Last Threads: https://warosu.org/diy/?task=search2&search_username=artbyrobotTo begin: the project goal: I am working to make a series of humanoid robots. I am using a Biblical theme of naming the first 3 robots I make Adam, Eve, and Abel. The goal is for these robots to have human body inspired musculoskeletal systems, advanced AI, and that they look human and pass for human to a casual observer at least at a distance. They must be able to walk, talk, run, dance, do sports, do chores, manufacture products, and make more robots just like themselves if not even better. My aim is to build a single robot arm and head and then add sufficiently advanced AI to that arm and head to enable it to build the rest of its own body for me. This way I am delegating the work of building the majority of my first humanoid robot to that robot rather than doing that work myself - and this is to save me time.In a like manner, my goal with the AI is to code just enough AI that the AI can begin coding itself and this way I don't have to code most of the AI myself because it will self create itself. I liken this to building a seed and that seed growing into a tree because for me to code that tree would take too long for me and just creating the seed would then save me time.
>>3005837Show prototype
>>3005837i often wait till i have something show to post but how you just make some place holder thread is gross. it tells me what kind of person you are
>>3005529>He's a bit meh at it sure, but I wouldn't say bad. He has helped me immensely in this area, bringing awareness to me of things I'd have never known even existed. This is invaluable. Game changing.lmao actual living people gave you tons of advice and valuable critique months ahead of your 'invaluable discoveries' with chatgpt. And all these discoveries are is just changing trajectory and plans due to complications that arise from implementing any significant advice by the AI. Chatgpt is a data skimming machine, btw. It will ask you a lot of questions about your work and ask for a lot of photos but in return won't even be able to draw correct pinout for an IC half of the time.The most insane part of all of this is that you pathologically defend chatgpt that's not even specialised in robotics or electronics but disregard real critique and advice from people with experience. You think you will become some kind of robot master while being led on a wild goose chase by a mediocre chatbot
>>3005837>code just enough AI that the AI can begin coding itselfwhy do you think you can do this when every major tech company is attempting to make their already capable ai systems better through self training and ai assisted engineers with literal trillions of dollars worth of incentivenot to mention this is near the top of the list for most major governmentsobviously if the best newest version of claude fable with unlimited vram and unlimited processing power couldn't make a better ai, why can you?
>>3005940>Game changing.What's even more astonishing is that people used to design and assemble absolutely amazing circuitry before AI was even a thing. The whole consumer/business chip ecosystem was set up in such a way that you should be able to find what you need because that's a sale for the manufacturer - so you had whole yellow pages sized catalogues of ICs and advent of internet brought in databases with every possible spec to filter your searches on wholeseller's website. All you have to do really is to go to a website and select what you're looking for. chatgpt has nothing to do with this.Your problem though is that you don't really know what you're looking for because you have no meaningful knowledge in the field and refused to gather it for past 12 years. Think about it; CTE takes 4 years. You're not a child anymore so you'd rather go to community college, which takes 2 years. Both of these institutions would teach you how to build basic electronics, giving you a good deal of momentum to build up on. Instead, you absolutely wasted those 12 years of project time so you're condemned to buying bogus parts and playing with kapton tape for 6 months by your own ignorance.
>>3005529>>if you can't be bothered to improve at soldering/handling electronics and try again?>the first try I probably got 98% there but it had to be 100% perfect to work.No shit, why don't you try again and improve to 100%?>And finding that last 2% was extremely difficult to do because you can't inspect connections that are visually obstructed on the bottom of a chip blah blah my scenario is so dramaticIf you can't inspect them then don't - just solder again. Do you think SMD soldering is some kind of fucking epic feat? Why can't you be bothered to improve at soldering?>Now I would gladly do this,Yeah you would if you were good at soldering, but you're not. You can't even hypothesize because you simply can't, you can't choose to do this because you never chose to improve.>but at that point, seeing how the frustration ramping was climbing rapidly, I looked for an easier alternative. To potentially act as a release valve of mounting frustrations.Yeah you gave up after one failed attemptalso>frustration ramping was climbing rapidlyJesus Christ we're talking about one smd here
>>3005837So, still 0 progress, no prototype no nothing?
for whatever abuse I get on 4chan I know one thing: if it weren't for you guys I would not have found the ESC I'm going to be using for my robot's motors so I'm thankful to you guys for that. That's a game changer for me. I'm working on its firmware now.
>>3006043CAD for all systems of entire robot done in details fully - checkPhysical skeletal structure of arm fully assembled and built and tested - checkFirst motor mounted and custom pulley based downgearing system designed and implemented (albeit roughly) - checkelectronics for that first motor made ready - checkcustom firmware for that first motor coding - in progress.
>>3006054I respect the fact that you want to be fully vertically integrated but don't you think testing one (1) motor and one joint before doing the full design would've been smart ?
>>3006049You know, for the ESC, there's a lot of 4in1 ESC that you can buy if you are concerned by size it might be interesting for you to look at, since one ESC can cover 4 motors. Actually I work in a very big company in Taiwan that makes ESC and QDD control boards so I might have a lot to say about your idea, I'm not going to read all your threads tho so just TLDR me if you are interested. For the firmware part you can consider Odrive, they have published the code already
>>3006063No I don't. To know which motor would fit and would be strong enough you first need the full design of the entire robot to have the ability to determine how much space you have to work with, how strong and fast it has to be etc. Then you can pick an appropriate actuator and select appropriate downgearing and if appropriate downgearing does not exist, make your own - which I did. Only then are you in a position to make the robot frame, mount electronics, mount the customer downgearing, and eventually then be capable of testing the first joint. And that test will expose all issues with the downgearing, the motor, the electronics, etc. Which is what we are moving toward ASAP but has eluded me for a long time due to life and due to rabbit holes upon rabbit holes I had to traverse to get this far.>>3006065no, I don't have space for a 4 in one ESC. Having each ESC separated enables me to distribute it one per motor next to the motor whereas a 4 wide would be too big to fit anywhere in the arm. The ESCs I have do fit already and will work I think. As far as Odrive, I thought that's FOC firmware made for Odrive boards. I am not aware of it working across just any ESC like you imply. I don't think it would. But in any case I don't think it would have the specific behavior we want and it probably is closed loop which I plan to not have in my firmware. In any case, the firmware of the motor has a HUGE role in overall control of the robot and latency as well and how the communication is as far as robustness and latency and so rolling my own firmware will ensure success in all of these critical areas and so I'm not open at this point to any firmware advice I'm already in DEEP to making my own and very far along and its going very well and I'm very positive about it, have great momentum and have total confidence in what I'm making. What I will say is I recall Odrive had like $100+ motor controllers and so I consider that company to be predatory ripoffs.
>>3006054Sticking primitives inside a 3d skeleton still isn't CAD, my dude.
>>3006090>designing a robot on computer 3d software is not computer aided design (CAD)ok I used my computer to make my design but just because the entire design was aided by the help of a computer doesn't mean it's comuter aided design - says the retard?
