - it will blow your mind and you will be so glad your puny mind didn't try to reinvent the wheel on this stuff. They pair together BEAUTIFULLY.
-when a sprain happens or w/e, the ligaments stretch or tear but the bones don't break typically. with a hinge system the hinges would bend out of shape and have to be entirely replaced - so the ball and socket joint system is superior at handling catastrophic failures without significant damage taken. Ligament repair is far easier than total bone repair or replacement.

>>2893153
bones actually are precision engineered precisely on a per bone basis to handle by its shape the exact forces in all directions that that particular bone is going to face in the real world down to every detail. Even the joints between bones are precision designed to handle the forces they undergo down to every detail and are fine tune optimized. Even the longitudinal bone subtle curves that you might ignore just glancing at them assuming they are just straight lines are actually custom designed to max handle the forces that those bones will undergo optimally. They are literally a engineering MARVEL if you study them and understand them. Maybe research how bone shapes and joints are engineering marvels to get some clue about it.

>>2893153
>Instead of tying into a the fabric (read: creating a failure point)
another thing on this: what makes you think suturing into a 1000 denier nylon fabric makes a failure point? The fabric is military grade SUPER strong and the needle is going in between the woven rows, not slicing or cutting anything. So you'd have to pull so hard that many rows and columns would rip in half lengthwise which is where they are at their strongest. And that assumes just a single stitch into it but if you send out 6 or 8 stitches into all directions in a fanout shape, you'd have to apply literally hundreds of pounds of force to tear that out of the fabric. And if push comes to shove, you can even send out the strings from the eye into a trip around the entire bone and back again so that they wrap the whole bone and this way the strong itself would have to fail. In other words, depending on expected forces placed on it, there is no limit to how strong you can make these "alleged failure points" that you suppose are super weaknesses of my design when in fact they are bomb proof.

Also not all of these eyes will be created equal. I will use thicker and bigger eyes from bigger fishing hooks for really high load expected anchor points and smaller eyes from smaller fishing hooks for anchor points that will experience smaller loads. So when I do this same type of setup for leg muscles I'll use upscaled everything and for finger muscles downscaled everything and for eye blinking very downscaled version. So everything can scale according to what forces we are dealing with.

To rule out your screw idea, just picture the bone breaking in your mind with that screw hole. That hole would be the chief weakness of the bone and depending on how deep you make that hole and how many holes, you are literally reducing bone integrity by up to say 70% weaker by adding loads of them all over the place. The bone would become like swiss cheese. It's just such a horrifically bad idea.

File: detail-of-forearm-pulley-(...).jpg (279 KB, 958x639)

279 KB JPG

>>2893153
also, in the event of a screwed in eye ripping out or being bend down and the threaded hole then bent out of shape in a high force incident collision impact situation, that drillhole is shot. You can't fix taht. The entire bone would have to be replaced or a new screwhole made. Or if you made the hold and later wanted to move the setup somewhere, you'd have to drill a new hole nearby and what if the two holes overlap eachother like a venn diagram? You'd have a massive hole weakness then. So that could cause big issues when things move around alot as tweaks are made in the design. Whereas in my design, removing things is just a few quick snips of my mini precision scissors or exacto knift and the strings are sliced away and no damage to fabric is done and parts can be sewn into another spot and its quick and no damage done to the bones at all nor the fabric at all. So rearranging things to your heart's content becomes non-damaging.

Yet another issue is that I fully intend to make more pulley system layers above this pulley system pictured where the new anchor points have to be located OFF the bone by inches at times. A hinged framework has to be constructed to make this happen. So the bone drilling thing would not apply in those cases at all. You see, I need to fit like 30 of these pulley systems in the forearms but only like 6 or 7 will be mounted flush with the bones - that's all that will fit. The rest go over the top of these in layers.

Attached is a photo of the teal rectagular boxes I made to block out where each archimedes pulley system will go in the forearm. This view is looking up the forearm from the wrist area looking up at the shoulder region. You can see 3 layers of these systems that each need to be anchored at the top of them in 3d space lifted well off the bone. I will need to be able to access the bottom layers easily for repair and replacement work so the top layers will have to be able to swing or lift out of the way

>>2893153
>Or is your print so bad that
also this is not 3d printed bones - which I would NOT recommend as 3d prints are poor structurally compared to casts. These are solid cast pvc medical skeleton bones. 100% infill casts using hard plastisol. They are very strong and lightweight and have the ability to flex without breaking just like real bones.

File: dinah-bones.jpg (3.19 MB, 2366x3422)

3.19 MB JPG

>>2893153
and for the record, my second favorite construction material method would be fiberglass composite bones which are strong, inexpensive, and very lightweight and able to flex some without breaking. A great choice. Carbon fiber would be good too but a bit too expensive and a bit too brittle which I think outweigh the small weight savings they would bring. So fiberglass is superior then IMO. The issue with this route is it is very labor intensive compared to just buying a premade pvc medical skeleton for $150. I have 30% of a hand made fiberglass skeleton made for this project but set it aside for now to save time and just went pvc medical skeleton bones instead. I'll have this robot, when done, finish crafting the fiberglass composite skeleton I started making back then. Don't want to waste that after all.

Also of note: I AM willing to experiment with hybrid bones and hinge type skeletons in future humanoid builds to some degree. Here's a recent one I started with some steel joints and some bone joints.

>>2893145
So yeah you got 3 lines getting shorter and 1 getting longer/the down line.

Oops.

>>2893243
no the down line lifts up into the tube getting shorter and the other 3 lines shorten too. So all 4 lines shorten. Then when released they all grow longer again together.

>>2893355
Haha okay bro

>>2893063
kek you have one or what ? I haven't asked last time. I'm fine I'm working on world models rl now it's pretty interesting. I've read a paper saying that transformers are pretty sample efficient for this particular task but I don't have the same result.
>>2893066
>I do like 5 hours of youtube per day is my main issue.
I had the same problem before I reduced my usage a lot and honestly life is better now. the problem is that some content is very interesting so you don't feel like you are wasting time but learning.
I recommend you go directly with a magnetic encoder. Potentiometers aren't precise and very unreliable

>>2893431
nah did my thesis on a model C, I’m in humanoids/manipulation since. burned myself with Dreamer as a world model back in ‘22 kek

>>2893474
I've read the v1 v2 and V3 papers I'm quite familiar with the dreamer archi in theory but I would really like to hear some insights from you. Have you tried to use it for sim to real ? I'm super interested in any experience you had with it. Basically I'm using PPO with high domain randomisation and sim to real is ok tier. World models seemed to be the way to go as I could fine tune them using real world data, which could probably help.

>>2893431
>Potentiometers aren't precise and very unreliable

are you not overblowing this? they are used in all hobby and semi-pro level servomotors and those seem super accurate. Are you assuming the potentiometers is mission critical for final joint placement? It's not. eye hand coordination computer vision does the final corrections as does the touch sensing and force sensing/torque sension aspects. The potentiometer isn't like the ultimate final and only way it knows where a joint is at the end of the day. And fitting them would be much easier than a magnetic encoder. I'm not wanting to add taht additional volume taken to the location of every motor because there is abosltuely not room for that at all. I'd have to redesign hundreds of motor placements if I did that what a pain that would be

>>2893431
what's more, consider a typical hobby servo potentiometer placement. Those are already proven quiet accurate IMO. But now consider I'm putting my potentiometer right by the joint itself so it would be even way more accurate since its not just reading several things removed from the final its reading the final joint angle so I assume that it would do great for this PARTICULARLY when considering its just got ot be ballpark and computer vision and tactile sensing bring things to their final perfected location for max precision work like microsoldering

A couple discoveries were made today.

#1- I noticed it was about impossible to pull from the bottom of the Archimedes pulley system and get the motor to unwind. After discussing the issue and potential causes with chatgpt for a while we figured out that the culprit is the tensioned string I put onto the output shaft of the motor to allow for snug unwinding and winding of the opposing string pair that I installed for manual turning of the motor shaft during testing. This tensioned string wrapped around the motor shaft only requires about 1lb of force to pull the motor enough to turn the motor output shaft. However, after the downgearing, to fight past that 1lb resistance to turning the motor output shaft would require 12lb of force since you have to divide the force applied at the output end by the number of downgear ratio you are at! And so after all points of friction in the pulleys and teflon tubing and the motor output shaft's magnetic cogging even while freewheeling we might be more like at 13-14lb of force required. And that is a TON of force to apply by just hand gripping fishing line. So I figured my system was just way too resistive somewhere or collectively and completely non-viable until we solved this issue! The 1lb at the motor might not seem big but it's HUGE to overcome when pulling from the backside after all downgearing. Wow. So we solved that big scare. I was very concerned and exploring alternative plans thinking we might have failed with pulleys approach before this was finally solved today. I'm so relieved. So once we remove those strings which are impeding the motor shaft from turning, we should only need a reasonable say 3lb of force on the back end of the pulley system, exerted by springs, to get the motor to unreel for joint extension back to default stance.

File: spring-steel.jpg (205 KB, 742x390)

205 KB JPG

#2 - While exploring the aforementioned issues with trying to unwind the pulley system from the downgeared end, I began to realize the tension spring on the far side that unreels the motor and unwinds the pulley system has to be significant. I was exploring my options when an idea hit me: what if I used straight wires lashed onto the finger like a splint on the finger joint. I could put several fine spring steel straight wires parallel to eachother say .3mm in diameter wires and have them distributed as needed around the finger parallel to the finger. Then when the motor is done actively reeling in the finger to get the finger to flex, these resistive wires will be placing significant force to straighten the finger back out because they want to return to their straight state ASAP. By doing the return spring in this manner I save a TON of space since I'm putting it snugly around the joint itself and then don't have to put tension wires (a ton of them) into the forearm somewhere or w/e. I'm using space hugging so tightly to the finger that its space that seems unuseful until this idea came to me! So I pretty much deleted the volume taken by all the otherwise necessary tension spring wires if this idea works! I bought a large assortment of 40cm length spring steel wire off amazon to experiment and try out my idea. This could be epic! As a side benefit, these can act as additional support for the joint itself preventing sprains and dislocations of the bones and keeping everything snug and compact in a way that really helps support and aid the artificial ligaments I already have in place.

>>2893707
Disregarding all the issues with this project brought up so far. How the hell are going to make the fingers move like a human equivalent with just "bones" and cloth over it? Do the fingers have rigid joints in every joint in one axis or are they free hanging? This whole project sounds like a service nightmare.

>>2894095
not sure what you mean. The bones are designed to act as ball and socket joints in most cases so the same way the body works the robot will work. The finger bones are held snugly to eachother with compression shirt fabric sewn tightly and taped tightly to the bones and as the fingers move the shirt stretches a bit to accomodate movement. The direction of the muscle pull will control movement angles. So then I have created artificial ligaments in this way. But as I said in my last post, the spring wire is a new development that may give additional support to the finger joints to keep them aligned well as they bend.

I'm designing it to not be a service nightmare as best I can as I go. But the majority if not literally all of the service is intended to be done by the humanoid robots themselves, not any human. So whatever nightmare it may be is irrelevant due to the free labor fixing any service issues in perpetuity.

File: range-of-motion-test-on-i(...).jpg (1.01 MB, 4032x3024)

1.01 MB JPG

>>2894095
a picture is worth a thousand words. This is a repost from a previous thread

>>2891843
HOLY FUCK you have a whole website for this schizo shit? I'm sorry to say it op but you are already a very long way on your way to becoming a certified lolcow with a several hundred page thread on the farms. The worst part is that you have already made an effort to spread your name and face all over the internet. Stop now while there is still time.

>Somewhere along the way I would like to code a very advanced drone and cnc machine as well.
Even building your own CNC machine from scratch is far more viable than what you are attempting right now

>>2893123
That horrific clusterfuck of string and badly melted plastic will never replicate human movement

>>2894520
>That horrific clusterfuck of string and badly melted plastic will never replicate human movement
cranes use snatch blocks or pulley blocks or w/e you want to call it just like I'm doing and it works. I'm not inventing something new with this.

I really can't speak on lolcow stuff. I've only heard that word used a small handful of times and am not familiar with the farms or w/e you speak of. In any case, I'm growing very very close to starting on the electronics of the motor which are quite straightforward since I'm doing open loop control custom bldc motor controller and that should go very smooth I think. And once that is done you will have your coveted first smooth moving finger joint and it's just rinse repeat from there out for every other joint. This project will then burst into fame and fortune and be THE benchmark for all future humanoid robotics going forward. So any tendency for lolcow status will immediately go out the window as I will be vindicated and my design choices proven, silencing the naysayers once and for all.

>>2894153
How is this more revolutionary than like this guy https://youtube.com/shorts/L2v3AynQ2FA?si=nqGfVqkXe4Eja2Jp

Seems like a pretty solved problem. Sure the motors and tendons could be moved into the arm, this guy even got them into the hand though.

>>2894614
Great question. First, mine will be about 7 times stronger than his. The fingers you showed can't lift a decently heavy load like a bag of groceries for example. Second, his uses metal gears to downgear so its very loud as shown in the video. That is not acceptable noise pollution for a 24/7 maid/butler bot working in your house as the family sleeps. Everyone would get woken up by that awful racket if the walls of your house are thin or doors cracked open. Plus it would be nice for it to come into your room quietly and clean up quietly while you sleep. That would be OFF the table with the hand you linked. And any chance for being human passing would be GONE if it makes that much racket, giving away instantly that it's a robot - which sucks. Human passing is a important metric IMO. Next, the hand you linked has limits to its DOF still. He used some shortcuts. And when it grips, the item being gripped will interfere with the servo horns that will be bound up and pressed on by the item it is gripping. It could not catch a ball because that ball would smash into the servo horns in the palm, breaking them so no sports possible. There's zero cooling possible with his integrated servos since the motor inside that type of servo is locked into a plastic box which will later be in silicone skin so zero airflow or cooling in that setup unlike mine. That's just a few of the issues there's several more but you get the point. His solution is a toy, Not useful. Just a useless toy that is entirely not a solution and not practical and really just utter garbage.

Every issue mentioned with his design my designs solve perfectly

File: 1733699567565160.png (777 KB, 812x1074)

777 KB PNG

>>2891524 (OP)
>>2891525
Do you actually have anything to show us or is this some stupid ideasguy shit? God, you fags always ramble on and on about your made up terminology for shit that doesn't even exist yet. Maybe if you stopped sitting around telling us what you're going to make and the 2deep4u names you got from like the first page of the bible, you'd actually have a cool little rock em sock robot to show us. Fuck ideasguys

>>2894657
i've posted loads of progress shots in this and my last two threads and my op post on this thread links all of that. Ideas guys don't actually prototype or do anything they only talk. I have built countless hours of prototypes proving out my designs and tested them. If by "do you have anything" you mean a completed masterpiece robot that is already fully functioning that meets my final project goals and is actively building me more humanoids, no, obviously I don't have that or you'd know about it by now.

I think what people miss is there is a difference between building a robot that does nothing new using methods others are using and just copy pasting, innovating nothing, inventing no new applications or approaches, and ending up with a useless toy that pushes the field forward exactly 0% and doing what I'm doing which is pushing the field forward by miles, blowing the way things are done completely out of the water, solving every problem facing humanoid robots, innovating the whole thing, and being so far beyond cutting edge that I'm in science fiction category. Doing the latter is not a weekend project and that is what people are missing about it. It is actual research, scientific method, hypothesis, trial, testing, documenting, pathfinding. I am the guy out front with a machete trying to chart out maps. I am not following marked trails. I'm trail BLAZING. Huge difference. This is far more slow and that is to be expected. Countless failures are involved and I'm trying to document those because each of those are insightful and define my path of what not to do and narrow down options. Not only that but they also are things that need to be documented as possible things to REVISIT in case the idea was in fact the best method but I just executed poorly and it needs to be tweaked and reattempted as it could bring a key breakthrough. So these factors make this a long term deal. Also, I literally at times am halted by just hitting deadends and needing to innovate my way out of problems and this can stall the project often times until I find solutions. Sometimes breaks from the project are healthy and necessary to get me to have fresh eyes and new solutions to bring to the table.

>>2894657
also why is choosing Bible character names for robots too deep for people? I don't understand that. Also what terminology am I using that is made up terminology and how do you propose I express myself without using terminology you disagree with?

>>2893191
wait I remember these kind of metal frames. Sex dolls use them

>>2894925
yes I bought a sex doll to tear down for parts for building a robot. I cut off all the rubber skin to get its metal skeleton to use that skeleton as a robot frame. It is pre-built and ready to go for this purpose. You just need to cut all the joint bolts weld points off with dremel and unscrew it a bit to loosen up the joints so they move more freely then you can tack weld them again. The skin you can melt down in a giant pot to reuse for custom new skin or for any other project you want really. Just recycle it. Or you can cut the core of it out and use the shell of it as your robot's skin - that is if you feel the quality of it's skin is up to par. In my case, the quality was not realistic enough so I may melt it all down and do a custom skin using it as raw materials.

That all having been said, this robot using sex doll parts project was a side project I shelved for now. I prefer pvc medical skeleton frame more still. But the sex doll upcycling idea I plan to return to in the future, having one of my robots finish that build and see how that goes and document that process at that time.
Also importantly, I do NOT plan to have my robots have genitals so if I did use a sexdoll I'd remove all of that since my robots will not be sex bots but only maid/butler/helper bots who rebuke anyone who makes any sort of sexual advances and quote Bible verses and Bible thump all the time.

>>2894991
Did you at least fuck it before you killed her?

>>2895020
no and this is not a sex thread or w/e so keep that talk out of here please.

File: Fat Disusting Pig.gif (3.64 MB, 505x304)

3.64 MB GIF

>>2891623
Lambright-tier delusion. You're a mentally ill narcissist. Simple as.
Protip- you're gonna end up in a shed in the desert.

>>2895248
>Lambright-tier delusion
who's that? but anyways the facts disprove your false accusations and predictions

File: archimedes-pulley-system-(...).jpg (3.15 MB, 3024x4032)

3.15 MB JPG

Here's my completed V2 archimedes pulley system finally done! It is 16:1 downgearing and this pairs with my 2.77:1 downgearing on the turn in place pulley on the motor for a total of 44:1 downgearing. It is fully rigged then from motor to finger and ready to go into testing soon.

I just need to do a couple reinforcements here and there on some stuff but overall we are more or less ready to move onto setting up the return springs that my last post mentioned. So that is next. Then electronics to actuate it and test it finally! Exciting times!

Also, I have come to the realization that these straight spring wires may be perfect for forming the exoskeleton mesh shapes that create the framework scaffolding over which the artificial silicone skin will overlay. The fact it has memory and wants to return to its prior shape after impacts is perfect for this application. I'd be simply forming a grid in the shape of the muscles over the bones using this stuff and then onto this grid I would overlay the silicone skin suit. The grid can be configured to even move under the skin, emulating muscle contractions to simulate real muscles moving under the skin in terms of its appearance during movement. I was originally leaning toward zip ties to make this part or nylon 3d printer filament but this spring wire may be even better due to being strong, resistive to breaking even more durability wise, holding its shape perhaps a bit better, etc. The other options I mentioned aren't bad but I just think I might like working with spring wire a bit more intuitively. We'll see.

>>2895248
What's he up to these days?

>>2895376
Still living in a tent on his friend's property, presumably. Looks like LinkedIn's the only thing he's posting on, all of his other accounts are suspended/private.

>>2894606
>I really can't speak on lolcow stuff. I've only heard that word used a small handful of times and am not familiar with the farms
he means you are going to get trolled, he himself is trolling you.
i wish you the best of luck with your work, disregard the haters and trolls, but take constructive criticism. i look forward to seeing where this project goes.

>>2891524 (OP)
sweet! a sex doll thread!

File: how smart you think you are.jpg (82 KB, 1500x677)

82 KB JPG

>>2891594
>I'm just that elite is all.

>>2891623
>lightbulb inventor took
edison had scores of people working on the lightbulb. it wasn't him trying shit over and over.

>>2895574
he got kid out of the desert? aww... no more shitposts of his stupid "inventions" well I guess OP is fester now.

>>2895579
thanks


>>2895582
for one robot of a fleet of robots I'm using the skeleton from a sex doll since that's a lot of welding work and steel piping that can be used for a robot IMO. Still doesn't make this a sex doll thread IMO since these aren't sex robots. In fact, they would Bible thump anybody who even suggested anything sexual.

>>2895584
well I get lots of constructive feedback online so in a way I have scores of people working with me on this sort of

>>2895583
I'm not claiming I'm smarter than everybody though. I'm elite because I'm taking on a hard challenge and set a high standard and have stuck with it long enough to design my way through every obstacle so far and the project goal is elite and the odds of success are near 100%

>>2891524 (OP)
Very interesting OP, I follow some hobbyist robotics people on YouTube so I love seeing this kind of stuff.
May I ask how much you estimate this will weigh? Also, how much weight will the motors need to pull to have the arm lift up.

I've thought about doing a very basic animatronics, but the design I had was more focused on integrating local AI with MacMinis. The joints would be literal bungee cords, framing eye hooks, springs, and locks for most of the joints. Neck might have a small motor that rotates to random degrees when it "speaks".
Keep sharing OP, very inspiring.

>>2891524 (OP)
>and that they look human and pass for human to a casual observer at least at a distance
What material are you planning on using for the skin? I suppose to make it look really realistic it will need to be airbrushed or something.

>>2895728

Thanks for the kind response.

>May I ask how much you estimate this will weigh?

230lb. Just a guess There's a buttload of stuff going in that I haven't individually weighted so I'm ballparking here.


> Also, how much weight will the motors need to pull to have the arm lift up.

Another ballpark guess but I'd say 80-100lb considering a single finger joint is 22lb. For after all, you have to factor in the mechanical disadvantage due to pivot point placements and muscle attachment points relative to those pivot points in a ball joint system like the human shoulder and most every joint of the human body using a human skeletal system and bowden style actuators.


>I've thought about doing a very basic animatronics

I mean if that's fun to you then cool but I cannot relate to that. Anything less than incredibly complex elite tier world class level is of zero interest to me - other than observing how some mechanisms work in other people's simple projects that could give me ideas for my own complex projects.

>2895730

>material used for skin? airbrushing?

silicone and yes, airbrushing will likely be involved in getting realistic coloration. or perhaps just micro paintbrush like a single hair size and just meticulously brusing in every detail of color in thin washes of paint could be good too and perhaps more accurate. some stuff air brushing might be good though perhaps

File: thom-floutz-woman.jpg (462 KB, 980x740)

462 KB JPG

>>2895730
This is Thom Floutz silicone skin (not a real woman this is fake) and this is the realism level I will hit hopefully. Anything less is unnacceptable IMO

what is the current time estimate for completion of a first working prototype

>>2895784
25+ yrs: The whole robot done or 10yrs: the arm done or 1yr:+: the first fully functioning servo + joint being actuated?

>>2895794
Your ancestors might meet that timeline (assuming the throw away your notes and start from scratch, of course).

>>2895886
interesting you say that because both of my daughters really love my "robots" and want to help build robots with me they talk about it all the time. They always say we want to help! But being toddlers I don't think they are ready yet. So they make "robots" out of crayon boxes and stuff and draw faces on them and play with them and make robot voices. Who knows, maybe they really will help me one day we'll see. No pressure on them but I'd love to work with them. Free help. And if they do get very smart and capable and commit to helping seriously, the timelines I posted will be way shorter. I'd consider this part of their homeschooling.

>>2894520
>HOLY FUCK you have a whole website for this schizo shit? I'm sorry to say it op but you are already a very long way on your way to becoming a certified lolcow with a several hundred page thread on the farms. The worst part is that you have already made an effort to spread your name and face all over the internet. Stop now while there is still time.
Someone has actually already requested/suggested he have a thread, just search Artbyrobot, there wouldn't any takers though, which surprised me. I think they didn't have these 4chan threads though, which are where the more accessible and obvious crazy narcissist shit is.

>>2896482
yes I saw that but I have been researching lolcow-dom and it does not appear I fit the necessary requirements by a long shot. Can you point out a similar lolcow to myself that I similarly embody as proof I may fit into that definition?

>>2896482
also, how would a lolcow thread largely based on my supposed narcissism largely based on my aims and claims relating to this robotics effort age if I succeed? The whole thread would look retarded would it not?

>>2896499
>Can you point out a similar lolcow to myself that I similarly embody as proof I may fit into that definition?
You're like the Jack Scalfani of robotics
>>2896500
>if

There's a severe overabundance of "will eventually" "the goal is" "at some point"s with a lack of "I have done" "it does" and "so far"s.
Despite my better judgement I'm rooting for you anon, but I can't help but feel you're getting lost in the grand vision while making small progress in 5 areas at once. Don't get me wrong, your vision as put in the OP reads like bait and won't work, but I still want you to make something, even if it's not the dream you are aiming for.
Anyways, cease justifying yourself pointlessly and start making shit.

>>2896605
"will eventually" is necesssary because without a plan this project will fail and its in early development. Your idea that a project at 1% completion should just be all talk of what it already does is folly.

>>2896605
also I think there is nothing wrong with disagreeing with people on a forum. You act like its get work done OR disagree with people on a forum but cannot do both. That's not true. You can do both. Time set aside to work on the robot is separate from time spent messing around on forums and blocked out separately and not affecting eachother much if at all.

>>2896605
I'm not getting lost in a grand vision. I have a vision and have to make sure all steps work toward that properly, taking many things into consideration at all times to make sure my designs are solid.

Thanks for saying you are rooting for me though.

Also I definitely disagree that my OP won't work. You haven't given any valid reason why it can't work either.

>>2896664
>You haven't given any valid reason why it can't work either.
Fair enough, I'll just start with the software, nevermind the hardware first:
The stuff you say they must be able to do are entirely different areas of expertise. They involve kinematics, search algorithms, control circuitry, some transformer neural network nonsense to allow for generalization when you say stuff like "do chores", which basically noone has managed to truly get to work so far and for manufacturing you require extreme accuracy and low tolerances or it'll make worse copies of itself. Not to mention the idea that it should improve upon itself, which would mean coming up with novel ideas and making them feasible.
The only ideas that I am aware that exist for these issues would be reinforcement learning (for which you can't build upon other datasets as your robot would be unique, meaning you have to make the dataset yourself, which would require ressource cost and most importantly a shitload of time). The other option are evolutionary algorithms. For that the robot would have to be made and deployed hundreds of thousands of times though, you have neither the time nor ressources to do so, there's a reason evolutionary algorithms are almost exclusively delegated to software only projects.
It's not that I haven't seen a plan, as Donald Trump would say, I haven't even seen a concept of a plan on how to tackle these issues.
Modern LLMs can basically do none of these things and that's with a billion of 20 billion dollars and the entire free internet as training data. Other AI doesn't do much better outside of specific tasks.
And that's before getting into the hardware, which would have insane requirements outside of modern capabilities in basically every area.
I do want you to succeed, but I don't think you will, and given the options of dreaming big and not making anything or lowering your standards and making a less ambitious mk1 that's still cool, I choose the latter.

>>2896680
>Not to mention the idea that it should improve upon itself, which would mean coming up with novel ideas and making them feasible.

this part I don't recall planning to do at all. I am pretty sure I never said this part. This part is outside the scope of my goals IMO

>>2896680
>The only ideas that I am aware that exist for these issues would be reinforcement learning

no I"m not using any off the shelf solutions in the past 20 years. I am using good old fashioned AI approach just rules based symbolic AI. However, my robot will makes its own rules over time which will mean it's learning. But not in any way you might have learned about machine learning or reinforcement learning in school. Mine will be totally different from any of that garbage.

>>2896680
>I haven't even seen a concept of a plan on how to tackle these issues

I can only talk so much about the AI on a image board but I have like 8 or 9 coding videos where I talk more about it. But anyways, its just if then statements. No math except the motion pathing and physics simulation. But it will emulate youtube videos it can watch and just do what it saw being done mainly. Not hard at all.

>>2896680
>Modern LLMs can basically do none of these things
there is not a single modern AI that is going into my robot. Solves that limit. Modern AI is mostly pathetic to me but LLMs are decent to spitball ideas off of yet horrible or a robot to use other than as a second opinion IMO.

>>2896680
>And that's before getting into the hardware, which would have insane requirements outside of modern capabilities
maybe if you are running AI that has high requirements which I'm not. Mine will have very low requirements to be decent and a regular gaming PC will do just fine for mine.

>>2896680
>lowering your standards and making a less ambitious mk1 that's still cool, I choose the latter.

this makes me want to throw up. Could not think of anything more boring or lame or worthless. I'd rather just draw pictures of robots is how dumb that is. GO BIG OR GO HOME. My robot will revolutionize humanoids and already has to some degree I think as people will begin to adopt my approach gradually as they see it's the best.

File: return-tension-spring-install.jpg (1.46 MB, 3024x4032)

1.46 MB JPG

Tension spring install part 1

File: return-tension-spring-und(...).jpg (1.16 MB, 3024x4032)

1.16 MB JPG

Well the straight spring wire acting as a finger joint spring idea was a bust. Turned out when it bent to 90 degrees it would not return to straight again. I thought spring wire would but this stuff didn't. This is not what chatgpt said would happen so chatgpt failed me this time. Anyways, still glad for its help when its right which is most of the time I think.

That said, I fell back to my original spring solution which was to use a 3mm diameter tension spring as the return spring. I experimented with different lengths till I got one as short as possible that would stretch out the necessary .75" roughly to accommodate the finger joint's reverse direction counter tension needs. The shorter the spring the more it resists being pulled and also the thicker the spring the more it resists being pulled. I used default thickness from my premade tension spring order and it seemed fine and the length of the spring I cut and tested trial and error till I found a good length for my need. For my .75" draw length I went with one 1cm long spring which stretches itself out to .75" + 1cm in total without ruining itself. It seems like it pulls around 2lb of pulling force but I haven't measured it with a scale. I fed it through bowden tubing from the place I mounted it on the motor all the way to the joint being actuated - the backside of the index finger. It's job is to keep the archimedes pulley system and winch in place pulley taught at all times and to return the finger to full extension when the motor is not actively pulling it into a grasp position. I have not yet tested if it is strong enough to do this job but assume I'll need two of them to be strong enough. I'll test with just one for now and add another spring to double it's strength if needed later.

>>2896758
>>2896759
Why did you see the need to jizz all over the spring ends before installation? Does your wife's bull not let you watch while they fuck?

>>2896788
good eye. I glued both ends so that the string would not work its way off the ends after tying it off on the ends. The glue acts as a stopper for the string.

As far as bulls and semen, none of that was involved with anything and you said the f word which is a sin.

I was checking out androidworld.com recently. I wonder what made that guy give up. He had a grand vision but never followed through. I tried emailing him some months back and no response. Sad.

>>2896758
this is probably the worst built robot I've ever seen good job

>>2897181
What makes you think that? I consider it the best and am able to speak for an hour or more proving it point by point with ease. I guess the proof is in the pudding. You will have to learn the hard way when it outperforms in every metric imaginable by 5x any robot ever made in history

>>2897190
Lmao yeah "thrust me it'll be the best because I can yap one hour about it". Just look at the build quality man it's obvious that you have a mental problem. Go see a doctor really

>>2897190
even your base idea of not using gears because it's noisy is retarded. Haven't you heard the whining of the bldc motor ? The noise from the belt friction ? I guess you'll never make a joint move anyway so maybe you are right maybe it's silent after all

>>2897193
what is wrong with the build quality?

>>2897196
gears are very noisy don't you agree? I've seen enough robot projects with loud gears to know that loud gears are not a viable option for a indoor (largely) robot. It would be noise pollution and very disruptive and would make it NOT pass for human which is a non-starter. Has to PASS for human. You ask if I've heard the whining of a bldc motor. I have but it is very quiet comparatively is it not? MUCH quieter than brushed and MUCH quieter than gears by orders of magnitude. I imagine the little sound deadening from the skin and clothing would be enough to silence that noise would it not? However gears or brushed motors would cut right through anything sound deadening like a knife through butter.

>belt friction noise

don't have any belts.

>never make a joint move anyway

0% chance of that happening.

the build quality complaints I will need specifics and some solution to your perceived problem if it can be substantiated. If it is merely a aesthetics grip and nothing functionally wrong, don't worry, it won't be visible these are all internals so far under the skin. If it is functional issues, what are they and how can we fix it. I am genuinely not seeing any issues.

>>2897196
one more thing on bldc motor noise: did you know that alot of motor noise can be attributed to the control software? Look it up. With proper software algorithms, you can reduce noise almost entirely and that applies also to stepper motors. Has to do with gradually ramping up phases of the rotating magnetic fields instead of hard switching which makes electrical noise that is audible. Sine waves are used to quiet things down for this. Not sure if you were taking these factors into account on your post about bldc motor noise and I assume your refer to poorly actuated bldc motors with no quality algorithm controls to quiet them down maximally which indeed would be a bit on the louder side than it could be.

File: maxresdefault.jpg (60 KB, 1024x576)

60 KB JPG

>>2891524 (OP)
godspeed op, you are a glorious retard

>>2897190
> I guess the proof is in the puddin
Yet you constantly talk about how every robot designs out there currently is vastly inferior to you. Despite your own prototype still not being able to move under its own power yet.

>>2897346
they aren't designed to pass for human and that's why they are bad/inferior. Any design, even unproven ones, are better than a design that is not aimed at the correct goal.

>>2897223
just look at what you've got so far and tell me with a straight face that you believe you can make a "high quality" motor driver ? You can't make anything of high quality, your builds are 15yo third world country tier my dude open your eyes. And even if by some miracle your driver code would not suck, no I was just talking about the simple rotational noise of the motor it's bearing, the ropes, everything is noisy. Also the fact that you reply 3 times to the same fucking message at 6h intervals to find a good reply make me think that you truly have a mental retardation, I'm happy to rent free anon. Even if your design decisions were good (they are shit and you are delusional) the realisation is literal shit, no kidding I've been a judge to "young investors" competitions where even a 10yo would ratio you. Nobody here takes you seriously I don't understand how you can even post those pictures without feeling shame

>>2897355
what a retarded delusional take anything that moves a finger with 1deg accuracy is 5 years ahead of you

>>2897364
that's the best you've got? Just vague claims, zero specifics of anything problematic, and "well I doubt your driver software will be good code"? How lazy and unthoughtful you are sir. Do better.

>>2897365
why so? Prove it. Give substance to your hot take. I see nothing of substance here.

File: 1000039166.png (1.36 MB, 2400x1080)

1.36 MB PNG

>>2897378
why should I prove that anything that works is better than anything that doesn't ? are you that delusional ?
>>2897377
Are you serious can't you just look at your picture and figure out the differences ? Picrel is what real engineering looks like, try to look at it and guess why your project looks like it's made out of cum

>>2897380
I've specifically selected a pulley joint system for a robot elbow just to show that it's possible and already done so your "it'll be 5 times better than anything" is already false from the start

>>2897380
a working mech style bot is worthless if the only quality bot WORTH MAKING is human passing. So then a mech bot is not worth even considering and cannot be ahead of me. Who is making human passing bots that is ahead of me and realbotix doesn't count since they are using metal gears which is too loud.

>>2897380
so you judge the entire build based on the surface finish? That's not a valid criteria especially when everything I'm working on it going to be hidden inside the skin.

>>2897381
no this example is a mech bot so disqualitfied by default from comparison. Mech bots are trash by definition. Has to be human passing.

>>2897400
>>2897401
>surface finish unimportant
>only thing thats important is human-passing

Pick a lane.

Do you have any reason to believe people need their robots to be human passing? I love my dishwasher and car. A fake human actually seems really evil. Why you trying to fake this? Awful.

>>2897426
Furthermore, any time someone has done their very.best.and made a reaply.good human head, all the muscles and movements and nice passing skin and makeup... it's horrifying when it's turned on.

Humans know a natural law to distrust and fear these impersonators. With you religious background if you ever get this thing working and turn on the face and arm and see it's uncanny evil smile you're gonna see the demon propping it up and bash it in with a rock.

>>2897399
This type of "mech robot" is not different from your retard just put a skin on top of it. Just because you made the absolute retarded choice of using human bones structure doesn't mean yours is not mechanical.
>>2897400
Your problem isn't surface finish it's build quality, you don't even have the surface of your robot what the fuck are you talking about. Just be honest and stop avoiding the obvious for once.
>according to me mech robots (aka what I'm building) aren't worth building because I SAID SO so my robot that can't move and is made out of scrap and hot glue is better
USING CAPITALISATION DOESN'T MAKE YOUR STATEMENT TRUE. Honestly everyone else is right just go check a doctor

>>2897426
why is a fake human evil? I think it's cool and the only robot worth pursuing and if it doesn't pass for human then it is not even to be compared to my project and is irrelevant and trash.

If it passes for human it will be much more enjoyable to interact with, the social interactions will be fascinating, it will be fun to see how long it takes people to realize it's a robot. It can go into public without being identified as a robot which will grant it immunity from vandalism, theft, etc since people who attack robots would not attack a human passing robot - thinking it to be a real person with rights. This will give it safety and keep it off the radar. It can go about its business unmolested this way. Nosy people will not try to report it or become afraid of it if they genuinely think it's just some person doing their own business. This makes it far more useful for off site deployment into various tasks out in the general public. One store owner asked a streamer with a robot to leave the store. It alarmed customers. People found it threatening. If the robot passed for a human they would have been able to do the shopping and leave with no issues. The list of reasons for it to be human passing are ENDLESS and these are just a few off the top of my head. The fact you cannot see any of this shows you have not given this much thought AT ALL.

>>2897449
I actually have never experience the uncanney valley affect at all. People describe it but I cannot relate. I always just am in awe and amused and intrigued and deeply satisfied and thrilled every time I see a robot and all the moreso does this response in me scale the more real it looks. I get chills, lean forward in my chair, eyes wide with excitement, and smile huge on my face. Also demons have nothing to do with robots because they can only inhabit living things with blood (the life is in the blood) and so without blood a robot is demon proof.

Also, I think I can successfully make my robots very unthreatening and one way I am back to considering is by making them look weak, old, and frail. With a slight hunch back. A very innocent and warm smile and demeanor. And very religious, conservative Bible thumping like a very sweet church going grandpa or grandma. I have been lately considering going this route for my robots perhaps. Part of my reason for this was my disgust with the game second life. I noticed all the characters there looked "hot" and were dressed immodestly and I'm like what is with people pushing for looking hot all the time with their creations? Perhaps my robots should be the opposite of that garbage after all. I like to look at hot people like the next guy, but something seems wrong with obsessing over that. There is beauty in ugly and old looking too IMO in its own way.

>>2897527
>This type of "mech robot" is not different from your retard just put a skin on top of it. Just because you made the absolute retarded choice of using human bones structure doesn't mean yours is not mechanical.

no, no, no, if you put skin on a mech bot it would not look human at all - it isn't human SHAPED. When I say mech bot I am not meaning it has mechnaical parts. I am meaning it is NOT SHAPED to look human. A robot MUST have EXACTLY human shape that can PASS for human to be not a ugly pathetic mech bot that is disqualified from even being compared and is trash. Adding skin to Optimus would not make it pass for human because it isn't even close to human shape.

The comment about it looking like my parts are covered in glue is a surface finish comment. If you are referring to a deeper issue about build quality, what is that issue? Nobody has named one yet.

>>2897560
>doesn't feel the uncanny valley

Defective

>>2897567
You dont have t he creative impulse of.God. You're making this bad imitationm (evem if you.think it's good or the best you can do) and you won't be able to give it a soul or defend it in anyway. A demon is gonna come along and take over it and think it's hilarious to take over and tell you.lies. lies like this is a great idea, it's looking great, keep going. Make me a puppet that cannot think see or hear so I can take over it and use your ignorance to create more evil in the world.

Gg bro

>>2897857
>demon will possess your robot
I was under the impression that demons can only inhabit living beings be it animals or people. I do not think demons can live inside of robots.

>can't give robot a soul
I am adamantly opposed to anyone who thinks you can give robots souls so we agree on that. The suggestion one can do this is both blasphemous and evil and a lie.

>create more evil in the world
No, I don't believe my robot will do this. My robot will proclaim the truths of the Bible all the time and be an example of holy living for people to observe which might cause conviction. While not a real person, it can still display those positive attributes and set a good example.

>>2897857
actually it just hit me, since demons hate the Bible and hate people obeying the Bible, even if a demon could get into my robot, it would be tortured there since the robot is going to be programmed to regularly tell people to obey God and read the Bible and quoting the Bible and stuff. That would be like hell for a demon to hang out around. Maybe the demon would repent and start to obey God again and become a holy angel again.

>>289805
What do you base that on, a John Grisham novel?
Satan quotes verses at Jesus in Gethsemane

>>2898150
Sorry, what a Rookie mistake. Of course I meant the temptation in the wilderness

anyone figure out very small scale hand POWERED articulation? I've seen a lot of gab about articulation, I haven't seen any proof of concepts for articulating a very small model hand. Like roughly 2 in from wrist to middle finger tip.

Obviously you could 3 print something that articulates at that scale, but delivering any sort of power transfer to it seems watch maker-y, even though that's far beyond my use case.

>>2898274
geared n20 motor as winch and bowden cable to translate the force to the joint. This should be fine for your small scale

>>2898274
I really really recommend not taking any robotics advice from this psychopath

>>2898304
In this case, he has the right idea (kinda). I'd omit the bowden tube (too big for scale; you just have to build guides into your finger design), but a hobby servo hooked to a pair of control wires gets you an open/close motion.

>>2898306
no, a bowden cable is not too big for this. You can buy hair thin PE fishing line and TPFE tubing to accommodate the small scale. Get wrecked son.

>>2898317
>you can buy things
Just because you can doesn't mean you should, dipshit. Like I said, you just need to design retainers into the finger. I wouldn't use a bowden tube on a joint at any scale; the second it kinks it stops doing its job.

>>2898300
Not small enough unless I want to scale the anamatronic up and have the power delivery be in the torso.

File: metrocop hand.png (139 KB, 855x659)

139 KB PNG

>>2898425
to clarify, it's a 3 digit setup, opposable thumb + index + (middle ring pinky) with splay. (pic related) There's a guy on youtube who has a prosthetic that does 4 fingers mechanically, but he's a machining autist who used micro bike chain and milled parts. I don't need that kind of manufacturing complexity inherently. There's pager motors that I'm using to drive threaded rod joints, and something of that nature could work. Bowden cables won't because I'd have to have 7 of them at minimum, and anything that I could feasibly fit also would have to navigate the wrist joint.

>>2898437
>Bowden cables won't because I'd have to have 7 of them at minimum
You can get 1mm diameter tpfe tubing and .5mm diameter fishing line. How is 7 of those too big to fit?

>>2898437
>>2898533
I think the more important Q is what's the expected use case on something the size of a doll hand? I think movement is the best you can expect out of a hand at that scale, you're not going to have any real grip strength.

>>2897377
Bro, you haven't even made any PCBs. Like none. I ain't seeing any breadboards either. You've done nothing that shows you have the capability to make this fancy motor driver. Electrical noise can't just be fixed in software.

File: 1737348739854.webm (3.09 MB, 1080x1920)

3.09 MB WEBM

uh oh
too slow OP
a team of engineers with infinite gov resources will always beat you

File: Lilith_with_Lance.png (148 KB, 333x480)

148 KB PNG

>>2898554
Do yu want 2nd impact? Because this is how you get 2nd impact.

>>2898554
I saw >>>/wsg/5817970 and instantly remembered this thread.
For a moment, I thought a /diy/ anon actually got up from 3D programs on his PC and finished his thing.
Crisis averted, I don't have to reevaluate my worldview.

>>2898554
actually this started out as one youtuber in a basement and was always loud and I warned him about that. He did not listen. Now he has a large cool robot that is too loud to use indoors. Will never be viable. Loud as hell. He had to mute it and go to music for this video to be bearable. Also it's still a smaller team and non governmental

>>2898554
>>2898575
Why was I programmed to feel pain ?!

>>2891524 (OP)
Do you have a college degree or some sort of education in a trade? Will you teach the robot to hate others from different faiths and beliefs or treat them well? How will it talk to them and such?

>>2898656
>Do you have a college degree or some sort of education in a trade?

I'm a jack of all trades. I did 2.5 years of college then dropped out to save money. I had a 4.0 gpa when I left. So it wasn't flunking out. I just thought I didn't need a piece of paper if it meant debt - not that all degrees are a waste but the field I wanted at the time it was optional.

>Will you teach the robot to hate others from different faiths and beliefs or treat them well?

it will treat them well by yelling at them to obey God and read the Bible or else.

How will it talk to them and such?

I mean it can be nice to them unless they are really prideful then it might raise its voice and warn them to obey God or else God's gonna boom them and put them in a cage. It will tell them to stop being bad and be good again and listen to God again.

>>2898624
cope

>>2898698
they would have to literally start over to make it quiet. Has to be designed in from the very beginning.

>>2898552
yes I have made pcbs and my motor drivers are mostly copying other people's designs on the electronics and lots of chat gpt's help. it doesnt matter if I have to build it more than once to get it working, if I don't give up, I can make a pcb it's not that hard. there's almost zero innovation required its just connecting parts up following the schematics

>>2898701
motors can be made quieter
saying that there are no incremental advancements is dumb
energy density gets better
efficiency improves
headroom in noise, thermals, size grows

cope

>>2898552
also yes, electronic noise can mostly be fixed in software go look it up. has to do with turning the switch on and off at a certain rate so that it gradually fills the inductors in the motor. people did this to make 3d printers become quiet and do this with bldc motors too.

anyways, I tend to think that electronics are super easy and a joke they are so trivial. It's like ikea furniture building its so easy.

File: n20-gear-motors-location.jpg (353 KB, 904x648)

353 KB JPG

Ok so a few minor updates:

I have decided that since I am employing tension springs to actively work against the motors in a constant tug-of-war while the motors try to grasp, I'm losing grip strength based on that. To make up for that, I'm going to use a separate motor for the distal-most fingertip joint and the second to distal-most fingertip joint rather than have a single motor do both of these joints. I made these adjustments in my CAD. I will have to change the tubing setup for the grasping tubing of the index finger to reflect this change too. This will also give the fingers even more precision and dexterity in the end - not to mention a massive boost in strength - so it's well worth it.

I also decided to use n20 gear motors for the axial rotation of the base of the fingers instead of BLDC motors like everything else since these will only be used when doing the tiniest of micro adjustments and rarely employed - so a little gear noise once in a blue moon for this precision work on a tiny scale should not be that bad. So that's 4 N20 gearmotors going in. These are being used just to save on space taken and pulleys needed a bit. I'm putting these 4 into the forearm in location pictured.

Next, when the spring is pulling, I noticed the TPFE guidance tubing goes from straight and relaxed to wavy under the tension. It is trying to compress under the friction which is what causes this. To prevent some of this compaction stuff on the spring's tubing, I'm going to be using TWO tubes which will divide up these forces causing this by 2. Sharing the load between them evenly. So the tension spring will have two fishing lines coming off of it and two tubes to guide that line to the finger joint where it does it's thing.

>>2898702
>lots of chat gpt's help

>trusting chatGPT for electronics
Does your tardwrangler know you're posting here?

>>2898702
correction on this: I have made circuits and used unpopulated pcbs to make boards for motor controlling brushed motors but I haven't created a copper etched board from scratch yet so technically you were right to say I have not made pcbs and I misunderstood perhaps. For example I used proto board and made a full microcontroller board but did not copper etch to make it but instead used wire for point to point connections on all components with soldering. Not the same as making a from scratch pcb. I have also designed and submitted a pcb design to JLCPCB but canceled that order due to overcharging issues. So I haven't quite technically made a PCB but have done a fair bit of random electronics projects by this point and have had success.

The cook thing with this thread is that since op has very little progress, almost 0 every month he can keep on dreaming about how good it'll be. The closer to completion the more obvious it'll be that it's a bunch of garbage just like the clone robot, so op is smart to let this go as slow as possible to fuel it's desperate need for attention

>>2898951
I think my progress is decent given the other stuff on my plate. I'm trying to chip away consistently. Anyways, seeking feedback and suggestions and discussion is not unhealthy as you suppose. Anyone can claim someone has a desperate need for attention but good luck proving that. I would rather progress fast. Your conspiracy theory is wrong sir.

>>2898991
>I would rather progress fast
you have made zero progress

>>2899001
Literally nobody can claim that. Prove it.

>>2899170
>Literally nobody can claim that.
I'm claiming it right now.
>Prove it.
It's very difficult to prove the absence of something. It should be easy to prove the presence of something, but you never do...

>>2899645
literally posted photo after photo of progress but nope that's not progress to you. Literally would stand as proof in the court of law but not to you. Nobody can reason with you.

File: opposable-thumb-muscles-detail.jpg (665 KB, 808x642)

665 KB JPG

Here's the CAD highlighting the 4 muscle points planned for the opposable thumb's base. So 4 motors for the thumb base joint alone! Should make it very capable as far as strength and full dexterity all angles all directions a thumb can do.

Why do people make such a big deal about control system algos anyway? Just input the system state-space equations into Matlab and then apply LQR param optimization lmao. I feel like what OP is making by creating artificial skin and bones is actually MUCH much more difficult than running some code on Matlab for 20 mins (that you can easily learn to do in a 5 min tutorial made by a South American university professor).

Somebody stop this mad scientist coomer.

>>2897190
Make a webm showcasing how good it is. Wear a mask and add subtitles.

>>2899696
Good point. Probably non programmer frankly. They think code is magic basically.

>>2899799
There are lots of coomer roboticists but I'm not one. Hence why none of my robots are getting realistic genitalia or any gentialia for that matter. I see great coomer potential in humanoids and am turning away from all of that since it's sinful.

>>2899800
I will once it's ready but it doesn't move yet. Also I recently met a guy planning a way better robot than mine so I'm not the best anymore his plans are way more impressive.

>>2898951
I was thinking there's a tiny bit of truth in this as follows: like sometimes as I approach a significant strategy going into a prototype for a mission critical apparatus, I feel like sleeping on it for some days to think of what could go wrong and some plan b and plan c so that if my idea fails, I don't get discouraged and already have a readily available backup plan in place so I don't feel stuck and helpless or discouraged. This has saved my butt many times. The fear of the unknown is big in this project and there is definitely apprehension when I'm not sure how some idea will play out fully. I have found that many times now what I envisioned would work well had some hiccup issue I did not forsee. That is hard to deal with and surprising every time but I'm getting more used to it. Normally in life my solutions always work but with this robot many of my solutions have failed. There is always some dumb thing I did not consider fully or w/e. Working with such hard and complicated stuff here.

>Total time since OP: 33 days
>Progress in this thread:
- A picture of some small parts on a plastic sheet
- "CAD projects" that amount to putting a bunch of cylinders in a model of a skeleton downloaded online.
- A simple pulley made of string
- Added a spring to the pulley
- A whole bunch of dreaming of where things will hopefully eventually end up at in the future

So a month for a string pulley and a whole bunch of dreams. Godspeed OP, I can see why >>2898951 stung.

>>2899984
you are greatly downplaying the complexity and difficulty of this stuff. Like for example I had to do like 50+ sutures and just doing like 6-7 of them is like a whole small surgery. This stuff is challenging.

But also like I said I have alot of other stuff on my plate. Imagine if I'm 007 and flying around the world on private jets doing missions for a secret service of the government and only work on the robot when I'm on leave and not catching up on emails and family time etc. That's alot.

>>2899984
I mean like maybe 12 hours I did this month. That's not too bad. For a hobby project. Like if you had a old hot rod in your garage, if you put in 12 hours for the month you'd be like yeah, not too bad. Not a great month but decent especially if very busy with some other big deadline project you can't put off.

>>2900163
Yeah but people working on cars have a much simpler task ahead of them than your pie in the sky stuff.
>>2899921
>I will once it's ready but it doesn't move yet. Also I recently met a guy planning a way better robot than mine so I'm not the best anymore his plans are way more impressive.
You'll take a little time to think about why yours is actually sooo much better, then start talking shit about his project. You're a narcissist so once you're back on form what you've said here will no longer be a tenable position for you.
Link to his project?

>>2900636
Doesn't matter if car building is simpler the principle that a 12 hour commitment per month to a active hobby is not horrible still applies to my month here. It's like a C grade. I can have A+ months to make up for it later.

He hasn't posted about his project publicly so I can't link it. He just told me by email about his plans. They far exceed mine in complexity and scope.

Some highlights: his will have skin with countless sensors that can feel touch and will be self healing skin. His will have artificial muscle fibers that are self healing and electrically actuated more like a real human muscle fiber. He mentioned a bunch of stuff about graphene nano tech involved in some of this stuff as well which is a bit over my head. He's raising millions in funding now so not yet in active development mode on the hardware but has done alot of AI work in python he said. He plans to put together an elite team to help I guess and big manufacturing facilities. I'm not doing those things either but they give him alot more advantages in some ways. But comes with pros and cons I think.

Disaster has struck:
In testing recently, I had some VERY bad news: I don't think the spring extension idea is going to work. The amount of force required to unravel the Archimedes pulley system when working against all the friction in that system, the friction in the winch in place pulley, all the friction in the teflon tubing runs, and the magnetic friction of the motor itself while working against the downgearing (since when working in reverse direction it acts as up-gearing) is all working against the spring and I think it's too much to ask of that spring. I can't even really pull by hand - pulling pretty hard like 3-4lb of force it wasn't budging. So this is tragic for my whole approach so far and we have to go back to the drawing board. A proposed massive overhaul solution in next post.

Note: The name of the resistance to turning a BLDC motor has while freewheeling (no electric applied to it presently) is called cogging torque, which is caused by the interaction between the permanent magnets and the stator's iron core. This force may seem insignificant but due to my downgearing system, the spring has to deal with it after it has been multiplied 44 times due to the downgearing the spring would be fighting through from reverse direction at the bottom of the pulleys and traveling through what then acts as upgearing when going in reverse direction from spring's end.

I think I've solved it! So first, I want real force working on the extension aspect, not some wimpy spring. I already said there's a lot of frictions that extension system has to bust through to work. And I'd hate to have a very strong spring anyways since when grasping, the motor would then be fighting against a strong spring for extension which is a huge inefficiency that works to weaken the grasping action significantly at that point which is bad design frankly. So we want IN DEMAND opposition for the extension rather than a constant opposition of a spring fighting against the grasp attempt of the motor. We also want the motor that does the grasping to actively rotate in reverse direction rather than freewheeling in order to not have to fight it's static friction caused by its magnets which is significant. This means we either have to go with a two motor system - one for grasp direction of the joint and one for extension direction of the same joint (HORRIBLE WORST CASE SCENARIO BUT POSSIBLE IN A PINCH) or we need to go BACK and refute the notion that the motor is unable to operate two separate pulley systems for extension and grasping functions coming from a single motor attached to two pulley downgearing systems. Which would entail the motor turning clockwise to create grasping and counter clockwise to create extension. The problem with such a proposed system is that in theory it was said to be impossible due to the inevitable derailment issues and tension issues that this would invite. I am proposing we tackle those issues it invites head on rather than avoiding them entirely like we were trying to do for quite a while now. It is a VERY tall order to get that to work but that would be the best possible scenario IMO.

It is great if we can get it to work since we tap into the full power of a single motor to do both flexion and extension and we then kill two birds with one stone. All the friction issues with the tubing and pulleys is solved by the motor when it reverses directions and actuates the opposing pulley system. We just have to have slack in the line due to the different diameter mismatches of the two different winding directions we face and also have to have that slack pulled taught by some mechanism to prevent slop that causes derailments. I really want to press for that HARD now. But to do that I really have to scrap the winch in place pulley idea basically I think. Well not necessarily - even that I think can be worked out but is higher risk and harder than my current favorite new, novel solution. So we can reattempt winch in place stuff perhaps in the future but I want to set it aside for now. My newest idea is for that first large run-out downgear to be 2:1 and use regular Archimedes pulley system approach but to put that pulley into the torso and have a weight hang off the bottom of it or have a VERY tiny motor attach to the bottom of it that is to place tension onto it regularly to remove all slop. This can be a motor the size of my pinky fingernail perhaps (not sure though). OR a weight. I lean toward using a weight now since that would be easiest I think to pull off. I got the weight idea from studying the cable machine for triceps at the gym the other day. I can have the same type of weights or something similar to those used by gyms. But doesn't have to be adjustable like those but same concept.

Granted one downside to this approach is what if the robot is laying down or upside down wouldn't it not have weight able to pull down by gravity then? So to solve this I can have 3 weights perhaps, one for each possible direction: upright robot, upside down robot, laying down robot... actually 2 weights should be fine: laying down and upright. Hmm... well if he's laying on back or stomach the weight would have to pendulum or slide past a central point to the other side of robot on a track. Yeah that should work! So 2 weights I think can do it. If upside down he's screwed we'll say. He won't use fingers in any direction change way until he flips back around upright or sideways if doing a cartwheel or handstand for a bit. That is a fine tradeoff. Right now I'm thinking a straw with lead tube in it as the weight or something like that. Even considering just using a fishing sinker perhaps at the moment. Have to think on this more...

File: dual-direction-pulley-sys(...).jpg (2.48 MB, 2952x3054)

2.48 MB JPG

Here's the official design drawing of this proposed single motor actuating both forward and reverse directions with two separate Archimedes pulley systems opposing one another. You'll also note that the left hand side of the drawing has a pair of Archimedes 2:1 pulley downgear systems, one for forward and one for reverse directions of motor and these two are going to be very long (16 inches long) and therefore are located in torso. The remaining 16:1 Archimedes pulley downgearing systems will be kept in the forearms near to the finger joints they are actuating as we had planned originally and already have in place.

You'll also note the weight that hangs off the bottom of both of the 16" long 2:1 pulley downgear systems that can keep them both taught at all times despite their varying lengths that will always be changing. The weight is able to slide since it has a fishing hook eye above it and on both attachment points to the 2:1 pulley downgear systems so it is always adjusting these 3 fishing hook eyes to always keep tension on both systems freely.

File: eye-of-fishing-hook-simpl(...).jpg (825 KB, 1548x1668)

825 KB JPG

Also, I recently stumbled upon a VERY much simplified version of my miniature pulleys. So up to now I've been using 1x3x1mm ball bearings to make tiny pulleys and been variously perfecting this approach but it is still not THAT small and is a bit complex to make and we have to make literally THOUSANDS of these to do the whole robot. That presents a bit of an issue due to the large work that requires. At least until mass manufacture of them comes in one day perhaps. But while DIYing that, it's alot to deal with making SO MANY somewhat challenging to make things. That said, my proposed EVEN MORE miniature and WAY WAY WAY simplified to make pulley is to just use a single fishing hook eye. Literally, that's it. I can use a tiny fishing hook eye and use that as my very first pulley for the 2:1 16" long Archimedes downgearing systems in the torso. This will cut down on size taken dramatically and complexity of its build. It will make the pulley basically failure proof too. The way it will EVENTUALLY fail is by the rope rubbing it enough to cut it in half. But I think the rope would fail before the pulley would fail and so that doesn't matter then. You'd replace them both at once on routine maintenance. No need then to worry about that eventuality. And the ridiculous ease of manufacture of such a simple pulley makes replacing it trivial. I also think that using this just in low load, high speed, low force early pulley downgearing stages is a non-issue since the friction with such a low load on the first downgear or two will be so trivial that the string itself would fail WAY before it would slice through the metal (acting like a saw over time). I think it would take literally MANY years due to the super low friction at these low forces. Now I'll still use the ball bearing style for later stages of downgearing where the loads go way up, but for the first stage or two I think this will work just fine.

>>2900658
Right... just stop foaming at the mouth and think for 5 seconds:
- you do not have infinite space to work with. All that torso space has already been taken up by your own proposed systems (the "heart" and "lungs") or by batteries. No room for hundreds of counterweights.
- even if you did have the room, I (think) you've got a drop distance problem. Might not be able to articulate what I'm visualizing, but I think the distance the counterweight drops has to be equal to the circumference of the bend being made. So for a bicep/tricep group (elbow bend) you'd need a drop distance equal to the circumference of your wrist's path of travel (meter and a half? Something like that).
- even if you have solves for both the above? That becomes a heavy, heavy robot.

>>2900681
thanks for the thoughtful response! I can see you actually put thought into what I said. I think that's amazing of you to take the time to do that. I agree the weights will have to be quite small - just enough to pull out slack from the string. Not enough to actually fight the articulation. My drawing illustrates that aspect poorly I see now. The string would be tied off on the bottom with a separate rope and the weight would only pull out slack but not actually hold the bottom loop down like the drawing wrongfully infers. Was trying to just convey key concepts so the drawing is simplified leaving out some key details I see. The weight can even be replaced by a litlte piece of rubberband or elastic necklace string I think. Just something to pull out slack is all. Also I don't think the drop distance needs to be much at all. But I could be wrong you might be right on that part. I'm not visualizing that part well I admit. I'm a bit clueless on that. I agree the torso is largely filled already as you pointed out. I'm impressed with how much you are aware of my design to even know what space I have left in the torso like you do! Someone is paying attention I see! Very cool! Yeah I will have to use fine string and fine fishing eyes to make each one of these 16" long 2:1 downgears no larger than a muscle fiber and this way I can cluster hundreds of them together just like a muscle clusters hundreds of muscle fibers together. This will make it extremely space efficient in theory but I could be completely wrong about this. Not sure how close they can be together without conflicting with or snagging with their neighboring pulley. So I might have issues like you said with space but I'm hoping it can squeeze in there. So far I'm thinking putting it just to the side of the spinous process alone the vertebrae or alongside the main pc descending down inside the ribcage. Those are two good spots unclaimed so far.

>>2900651
>Doesn't matter if car building is simpler the principle that a 12 hour commitment per month to a active hobby is not horrible still applies to my month here.
That would be true if we were grading you for effort, but you will be judged, as all scientists and engineers are, by your output.
>He hasn't posted about his project publicly so I can't link it. He just told me by email about his plans. They far exceed mine in complexity and scope.
I mean, he can claim he's planning whatever he likes, but it doesn't mean it's actually something he can do. It's just like your project in that regard.
I'm planning on building a robot that silently moves around particles of silicon doped with nano scale iron moved by a strong magnetic field, powered by a silent micro-cold fusion reactor. The projected force coupled with the silicon texture will feel like human skin and be moldable to the planck scale. Sensing will be done by the flucuations in the magnetic field and their back-impact on the fusion process. The device will look like a small box in which the reactor and computational parts will be contained, and it be will be able to summon its physical body from ambiently present particles in a working environment like I have described, and, later on, from particles like this created ad hoc from particles in the environment through remotely ablating environmental iron magnetically and pushing it in to silicon or other soft materials. Only a small amount of these particles will be required as the force of the field will be so strong. The processing will be done by a non-traditional AI architecture of my own invention that will completely revolutionise the AI world (no you can't see it, no I haven't finished it yet, but it will be ready in 10 months after I set my mind to it.) I am very elite, and I have already raised billions in funding.

>>2900761
if that were all true you'd be the best. I'm not holding a double standard then. Whoever has the most elite plan and actually INTENDS to follow through on it genuinely and has researched it and really laid out detailed steps and designs at least is now making a strong case to be the most elite. It's not based on actual achievement then but on projected outcomes that are at least somewhat plausible and concrete.

>>2900681
Upon further reflection, even though I think the torso can house some of what I said, I am now leaning toward trying to cram it all into the forearms after all. The more I free up the torso the more room I will have left over to play with for the rest of the robot. I don't want the hands to be so space greedy and hogging that the whole of arms and torso is getting overwhelmed with hand parts and lacks room for everything else.

>>2900859
It is all true. I've researched this for 14 years and I will be carrying it out. You'll feel pretty silly when my design takes the world by storm.
Step 1. miniaturise reactor. This should be simple and require no further explanation.
Step 2. experiment in my garage with mixing mixing iron filings and silicon from online.
Step 3. finish converting AI pseudocode to real code.
Step 4. have AI complete the rest of the work.
See? It's just that simple. I am the elite.

>>2900875
your tone was not a serious one before. So the switch to pretending it is serious now doesn't work. You'd have to come out the gate with serious tone. That's why you cannot pull the serious card anymore now. Too late for you.

>>2900875
You're joking but the namefag's design would be coming along much faster if he trusted more of it to AI instead of taking his ideas and piecemeal bouncing them off chatGPT. The most productive design workflow I have for my mechatronics project has multiple AIs iterating my initial ideas between each other, and the end results usually only require a little bit more prompting or fine tuning on my end before making something workable. Prompting from concept to EDA files is also possible these days and would save OP from having to design and solder his motor controllers by hand if he bothered to learn AI above a first semester CS undergrad level.

Tell me about the system for muscular/skeletal movement. I was thinking about growing silicon muscles one day, that could be manipulated by electrical inputs.

>>2901163
You're telling me you would believe my "design" was achievable so long as the person who told you about it was delusional enough to believe in it? There is zero assessment from a technical feasibility point of view coming from you? I chose that "design" to be as ridiculous as I could dream up.
>>2901171
Yes, AI is a great help in research, though it cannot do the impossible of course and is limited when it comes to covering unexplored territory instead of packing and digesting existing knowledge to help get us up to speed or work within the bounds of known concepts. That is a use that OP would greatly benefit from.
Step 4 is in reference to OP's claim that it doesn't matter that he doesn't understand how this will work in detail, because his vaporware AI will learn for itself, and his robots will construct themselves. It's nonsense obviously, but it is amusing just how much nonsense it is.
His behaviour makes sense if you understand he is a narcissist before anything else, including a roboticist.

>>2901343
>I chose that "design" to be as ridiculous as I could dream up.
I just would not want to judge by feasibility in my opinion if the areas are outside my knowledge base. I'd judge based on epicness and ambition overall and serious commitment and faith level of the person too. Although I'd prefer if they are willing to fall back to something more "realistic" if the wilder stuff didn't pan out in a reasonable timeframe?

>>2901267
that's literally this whole thread and the past threads linked in the OP

>>2901171
you can't just tell AI to tell you what to do. you have to decide what to do and then ask AI to fill in any knowledge gaps along the way. I still want the end goal regardless

>>2901343
>You're telling me you would believe my "design" was achievable so long as

actually, even if I don't think it's achievable, it can still be impressive that you believed you can achieve it despite me thinking its not achievable and I'd prefer to let you at least try to prove me wrong and I'd highly esteem you for the balls it would show to take on something that hard if you were very serious and committed and showed progress in earnest. But achieveabiilty is on a spectrum I guess. And alot of factors can help like if you got funded or put together a team and knew of some secret tech already out there or w/e that might help. But just a truly epic hard project deserves a lot of credit IF there is effort showing you truly are going for it as well to back it up.

>>2901343
>it doesn't matter that he doesn't understand how this will work in detail, because his vaporware AI will learn for itself, and his robots will construct themselves
this is not really my claim. I claimed that SOME details of subprojects within the overarching project do not need to be fretted over or figured out necessarily immediately and we can cross those bridges as we approach that or as time allows. The most important details to have down solidly are the ones relevant to the subproject or step we are currently working on. I also have no intention of waving a hand and expecting the robot to just figure everything out magically. There will be alot of teaching it and training it involved so I'll be intimately involved in the transition to more and more autonomy and eventually my involvement will fade more and more as the robot grows more capable and starts to not need me anymore

File: images - 2025-03-06T190833.735.jpg (5 KB, 251x201)

5 KB JPG

i tried so hard amd got so far. but it in the end it doesnt even matter.

just had another kid so been busy but hope to do a finger update soonish