My Advanced Realistic Humanoid Robots Project

[artbyrobot]

As to all the silly sayings like "just give the motor to pulley a go" - I already addressed this.  There is no reason to go down the rabbit hole of iterative prototyping and troubleshooting my novel microcontroller circuitry and motor controller circuitry to be able to drive the bldc motor as well as all the software dev behind that when I still haven't gotten the pulleys finalized and you seem to conveniently ignore the fact that the pulleys have undergone many revisions and redesigns and are not yet done.  So your idea to just skip all of that or snap my fingers and all the pulleys are just randomly done seems very illogical.

Disregarding the many previous actuator plans I have tested and eliminated as possibilities, the gist of our current progression approach is: I first mounted the motor, I rigged the finger to be moved by that motor, and next I'm to finalize 32:1 downgearing by way of pulleys in order to bridge the transmission of power from the motor to the finger joint.  Only when that is done will I iteratively finalize my microcontroller circuitry and motor controller circuitry - in that order and install those.  Along with that I have to design my battery supply circuitry.  Then and only then will I be able to develop the software to run the microcontroller/motor controller setup and begin testing the movements.  Not sure what the whole "just make the finger move" comments point even is.  Observe the order of steps and time that each takes in order to present logical suggestions that are relevant.  Not hard to do.

[Leiothrix]

I didn't twist anything.

You have quite clearly demonstrated that you don't have the base level of knowledge to decide what is worthwhile and what is not.

You have just deemed everything that you don't already know to be stupid and worthless.

And again with not answering questions.

Edit:  And huh?  So are the epic pulleys epically finished or not?  If they are not finished then you can't declare them as working because they are not finished.  If they are finished you have to actually test them to figure if they work or not. 

Or are you just backtracking because you can't actually produce something?

[artbyrobot]

You have quite clearly demonstrated that you don't have the base level of knowledge to decide what is worthwhile and what is not.

that is your own misinformed opinion and incorrect.

And huh?  So are the epic pulleys epically finished or not?

of course they are NOT finished.  When did I say they were?  I did not.  In fact I clearly stated they weren't repeatedly.  When they are, updates will go out indicating they are.  The latest iteration of the winch turn in place single pulley mounted on the motor is done.  The archimedes pulley system I already announced was about to get a total rebuild and redesign.  You seem to not even be following along this project as this question proves.  You are just here to talk stuff but not even have a clue it seems.

[Leiothrix]

that is your own misinformed opinion and incorrect.

Well it seems to be the popular opinion here, if it is wrong you don't seem to be doing anything to correct it.  If the whole world was against you one would think that there would be a point where one would do a little bit of introspection, but nope not here.  Meh, it takes all types.

You seem to not even be following along this project as this question proves.

To be fair it is pretty hard to follow as your story changes every other post and instead of answering questions to clarify your position you change the subject and go off in a different direction -- all while insulting everyone and everything.  But the overall summary is "no progress" so it is kinda easy at the same time.

Observe the order of steps and time that each takes in order to present logical suggestions that are relevant.  Not hard to do.

You have your heart set on a BLDC but you don't even know if that is appropriate or not because you haven't even tried driving one. 

The order of steps is to design the power system either before or in conjunction with the transmission, not plonk it in as an afterthought.  You have no clue of the size of motor you need, the physical mounting arrangements, the torque or speed you need, the positional accuracy, electrical power or end reduction that you need.  Or slop, backlash, holding power, heat dissipation, duty cycle, shock absorption, alignment, serviceability, connectivity, weight, bulk or any of the other "boring unnecessary distracting completely worthless" stuff that you couldn't be bothered to think about.

You completely reject the notion of doing any actual engineering or even just listening to a bunch of people who have actually done some of this stuff before.

You are incredible, just not in the way that you think 

[artbyrobot]

This statement:  "You have your heart set on a BLDC but you don't even know if that is appropriate or not because you haven't even tried driving one. "  has got to be one of the most ill thought out ones I've seen on either thread.  Congrats   

News flash:  when you drive a bldc motor, it spins axially.  There is nothing novel or exciting or unexpected about this.  There is no need to "get it to spin" for me to know what I'm dealing with even down to the speeds possible, the acceleration possible, the torque that it will output, etc.  There is no mystery there.  So to suggest I won't know if bldc is viable until I see it spin is absurd.


That said, I was thinking about it and I can conclude I think I've put in 4-7 months of dedicated fulltime development on this project in that 10yr timeframe.  That is NOT including research on all the related fields and note-taking to get me to elite levels of knowledge required to pull this off.  That 4 months is assuming full time working on it 10 hours a day 6 days a week with total focus and no distractions during that and adequate sleep and no other major responsibilities competing for my energy and I have great health the whole time.  In other words, I think I'd be able to easily reproduce everything I did progress-wise in all aspects in that timeframe given these resources I just named.  That is a rough estimate best guess though.  But then again, so much of the design work and stuff was necessarily spread out since you can't rush genius or innovation and invention.  They come over time.  As inspiration strikes and solutions just come to you over time.  So really even making this estimate is very subject to revision it is just a ballpark I came up with lacking sufficient time to carefully consider every possible thing.

[artbyrobot]

If the whole world was against you one would think that there would be a point where one would do a little bit of introspection, but nope not here

This is actually factually incorrect.  I have a significant 50-60+ people who have come out and said that they thought my robot project and skills and knowledge level in the relevant fields etc were all elite, world class, and bound for success in their view.  They staunchly disagreed with the haters.  So it is NOT at all the "whole world".  Also, of the countless lurkers who have read these threads here, I am confident many are silently rooting for me but don't want the drama of defending my project or myself, seeing it would be a waste of time to publicly attempt to fight the zombie hoard of mindless jeering going on.  Over and over I've proven such gross errors in logic and failure to even pay attention to macro level details of my project progression and direction yet no humbling happens.  Just the same smug attitude without apology for the constant barrage of ignorance.  So why would the fans come out of the woodworks to face such rabble?  What would come of it?  You see, on forums and social media people just enjoy to be haters.  They love the anonymity and negativity.  Its a well known phenomenon.  A way to lash out because of one's own unexpressed unhappiness or w/e.  They love to just bully.  It's sad really.  Most people don't want to be on the receiving end of bullying so they get quiet in these situations.  That's why it seems very one sided on these threads where the bullying is in full force.  I don't mind it.  I use it as motivation fuel.  And I have thick skin.  And unwaivering confidence.  So the optics here can be very misleading.  And all of this is not to mention the "herd mentality" and "social pressure to comply" aspects that go on.  The mob mind.  People tend to go with the flow.  I'm against that.  People should be rational and think independently and stand for what they really think without just going with the mass opinion like a drone lemming.

[ebastler]

News flash:  when you drive a bldc motor, it spins axially.  There is nothing novel or exciting or unexpected about this.  There is no need to "get it to spin" for me to know what I'm dealing with even down to the speeds possible, the acceleration possible, the torque that it will output, etc.  There is no mystery there.  So to suggest I won't know if bldc is viable until I see it spin is absurd.

How much thought have you given to controlling rotation speed, acceleration, and torque?

• Over what range do you need to vary each of these parameters in your various robot "muscles"?
• Do you need to vary them independently of each other? Over which ranges?
• What parameter ranges can your motors of choice provide?
• Do you have plans how to achieve that via electronic control?

[artbyrobot]

But the overall summary is "no progress"

ss

Here's a patently absurd comment which really highlights the lack of arguing in good faith, sincerity, or rationality going on here.  So I have objectively a good 200 hours of CAD work done and on display for all to see.  That is not "no progress".  I have schematics I've made for microcontrollers, motor controllers, power supply, battery packs, etc that took 80-100 hours not including all the research that went into it to be able to design it all.  That is not "no progress..  Then we have the many tried and dismissed actuator solutions that are documented.  That is not no progress.  Then we have 250 or so youtube videos on the project covering hundreds of hours of detailed content and progress.  That is not no progress.  We have 3 robot frames with each 30-200 hours of dev work done on them.  We also have prototype power supply 80% done, microcontroller programming and dev work done, many iterations of pulley system designs done and ever improvingly so, all the parts purchased and selected for almost every single part of the whole build and all the vast research that went into finding out the best path for each of them.  We have a motor controller prototype 80% done.  The list goes on and on.  But someone being comletely illogical can point to this 800 or so hours of work and call it "no progress".   

[Leiothrix]

News flash:  when you drive a bldc motor, it spins axially. 

Yeah, and so do brushed DC motors, stepper motors, servo motors or any other style of motor.  You additionally could use linear actuators to provide rotation or all sorts of other things.

However you don't know the physical size, power, torque, speed, accuracy, acceleration, stall current, holding power, shaft size, slop, backlash or any other engineering requirements because you don't seem to think that any of it is important.  Even if you bothered to read a datasheet that is not a substitute for actually seeing it in person.

So you don't do the engineering effort but make up for it by not just giving it a go and seeing what happens 


With regards to progress -- you don't have an actual schematic (unreadable paint diagrams don't count), a usable circuit, a light that blinks, a motor that spins, a linkage that works, a finger that moves, or software that does anything.

Just because you spent a lot of time on something doesn't mean that it got you anywhere.

[artbyrobot]

How much thought have you given to controlling rotation speed, acceleration, and torque?

Speed and acceleration go hand in hand.  I have a deep and thorough understanding of how bldc motors modify these two via switching on and off the mosfets leading to the 3 leads that cause the rotation of the rotating magnetic fields.  I can control this via code.  I have fine control of all aspects of this.  Even torque can be controlled by not just switching but pwming the switching in such a way as to throttle the formation of the magnetic field in the inductor coils so that the magnetic forces are able to grow and fall with precision.  I understand it all.  It is a piece of cake to me.

[artbyrobot]

you don't know the physical size, power, torque, speed, accuracy, acceleration, stall current, holding power, shaft size, slop, backlash or any other engineering requirements because you don't seem to think that any of it is important. 

How absurd are you going to be?  I mean really?  I literally created 3d models to scale of every motor assigned for each and every relevant muscle in the entire human body in my CAD and had to carefully measure out the space required to get them to fit into the human form factor.  The motor can and shaft had to be measured and modeled to the millimeter.  So where are you getting these wild statements?  I have also LITERALLY done all the math for calculating the speed, the torque, and the downgearing for the finger motors and submitted it to this forum on my threads and they were questioned some only to be ultimately approved as rock solid math without flaw whatsoever.  The main person questioning it had to admit he made some errors in his math as he tried to follow my math and had to admit all my math was spot on. 

Also slop and backlash are moreso gearbox terms and I'm not sure how strictly they apply to a pulley based system which is always under tension and is intended to be back driveable and compliant.

As far as accuracy, I know much about this.  I know that my motor control plan enables me to stop the motor turning at least within 20% give or take increments of a single motor turn and I know that motor is turning enough to reel in 32 inches of thread like a winch.  That is a TON of turns and we have sub turn level accuracy to stop the motor on a dime.  That is extreme accuracy capability.  The limiting factors on accuracy will be the joints themselves and feedback speed I think.  The motors and motor controllers will not be the bottleneck on accuracy capabilities.  Also accuracy in a industrial robot arm that is gcode ran or w/e is far more relevant than accuracy in a bio-inspired humanoid type design where eye hand coordination and regular feedback visually and in terms of sensors and ability to make micro adjustments and improve position dynamically are going on there.  It is apples and oranges and you don't need the same kind of accuracy in the latter case.  It is a different world at that point.

[artbyrobot]

Even if you bothered to read a datasheet that is not a substitute for actually seeing it in person.

Who said I will not ever test?  How absurd.  But I do have all the specs already and don't have to test to know them up front and select accordingly.

So you don't do the engineering effort but make up for it by not just giving it a go and seeing what happens

I have done massive engineering effort so wrong again there.  Also I do intend to test the robot as I get to those stages.  I have been testing my pulleys as I go have I not?  And I already told you the order of operations as it pertains to the motor, pulleys, and electronics portions of this actuation system.  Testing the electronics is not possible when they are not yet setup and are not next up even.

you don't have an actual schematic (unreadable paint diagrams don't count)

I literally 3d modeled my microcontroller and motor controller schematics in addition to the photoshop diagrams.  Can't get any more thorough than that.  And if you are holding your breath for a traditional (read ugly and unreadable) schematic like you probably use, don't bother.  Will not EVER happen.  And yes, schematics do count regardless of the format preferences.  They are still schematics and news flash:  mine are readable and many here read them and commented on their concerns with them - all of which I addressed.

a usable circuit, a light that blinks, a motor that spins, a linkage that works, a finger that moves, or software that does anything.

I made all of the above multiple times in some cases.  I did a full software dev for brushed motors (to completion) as well as all the controller work done and tested it fully.  Forward and back.  I also did blink of bulbs.  I also moved the finger by hand.   Also developed the software for the AI extensively from custom 3d engine from scratch to custom web browser to computer vision to speech to text to morals system to dynamic custom coding language that the robot and I will code in and it will be runtime read and executed in real time etc.  And much more.  Even began a custom operating system.  Most of this stuff was only a firm foundation laid and not done to completion yet though.

So really you are just spouting factually incorrect statements over and over to your own shame IMO.

[Analog Kid]

Obviously this guy doesn't know the First Rule of Holes.

[artbyrobot]

Actually, digging holes in my view is happening to the haters/jeering critics.  The more they say in opposition, the more deep their own hole gets and the more embarrassing it will be for them when I pull of the "impossible" with this project.  I'll grant that the more I say in favor can be more embarrassing in my case if I fail.  So both sides are digging but unlike the jeering critic/hater, I have the wheel and they are powerless over the outcome.  I have the ability to make it happen whereas they can only stand by helplessly as I move powerfully close to each major landmark milestone with this project.  So my gamble is safer since I have control of outcomes and you only have speculation.

[ebastler]

Even torque can be controlled by not just switching but pwming the switching in such a way as to throttle the formation of the magnetic field in the inductor coils so that the magnetic forces are able to grow and fall with precision.  I understand it all.  It is a piece of cake to me.

Oh, really? Better remind ChatGPT that PWM is out of the question:

I will NOT be using PWM functionality or channels at all. [...]
Since you’re not using PWM, just basic digital output, you should be able to configure all 70 pins as digital output for driving MOSFETs.

[artbyrobot]

Even torque can be controlled by not just switching but pwming the switching in such a way as to throttle the formation of the magnetic field in the inductor coils so that the magnetic forces are able to grow and fall with precision.  I understand it all.  It is a piece of cake to me.

Oh, really? Better remind ChatGPT that PWM is out of the question:

I will NOT be using PWM functionality or channels at all. [...]
Since you’re not using PWM, just basic digital output, you should be able to configure all 70 pins as digital output for driving MOSFETs.

Nice try but fail on your part again.  Although the switching of the mosfets will be happening in a pulse width modulated way, hence using pwm in that sense, I will not be using pwm in the sense of the pwm specific pins on the microcontroller designed to do pwm.  The difference between the two is huge and you fail to see it.  Some pins are labeled pwm and have hardware pwm capapbility or w/e and that is what I was saying I won't be using.  That does not mean a regular digital i/o pin can't be turned on and off in a way that is called pwm.  So you are conflating two unrelated things.

[ebastler]

Although the switching of the mosfets will be happening in a pulse width modulated way, hence using pwm in that sense, I will not be using pwm in the sense of the pwm specific pins on the microcontroller designed to do pwm.  The difference between the two is huge and you fail to see it.  Some pins are labeled pwm and have hardware pwm capapbility or w/e and that is what I was saying I won't be using.  That does not mean a regular digital i/o pin can't be turned on and off in a way that is called pwm.  So you are conflating two unrelated things.

Don't worry, I understand the difference: Driving 60 pins at relevant PWM rates via hardware timers is possible (if you had enough timers); driving them in software on an ATmega is not.

I am really wondering (and a bit concerned) what the end game of this project will be. Will you just quietly drop it and take a proper job at some point, when economic realities force you to do so? Will it end in a breakdown when you realize how you have been chasing a mirage for so many years? Or will you just keep at it and turn into an angry old man talking to ChatGPT?

(Oh, I forgot. You will rule the world and humble all of us, right.)

[PlainName]

If the whole world was against you one would think that there would be a point where one would do a little bit of introspection, but nope not here

This is actually factually incorrect.  I have a significant 50-60+ people who have come out and said that they thought my robot project and skills and knowledge level in the relevant fields etc were all elite, world class, and bound for success in their view.  They staunchly disagreed with the haters.

Identify 10 of them. Or just two if they have standing in an appropriate field.

[xrunner]

That said, I was thinking about it and I can conclude I think I've put in 4-7 months of dedicated fulltime development on this project in that 10yr timeframe. That is NOT including research on all the related fields and note-taking to get me to elite levels of knowledge required to pull this off. That 4 months is assuming full time working on it 10 hours a day 6 days a week with total focus and no distractions during that and adequate sleep and no other major responsibilities competing for my energy and I have great health the whole time.

Hmmm. I just got up and I see your posts span from 11:37 pm yesterday to 6:35 am this morning.

11:37:25 pm
12:18:27 am
12:33:13 am
12:50:56 am
1:58:48 am
3:34:30 am
3:52:38 am
3:56:14 am
4:05:31 am
4:35:29 am
4:42:56 am
4:50:33 am
4:59:16 am
5:12:27 am
5:22:26 am
5:38:13 am
6:35:08 am

Perhaps if you'd use your free time to work on your robot instead of trying to convince others how great you are through the night (it doesn't appear from the responses here it's working), you'd get a lot farther along.

I have a significant 50-60+ people who have come out and said that they thought my robot project and skills and knowledge level in the relevant fields etc were all elite, world class, and bound for success in their view. 

Simple logical fallacy there ^^^ which proves exactly nothing. Argumentum ad populum. The number of people that believe something is true does not in and of itself make it true. Only factual consideration and critical thinking observations makes something true. My observations based on the evidence you've presented in several public areas does not support the claim you have "skills and knowledge level in the relevant fields etc were all elite, world class, and bound for success." Rather than exclaim that to the world over and over it would be better to be humble and show results that prove you are such a person, which I've not seen a single post showing.

[artbyrobot]

Simple logical fallacy there ^^^ which proves exactly nothing. Argumentum ad populum. The number of people that believe something is true does not in and of itself make it true

I didn't say it did.  I said it to refute the claim that "the whole world is saying you are going to fail" when that's not true.

[artbyrobot]

Will you just quietly drop it and take a proper job at some point, when economic realities force you to do so?

Part of the 10 years was spent establishing myself financially and getting multiple streams of income in the works.  I don't need to get a proper job I already work for myself.  So your inference that I'm trying to crank out a robot while ignoring my finances and working on it on borrowed time is inaccurate.

[xrunner]

Simple logical fallacy there ^^^ which proves exactly nothing. Argumentum ad populum. The number of people that believe something is true does not in and of itself make it true

I didn't say it did.  I said it to refute the claim that "the whole world is saying you are going to fail" when that's not true.

Stop focusing on what people say or don't say. Focus on DOING. Your work will speak for you. Do something that shows elite programming or other robotic skills. I've seen no work that speaks to that  - none - Zero - nada.

Now I predict you will again focus on what I'm saying as if it will change any perception of the work you show here. Instead, turn away and make something that's elite, or you will be wasting even more time.

[joeqsmith]

Part of the 10 years was spent establishing myself financially and getting multiple streams of income in the works.  I don't need to get a proper job I already work for myself.  So your inference that I'm trying to crank out a robot while ignoring my finances and working on it on borrowed time is inaccurate.

I know about two of your income streams which are published on the web, including your own recent statement of total cumulative sales in one of them. I have also read your (somewhat older) posts about your living situation. Unless you have something else up your sleeve, I remain unconvinced.

I just assumed it was a scam to generate income. I would say less elaborate than many of the scams posted and less investment on OPs part.  In it for the long haul like so many of them.  But again, at least it's not free energy, perpetual motion, Tesla BS this time. Outside of that, I don't see much difference. 

I have a few small businesses I'm working at, and I'm far from "all set" financially, but I am working on the robot anyways. Plus, I was able to get my girlfriend on board by explaining to her that I could receive money WHILE building the robot in the form of Youtube adsense revenue, pay-per-click ads on my robotics website, and even donations from fans. So with all of that going in my favor, nothing was stopping me and I finally jumped in and started the robot in 2014.

http://www.artbyrobot.com/what-inspired-me

How much money do you think this project will cost?

I anticipate the build will cost me between $20,000 and $30,000 in raw materials. The rest of the cost is my time and opportunity cost. However, I think those will be VERY well compensated for before this project is through!

My twitch streaming name is "Larry Robotnik from Chang Tang, Tibet". I have a background in computer programming as well as a lot of trades (electrical, mechanical, construction, etc). I also am an artist - sculptor, painter. I consider myself an inventor and an innovator. My current inspiration is God.

I possess a photographic memory. Just by hearing the name of an item, I am be able to envision it in realistic detail. I often visualize an invention in my mind with extreme precision, including all dimensions, before moving to the construction stage, a technique sometimes known as picture thinking.

https://www.twitch.tv/artbyrobot/about

[joeqsmith]

Talking about some of their testing and other photos of pulley/string routing not shown here:
https://forum.arduino.cc/t/my-advanced-realistic-humanoid-robot-project/1006814/51

Ok, so 160 pulleys and 40 motors to make one arm.  Then the robot makes the remaining pulleys.   

As I'm only building the one arm and the head and letting the robot build the rest of its own body from there, I just need to build 160 of these. 4 per motor and 40 motors roughly. Also I do intend the robot to have this level of dexterity and I hope it does. If it doesn't then the project is a massive disappointment and failure

***
Doing a search,  I'm amazed of how many fishing nets they have out searching for the next mark.   Must be working for them.  Here they show "Earning Massive Money & Scaling a Videogame Botting Company "

[artbyrobot]

Doing a search,  I'm amazed of how many fishing nets they have out searching for the next mark.   Must be working for them.  Here they show "Earning Massive Money & Scaling a Videogame Botting Company "

If I put out videos on my youtube channel that is meant to be information sharing and telling my story with a DIY bent, how is that "searching for next mark".  What is your definition of searching for a mark?  In my understanding, that means searching for someone to scam or steal from.  If you are accusing me of trying to steal from someone, what is the viewer of this video having stolen from them?  If the person clicking on a video about videogame botting has an interest in that hobby/career path, then this video will be helpful to them and interesting to them to learn about someone elses journey in that field.  Nothing about such a video steals from someone or scams them.  There isn't even a call to action involved.  In fact, it is a generous sharing of my time to create a video like this to share it with others for their benefit.  I used youtube for years to learn how to do things there wasn't much information about elsewhere and a big part of my channel was motivated by my intention to share things I know with others to give back as a way to show gratitude to the many DIYers who shared their fixes or w/e that helped me to DIY my own fixes and learn from their solutions.  So then my channel is largely a rich resource of generous information sharing to add to the database of informational sharing videos out there.  So your mark seeking accusation is a false one.  Do you enjoy false accusing others like this?

Also, as it so happens, my robot project progress sharing is largely motivated by the same generous intentions as much of my youtube channel motivations as laid out above.  To help others who want to build advanced robots but are finding a lack of information on how to do so and give them a open source hand up to achieve their goals.  This is a kindness to share this way and I'm a fan of opensource movement.