I mean, dude, for sure Hazard prevention is your topic. Out of the sarcasm, bad programming can happen, or failure can happen internally. That's why robotic arm setups have many panic stop buttons, also that's why there is a line on the floor so people is out of reach of the arm's movement range.
@@nicolassepulveda4878 yeah I weld with a dual arm yaskawa everyday, unless your stupid and can't control your own actions of input into the robot suurrreeeeeee yeah it will smash through his wall, but I doubt anyone here is oblivious to their actions
@@nicolassepulveda4878 the stops and guards are for auto operation, everyday i move arms that im standing next to you can respect a machine and still be safe
What a challenge, man! It doesn't sound to be a piece of cake, but is definitely worth the effort! And the knowledge you get from this experience is priceless 👌
I'm a robotic student and your video is full of interesting informations, thank you so much ! I had never thought about doing something like that, buying real industrial robot is awesome and I look forward to see the day I'll have the space and knowledge to do what you did there 👍
This is amazing! There is a huge market for aftermarket Fanuc retrofits. If you decided to sell that people would buy them like crazy for old CNC machines running Fanuc drives.
@@theaveragepro1749 On what basis? Reverse engineering for the purposes of compatibility is specifically legal in the US. No one whom Fanuc would hope to sell to would be using a homemade DIY solution instead.
Great vid that shows what you're up against if you decide to acquire one of these units. Unless you want a forever-project, $5-10k sounds like what should be expected.
First time I've heard of a practical use of FPGA-s that I'd be interessted in lol. Also the fact that you can get your hands on a robot like that is next level, if for nothing else, for practice. Just make sure it doesn't put a hole in the roof I guess lmao
This is what hacking means in today's world. The vast spanse of knowledge in various domains like electronics, electrical, mechanical, software, hydraulics, CNC, troubleshooting skills, what not is required to make this happen. Kudos to unkerfuffling so much complexity and getting a working model. Hat tip good sir!
Brilliant work! You stand among the great! Way to push through! You're helping drive manufacturing for the everyday guys. I build machines and could totally use that controller. I just drove the cost on a part from a $1.05 to $.09 and we make a thousand of these parts a day minimum. We need to load the machine once a month while we maintain the ability to pull 1,600 off a day. "So we never have to load it and we're able to pull in $$ish?" Yessir. I'm stoked... I started electrical and mechanical and then found manufacturing. Building with my hands man.
I used to work for an integrator and used Fanuc robots for 5 years. For the past 23+ years, I have worked for Motoman ( Yaskawa). You have your work cut out for you, since you don't have the matched controller for that arm. Great work and way to DIVE in. Look forward to watching your progress. Cheers
Been a minute since I worked on a Motoman. Ours was so old that I ended up buying an old 1.3M pixel camera with a CF slot in it so I could format the card properly (Motoman required a different version of FAT12 which Windows no longer uses).
This was a really great insight into the world of industrial robotics. I find it amazing how, in recent years, industrial machinery like this that could theoretically produce thousands or millions of dollars worth of product over its lifetime can be bought for the same price as a month's worth of groceries. The inflation rate of basic necessities has been far greater than the inflation rate of heavy machinery like this. It makes me think that a surprise North American manufacturing boom might be one of the things that can save our economy.
you are a hero! please keep us updated on the progress on spinning a custom pcb to control these, we use a few fanucs and i've always wanted to get a smaller one to tinker around with but i was worried about all the licensing bs.
I used to work for FANUC North America. If maintained, these robots can last a looooong time. Industrial customers can order replacement parts directly from FANUC or an authorized FANUC dealer with a ZDT contract.
Nice! I was thinking that you could make the robot run faster by changing the mosfets that originally come with the o-drive. something that can handle 300V. You might probably have to design a new PCB for that, but it shouldn't be a big deal. This is really amazing and inspiring work!
Great video! I wish I had enough electronics/software knowledge to really understand everything you presented here. With the right attachments for the end of the arm, this robot would be handy for automotive rebuilders (imagine being able to quickly and easily manipulate the position/orientation of heavy parts during disassembly/reconditioning/assembly). It could also replace a miniature crane in certain applications, especially if the robot was made mobile by some means.
For the record, roboguide ships with all the controller firmware variations so you could have downloaded that and then restored from a virtual robot while using the trial version and saved yourself the custom driver but this is cooler
i'm a cnc machinist and understand cnc code pretty well but understanding and overriding robotic arms is kinda the next level above me. i know these are usually noncollaborative robots and not even sure if these come with a pendant.
Imagine programming one of those crazy industrial robots to play with a video game controller. Everything starts out quaint and innocent. But then the robot loses a life in whatever game it ends up playing, and the video ends with the robot violently bashing the controller and chucking it through a window. I genuinely think that'd be the funniest thing to watch!
I used that R2000 for mig welding at a factory. At floor level its only 4-5 feet tall. You could disable the machine guards to make adjustments. If you get in that things way, it will toss you like a rag doll. I'm 200 pounds!
Amazing! I just saved a arcmate 100ib and r-j3ib controller with pendant from being scrapped. Id love to know exactly what 3phase converter and transformer you used, the vid still is hard to see. Wanting to use it for plasma cutting steel sheets if I can but am mega new to these robots. Any help would be appreciated. ❤
I am not sure if I understood the issues correctly but anyway was an interesting video from start to the end and it seemed the robo was working at the end.
there is a fanuc post-processor for fusion online, I haven't tested it personally, as I still can't get communication from any computer to the rj3 controller on the robot working yet.. definitely an interesting project though, good on you having the patience to go through all this I would have just bought another robot (to be fair that is probably part of the reason our shop has 3 idle robots at the moment though)
I looked into the available post processor a bit, I wasn’t too confident it would work. The only way I had to transfer files to the robot was an ancient flash card.
@@ExcessiveOverkill yeah, my local fanuc guy has been awesome, and I talked with a few people online and go kfloppy all setup, but either my rs232 wiring or something else was never quite sorted so i didnt get file transfer working.. tbh its pretty far out of my element so wiring could very possibly be the problem..
It’s looking like the components to make a higher voltage servo drive are all back in stock, I’m probably going to switch over to those once I get them working so I can go at full speed and get much better control.
@@ExcessiveOverkill Have you looked into servo amplifiers that can be reconfigured to use your motors? I've recently picked up a few Kollmorgen S700 and from initial pocking around in the configuration software they seem to be very flexible as to what they can drive, even ordinary AC motors are supported. They are FPGA based and have ethercat onboard, which is also something I highly recommend looking into.
Your final speed limit is so low because a motor's max speed is determined by it's driving voltage while it's torque is determined by it's driving current. The neat thing is that it's almost a completely linear relationship with those.
Yup, I was expecting a tiny bit faster but it was still usable. The parts for a full voltage servo drive are finally back in stock so those are going to to be one of my next projects.
How did you mount the robot on the concrete? Bolts probably, but I mean they need special concrete block afaik. What if the floor concrete cracks open? Ty.
You are a very ambitious young man! I used to work for FANUC as a field service engineer and your project would be so much easier with the correct software. Did you try to contact FANUC and give their support engineer the F number from the base of the robot? Tell them the controller F number and request an USB be sent to you. Don't tell them anything about this mad scientist stuff or they will not work with you. Just tell them that you bought the robot for educational and demonstrational purposes. The FANUC servo amplifier and controls are far more capable than any home built system. Kudos for getting anything to move safely. Yeah, don't mention that. Also don't talk about the disembowled control cabinet either. Siemens programmable logic controllers can interface with both a FANUC robot and also interface with Arduino controllers running Linux. Oh, using the FANUC servo amplifier allows collision detection which can prevent damage or injury from happening. Good luck to you and work safely!
I always thought there is some dark magic involved in running such robots servos as the dynamics are complex and you need smooth motion, small following error and reduce things like harmonics and etc. I read up a book on servos and the math was quite heavy :D
What was even cooler, back when i worked at Fanuc, was the "float" mode. You can grab the end of the robot arm, and push it around. It will sense you pushing on it and "let" you push it. Obviously you're not strong enough to actually move it, but like a toddler holding your hand, it sense the tiny force you put on the arm and move in that direction. It's also quite impressive to watch the raw encoder values with with the robot sitting still, and just touch the end of the arm. The encoder is so precisce that it can register the imperceiptible movement you add.
Older FANUC robot parts are not harder to get. Fanuc can provide parts for the first robot placed into production. This is true for the software as well.
They can, but I think the real question is if they will. Why help some hobbyist repair an old machine when you could tell them to buy an expensive new one?
@@theninjascientist689 I wasn't replying to "the real question" (what ever that is) but the statement of availability. If you're a hobbyist what you are doing is taking an industrial robot and Frankensteining it. If you have the controller made for that robot (matching F# or E#) with the registered software that robot will perform as good as the day it left the factory as long as the bearings and gears are not worn. There are some much older than that robot still in operation without as much as a fuse blown. You got one that was either abused or installed incorrectly. It will never be worth anything otherwise. And you cannot simply take the core software from another robot and copy it into another controller for two reasons. One if the version is off by a little bit it may not perform correctly. But secondly, it is a crime to copy this software made for another robot. If you had the controller made for that robot you could simply call Fanuc's tech support and they would assist you. If the software is lost (if it's on the controller it's never "lost" maybe corrupted?) you can ask for a reburn and it will cost you about $500. They will email the software to you. Phone support is free as long as you are the registered owner. (Just show proof of purchase if not) Lastly, if that robot could operate the way it did out of the box, you would be placing your life in your own hands if you ran it with it moving in automatic standing that close. Also, with that robot comes documentation on the software and how to setup and program in TPP. Try calling FANUC's hotline.
Haha wtf I've used the ABB EGM interface and was already triggered on how much work it was to get it to run properly but this is a whole new level of external robot control
$450 for a 1.5 ton robot arm?!? I know what I'm getting for when I do my next engine swap! imagine just getting it to lift the engine out or pick up a car or something, that would be so cool!
Is there gonna be more on this? I just purchased a arc mate 100i. I was planning on buying a matching rj2 controller and rewiring for 220v. With the thought in mind you can creat programs with the tech pendant.
Yeah I plan on revisiting it this spring/summer hopefully. It still needs a ton of work to be good though. As long as the controller has the software for that exact robot it should work, but they seem to be pretty picky.
This is very interesting- I am wondering what prevents making a program that accounts for the difference between kinematics set A (controller) and set B (actual robot)? There are off-the-shelf 6 and 8 axis kinematics modelers, maybe more, if you know the length of each offset in Robot A (the controller is programmed for) and B (shown in your schematic) I think it would be workable. Or at least possible.
There are different communication methods between joints and the CPU on various robots, those require more work than simply translating between two vector spaces. These vary in timing, power requirements, and overall data architecture.
Amazing work I love it thanks for sharing. I wonder if you gave up on the fanuc servo amplifier too easily. The optical fiber connection shouldn't be so difficult to tap into. Maybe it's just ethernet and you can wireshark the packets? Someone out there has information on that protocol.
I think it is some custom realtime protocol, I have a large capture of it saved from when the controller was still together. I'll add it to the github repo if anyone wants to have a go at it. I asked around all the forums and found zero mention of it.
That's somewhat interesting to me that the firmware isn't hosted somewhere and readily available. The fact that you have to buy a physical part... to get a piece of software in 2023 is ridiculous
just an FYI. I am not sure if you are aware it seems like you have done a good bit of research on the robot already. I used to build production machiner around the fanuc robotics line at my old job. If you were inclined to you can get the teach pendant for your model and do all of your programming that way. No external software, no license to buy, nothing extra to figure out. Just plug and play. If i cna be any help to you please let me know.
Hi I would get data about status (start, stop, allarm, type allarm) from an olc CNC Fanuc OM-B. What do you think? In your opinion is it possible? How? Are the any parameters sending on serial and CNC answer on serial and I collect on PC? Thanks.
And that folks is why you look for a second hand Kuka robot to play with:) I am now the proud owner of three Krc1 models running windows 95 where you can both upload and download the firmware to your particular robot PLUS it's available on the second disk partition PLUS if you have lost it Kuka will let you have a copy. Still fun to get them doing anything.
Definitely balls of steel having that arm inside where it's well in range to smash through all of those floor joists lol
Not like its gonna move where not programed it doesn't have a mind of its own
I mean, dude, for sure Hazard prevention is your topic.
Out of the sarcasm, bad programming can happen, or failure can happen internally. That's why robotic arm setups have many panic stop buttons, also that's why there is a line on the floor so people is out of reach of the arm's movement range.
@@nicolassepulveda4878 yeah I weld with a dual arm yaskawa everyday, unless your stupid and can't control your own actions of input into the robot suurrreeeeeee yeah it will smash through his wall, but I doubt anyone here is oblivious to their actions
@@nicolassepulveda4878 the stops and guards are for auto operation, everyday i move arms that im standing next to you can respect a machine and still be safe
@@codeack101wlck
I just started using Yaskawa but not for welding. How did you get into that?
you're an absolute legend
Jeez Louise, what a project. This looks like a full-time job
What a challenge, man! It doesn't sound to be a piece of cake, but is definitely worth the effort! And the knowledge you get from this experience is priceless 👌
Trust me when I say that he's simplified a lot.
Like, A LOT.
I'm a robotic student and your video is full of interesting informations, thank you so much ! I had never thought about doing something like that, buying real industrial robot is awesome and I look forward to see the day I'll have the space and knowledge to do what you did there 👍
JUST DOOOO IT, DOOOOOOO IT!
This is amazing! There is a huge market for aftermarket Fanuc retrofits. If you decided to sell that people would buy them like crazy for old CNC machines running Fanuc drives.
I want one
but also might risk legal action from Fanuc
@@theaveragepro1749 On what basis? Reverse engineering for the purposes of compatibility is specifically legal in the US. No one whom Fanuc would hope to sell to would be using a homemade DIY solution instead.
i want a hundred!
Wow. Learning VHDL for a project like this is extremely dedicated! I am very impressed.
the planned obsolescence of these huge indestructible machines is shameful.
Great vid that shows what you're up against if you decide to acquire one of these units. Unless you want a forever-project, $5-10k sounds like what should be expected.
First time I've heard of a practical use of FPGA-s that I'd be interessted in lol. Also the fact that you can get your hands on a robot like that is next level, if for nothing else, for practice. Just make sure it doesn't put a hole in the roof I guess lmao
Whoa, instant sub. You are straight amazing. Electrical engineering still feels like black magic to me.
Yep... instant sub here too... 😎👍☘️🍺
Yes, it looks like a delicate operation testing there. You could scoop out a huge chunk of overhead floor with only a few seconds of inattention.
I hope you will one day develop this project into a business converting discarded robots into useful machines
This is what hacking means in today's world. The vast spanse of knowledge in various domains like electronics, electrical, mechanical, software, hydraulics, CNC, troubleshooting skills, what not is required to make this happen. Kudos to unkerfuffling so much complexity and getting a working model. Hat tip good sir!
Brilliant work! You stand among the great! Way to push through! You're helping drive manufacturing for the everyday guys.
I build machines and could totally use that controller. I just drove the cost on a part from a $1.05 to $.09 and we make a thousand of these parts a day minimum. We need to load the machine once a month while we maintain the ability to pull 1,600 off a day.
"So we never have to load it and we're able to pull in $$ish?" Yessir. I'm stoked... I started electrical and mechanical and then found manufacturing. Building with my hands man.
This is an insane amount of skill and effort. Good job, and that is a wicked project!
HOLY MACKEREL.
This is darn close to being an open source drop-in replacement for an R-J3
I used to work for an integrator and used Fanuc robots for 5 years. For the past 23+ years, I have worked for Motoman ( Yaskawa).
You have your work cut out for you, since you don't have the matched controller for that arm. Great work and way to DIVE in.
Look forward to watching your progress. Cheers
Been a minute since I worked on a Motoman. Ours was so old that I ended up buying an old 1.3M pixel camera with a CF slot in it so I could format the card properly (Motoman required a different version of FAT12 which Windows no longer uses).
This was a really great insight into the world of industrial robotics. I find it amazing how, in recent years, industrial machinery like this that could theoretically produce thousands or millions of dollars worth of product over its lifetime can be bought for the same price as a month's worth of groceries. The inflation rate of basic necessities has been far greater than the inflation rate of heavy machinery like this. It makes me think that a surprise North American manufacturing boom might be one of the things that can save our economy.
you are a hero! please keep us updated on the progress on spinning a custom pcb to control these, we use a few fanucs and i've always wanted to get a smaller one to tinker around with but i was worried about all the licensing bs.
This is amazing who would have thought to use odrive on such a big robot, genius
I used to work for FANUC North America. If maintained, these robots can last a looooong time. Industrial customers can order replacement parts directly from FANUC or an authorized FANUC dealer with a ZDT contract.
Nice! I was thinking that you could make the robot run faster by changing the mosfets that originally come with the o-drive. something that can handle 300V. You might probably have to design a new PCB for that, but it shouldn't be a big deal. This is really amazing and inspiring work!
Great video! I wish I had enough electronics/software knowledge to really understand everything you presented here.
With the right attachments for the end of the arm, this robot would be handy for automotive rebuilders (imagine being able to quickly and easily manipulate the position/orientation of heavy parts during disassembly/reconditioning/assembly). It could also replace a miniature crane in certain applications, especially if the robot was made mobile by some means.
Thats some perseverance man gj getting this far
Just wow. Excessive overkill is right! This was inspirational. Thanks for making this video.
This is the best project on youtube!
Video stalls at 8:14. This is the second day in a row. Skip ahead a few seconds and it plays normally.
Respect. Mad skills. Thanks for publishing this inspiring adventure!
So much work. You deserve a brake from “what is wrong this time?” .
For the record, roboguide ships with all the controller firmware variations so you could have downloaded that and then restored from a virtual robot while using the trial version and saved yourself the custom driver but this is cooler
i'm a cnc machinist and understand cnc code pretty well but understanding and overriding robotic arms is kinda the next level above me. i know these are usually noncollaborative robots and not even sure if these come with a pendant.
Thanks for sharing. An impressive debug and re-engineering project
I wonder, why have i not seen this on LinuxCNC forums? Or did i ? 😃 Magnificent job on bringing the beast back to life!
Впечатляющая работа! Я долгое время работал с похожими, в буквальном смысле, роботами - Hyundai Hi5, Hi5A.
this is beyond impressive, good job
Massive credit to you for tackling that box of worms..
Imagine programming one of those crazy industrial robots to play with a video game controller. Everything starts out quaint and innocent. But then the robot loses a life in whatever game it ends up playing, and the video ends with the robot violently bashing the controller and chucking it through a window. I genuinely think that'd be the funniest thing to watch!
I used that R2000 for mig welding at a factory. At floor level its only 4-5 feet tall. You could disable the machine guards to make adjustments.
If you get in that things way, it will toss you like a rag doll. I'm 200 pounds!
Amazing! I just saved a arcmate 100ib and r-j3ib controller with pendant from being scrapped. Id love to know exactly what 3phase converter and transformer you used, the vid still is hard to see. Wanting to use it for plasma cutting steel sheets if I can but am mega new to these robots. Any help would be appreciated. ❤
I am not sure if I understood the issues correctly but anyway was an interesting video from start to the end and it seemed the robo was working at the end.
Holy shit... Fantastic work. Thanks for sharing.
Excellent work, Sherlock!
I got lost after 1 minute but pretended to be smart by continued to watch to the end
Also, thanks for increasing the cost of robots as I was considering buying a robot. :)
Guys who use a damn logical analyzer are just incredible
Complexity of this sort is often due to layer caking solutions rather than working from first principles. Good luck on your project.
This is a crazy amount of work for a homemade fifi
Was contemplating taking on a similar project, now I’m a little nervous.
Bro is out here rewriting robot arms and I can’t even level my 3d printer bed
there is a fanuc post-processor for fusion online, I haven't tested it personally, as I still can't get communication from any computer to the rj3 controller on the robot working yet.. definitely an interesting project though, good on you having the patience to go through all this I would have just bought another robot (to be fair that is probably part of the reason our shop has 3 idle robots at the moment though)
I looked into the available post processor a bit, I wasn’t too confident it would work. The only way I had to transfer files to the robot was an ancient flash card.
@@ExcessiveOverkill yeah, my local fanuc guy has been awesome, and I talked with a few people online and go kfloppy all setup, but either my rs232 wiring or something else was never quite sorted so i didnt get file transfer working.. tbh its pretty far out of my element so wiring could very possibly be the problem..
Yo. 8:42. From 0 to hero right there. Nice work!
Awesome video, im in the same project right now, but i already started the motor swap
This is impressive ! Good job 👍🏻👍🏻
Such an awesome video & channel !!
love the video! I really want to make the jump from mills into CNC with robot arms like you.
Kudos to you for taking such a herculian task upon yourself :)
Any updates on the project?
It’s looking like the components to make a higher voltage servo drive are all back in stock, I’m probably going to switch over to those once I get them working so I can go at full speed and get much better control.
@@ExcessiveOverkill Have you looked into servo amplifiers that can be reconfigured to use your motors? I've recently picked up a few Kollmorgen S700 and from initial pocking around in the configuration software they seem to be very flexible as to what they can drive, even ordinary AC motors are supported. They are FPGA based and have ethercat onboard, which is also something I highly recommend looking into.
Your final speed limit is so low because a motor's max speed is determined by it's driving voltage while it's torque is determined by it's driving current. The neat thing is that it's almost a completely linear relationship with those.
Yup, I was expecting a tiny bit faster but it was still usable. The parts for a full voltage servo drive are finally back in stock so those are going to to be one of my next projects.
channel name checks out.
This is something that I cant even dream of..!
awesome work man!
I just bought two....anchored them in my garage 4 ft apart and I had them arm wrestle and the one loss because the anchor bolt came out of the ground.
It's really owesome. It definetely needs a lot of time. Wondering java or c supports automation
Wow! Very nice work. Thanks!
Great Video!!
How did you mount the robot on the concrete? Bolts probably, but I mean they need special concrete block afaik. What if the floor concrete cracks open? Ty.
You are a very ambitious young man! I used to work for FANUC as a field service engineer and your project would be so much easier with the correct software. Did you try to contact FANUC and give their support engineer the F number from the base of the robot? Tell them the controller F number and request an USB be sent to you. Don't tell them anything about this mad scientist stuff or they will not work with you. Just tell them that you bought the robot for educational and demonstrational purposes. The FANUC servo amplifier and controls are far more capable than any home built system. Kudos for getting anything to move safely. Yeah, don't mention that. Also don't talk about the disembowled control cabinet either. Siemens programmable logic controllers can interface with both a FANUC robot and also interface with Arduino controllers running Linux. Oh, using the FANUC servo amplifier allows collision detection which can prevent damage or injury from happening. Good luck to you and work safely!
I didn’t have any luck sneaking past Fanuc, even with the education school excuse, couldn’t even get manuals from them☹️
The price of that robot is cheaper than most CNC tooling you have to buy, insane
Good robots, terrible company.
Nice work!
Mad skills! Well Done!
I always thought there is some dark magic involved in running such robots servos as the dynamics are complex and you need smooth motion, small following error and reduce things like harmonics and etc. I read up a book on servos and the math was quite heavy :D
What was even cooler, back when i worked at Fanuc, was the "float" mode. You can grab the end of the robot arm, and push it around. It will sense you pushing on it and "let" you push it. Obviously you're not strong enough to actually move it, but like a toddler holding your hand, it sense the tiny force you put on the arm and move in that direction.
It's also quite impressive to watch the raw encoder values with with the robot sitting still, and just touch the end of the arm. The encoder is so precisce that it can register the imperceiptible movement you add.
Great video. Where did desk chair come from?
Amazing project!
I am encouraged to buy one now.
Older FANUC robot parts are not harder to get. Fanuc can provide parts for the first robot placed into production. This is true for the software as well.
They can, but I think the real question is if they will. Why help some hobbyist repair an old machine when you could tell them to buy an expensive new one?
@@theninjascientist689 I wasn't replying to "the real question" (what ever that is) but the statement of availability. If you're a hobbyist what you are doing is taking an industrial robot and Frankensteining it. If you have the controller made for that robot (matching F# or E#) with the registered software that robot will perform as good as the day it left the factory as long as the bearings and gears are not worn. There are some much older than that robot still in operation without as much as a fuse blown. You got one that was either abused or installed incorrectly. It will never be worth anything otherwise. And you cannot simply take the core software from another robot and copy it into another controller for two reasons. One if the version is off by a little bit it may not perform correctly. But secondly, it is a crime to copy this software made for another robot. If you had the controller made for that robot you could simply call Fanuc's tech support and they would assist you. If the software is lost (if it's on the controller it's never "lost" maybe corrupted?) you can ask for a reburn and it will cost you about $500. They will email the software to you. Phone support is free as long as you are the registered owner. (Just show proof of purchase if not) Lastly, if that robot could operate the way it did out of the box, you would be placing your life in your own hands if you ran it with it moving in automatic standing that close. Also, with that robot comes documentation on the software and how to setup and program in TPP. Try calling FANUC's hotline.
Haha wtf I've used the ABB EGM interface and was already triggered on how much work it was to get it to run properly but this is a whole new level of external robot control
Got a price from kuka. 360 000k for a cnc milling robot.....
Includes setup and startup.... But still..
$450 for a 1.5 ton robot arm?!? I know what I'm getting for when I do my next engine swap! imagine just getting it to lift the engine out or pick up a car or something, that would be so cool!
Wow sick!!! Dude you are a genius
What a rabbit hole...
Is there gonna be more on this? I just purchased a arc mate 100i. I was planning on buying a matching rj2 controller and rewiring for 220v. With the thought in mind you can creat programs with the tech pendant.
Yeah I plan on revisiting it this spring/summer hopefully. It still needs a ton of work to be good though. As long as the controller has the software for that exact robot it should work, but they seem to be pretty picky.
Algorithm has blessed You lol. Great video, You should post an update or something :) What did You plan on using it for? Just for fun?
which programs do you use for motion planning ? Like MoveIt ?
Great work!
make a giant 7 axis CNC center, place the robot on a circular rail and add tool to its arm
Would this be overkill for an automated litter box cleaner? My cat poops a lot and those clumps are surprisingly heavy.
Amazing project and rundown of Fanuc servo interface. Thanks a lot for the video!
Nice work!
This is very interesting- I am wondering what prevents making a program that accounts for the difference between kinematics set A (controller) and set B (actual robot)? There are off-the-shelf 6 and 8 axis kinematics modelers, maybe more, if you know the length of each offset in Robot A (the controller is programmed for) and B (shown in your schematic) I think it would be workable. Or at least possible.
There are different communication methods between joints and the CPU on various robots, those require more work than simply translating between two vector spaces. These vary in timing, power requirements, and overall data architecture.
seriously impressive ..
Nicely done.
Very nice, if you can come up with a package that replaces the original Fanuc controller boxes I'm sure it would have a decent sized market.
Amazing work I love it thanks for sharing. I wonder if you gave up on the fanuc servo amplifier too easily. The optical fiber connection shouldn't be so difficult to tap into. Maybe it's just ethernet and you can wireshark the packets? Someone out there has information on that protocol.
I think it is some custom realtime protocol, I have a large capture of it saved from when the controller was still together. I'll add it to the github repo if anyone wants to have a go at it. I asked around all the forums and found zero mention of it.
wow, great job keep it up
That's somewhat interesting to me that the firmware isn't hosted somewhere and readily available. The fact that you have to buy a physical part... to get a piece of software in 2023 is ridiculous
just an FYI. I am not sure if you are aware it seems like you have done a good bit of research on the robot already. I used to build production machiner around the fanuc robotics line at my old job. If you were inclined to you can get the teach pendant for your model and do all of your programming that way. No external software, no license to buy, nothing extra to figure out. Just plug and play. If i cna be any help to you please let me know.
I actually did have the pendant but I wanted to try to find a way to run more complex programs like CNC toolpaths.
Hi I would get data about status (start, stop, allarm, type allarm) from an olc CNC Fanuc OM-B. What do you think? In your opinion is it possible? How? Are the any parameters sending on serial and CNC answer on serial and I collect on PC? Thanks.
And that folks is why you look for a second hand Kuka robot to play with:)
I am now the proud owner of three Krc1 models running windows 95 where you can both upload and download the firmware to your particular robot PLUS it's available on the second disk partition PLUS if you have lost it Kuka will let you have a copy. Still fun to get them doing anything.
I'd have considered reverse engineering the ROMs as one of the options. It sounds like you'd just have needed to locate and change some constants
You actually do have access to those parameters through the controller, but knowing what they do and which need changed is a whole other issue
You have a massive brain - that is awesome :)
Excellent video!!!