3D Printer Farm - Robot-based Automation

[@ApacheFPV]

Ah yes a robot to take care of my robots

[@DavidGarcia-nx2gj]

@Nature Taco how profitable could be that? because I don't think it is

[@brndto]

@@DavidGarcia-nx2gj gcode is free and needs no materials??

[@DavidGarcia-nx2gj]

@@brndto and the robot is free or what? That doesn't look cheap not a single bit

[@gwagwa76]

@Nature Taco yeah it’s a bit over engineered lol

[@thunderbolt997]

@@DavidGarcia-nx2gj in a worst case scenario (high cost of automation low profit on products) say this robot solution costs 20 000 euro and you sell 3d printed products with an average price of 20 eur you will pay this robot back in 1000 orders if average printing time is 4 hours and produce parts 24/7 it will take 5.5 months to pay back

[@RomanDvoryadkin]

I think you can use one spare flexplate. While one print job is finished, robot can grab flexplate with printed part and immediately install spare one to minimize printer wasted time while printed part cools down.

[@D4rkS7der]

I had the same thoughts when watching this :-)

[@jed3219]

the only con I can think of to doing this is the spare plate would need to be preheated in order to reduce downtime on the new print as far as bed heating is concerned, but yeah that would definitely shave off some time that'll surely add up on longer runs!

[@MikiCab1]

haha. That thought came into my head while watching. 2.0 SW/FW upgrade with minor hardware upgrade.

[@TheyTookStrawb]

​@@jed3219the plates are so thin they probably cool down so much some the parts are being removed that the difference is negligible

[@JohnAldred]

This is beautifully overkill! :)

[@WisienPol]

it is just demo version ready to scale up to 10s of printers IMO

[@snan1384]

Not necessarily overkill - it seems to be easily scalable, and if you had to hire employee to do the same thing, it will eventually become cheaper at some point of time.

[@andrewnelson3714]

it is most certainly overkill, the same result could be acheived with a cold release buildplate and simply modifying the Gcode to push the part off the bed using the extruder, into a bin.

[@Rodney.or99]

@@snan1384 If you run this 24/7 then you break even pretty fast.

[@meercreate]

@@andrewnelson3714 I never thought of using the extruder as a pusher

[@DJWhitelow]

Im pretty sure Prusa would have done this with their 500+ print farm if it genuinely could be so hands off, but still really impressive! Can totally see the potential when the printers themselves get reliable enough.

[@dumle29]

Easy work-around for that is just to not try and remove the prints from the spring steel, and just have cleaned and prepped spares. I suspect the main reason is that the room the automation takes up could be better used for more printers.

[@CrashPCcz]

Yes, one or two technicians can handle dozens of printers and troubleshoot. If anything, then belt printer might be better solution.

[@TrolloTV]

This is just crazy overengineering and not at all practical for running 3d printers that are that cheap

[@aimansoul]

the (almost) same system is also used in my University to fully automate 3d printing and it look so cool seing a robot arm doing all the work

[@LilApe]

This just keeps proving further how great the prusa printers are.

[@Marco-yk8kp]

or how smart people are, this could be done with ender 3s, vorons, even ultimakers. Prusas are the least impressive part of this

[@LilApe]

@@Marco-yk8kp That is true, But i'm just implying that using prusas in an automated production environment is just a testament to their known reliability. To simply press print and walk away, knowing you wont have errors or failed prints is what makes a prusa, a prusa. With enders, you're going to periodically have failed prints, errors or replace parts. Many people i know just throw out their creality printers and buy a new ones when parts wear out.

[@yourlocaltoad5102]

@@LilApe then you know people that seemingly have more money than brains. A Ender 3 with a few simple upgrades and proper slicer profile will print just as good as a prusa for a much cheaper price.

[@LilApe]

@@yourlocaltoad5102 no, they have more brains and money. Because anyone in a serious production environment isn't going to spend time modifying printers, just to get them to work as well as a stock prusa. That's lost time in manufacturing. For a hobbyist, sure. But a serious operation needs a reliable printer that will work lights out and un attended, out of the box. That's a prusa.

[@75echo]

Impressive, now I need to buy one to feed me grapes while I lie on the couch watching south park.

[@l3d-3dmaker58]

OMFG yess I need to make one but diy! I've wanted this for so long, especially the individual bins

[@tizio5103]

Dang, I wish my 3DP was reliable enough to even consider this. I'd end up with a pile of filament hair in my bin.

[@Ender_Wiggin]

I love it! you could have a factory on the side of your garage!

[@theodoreiliaskos]

well let me show you what over-engineered really means...

[@crowguy506]

Now use Vorons and the room is flooded with finished prints in a heartbeat.

[@dunichtich100]

That was also my thought when I saw this setup 😂👌🏼

[@ryanokeefe12]

The pace that these things will be outputting parts would not justify the additional complexity for sorting into bins.. You could put the print farm between a desk and the bathroom/kitchen and each time there is a part that has been removed from the printer, someone could part it in the correct bin... The sorting machine is not saving any time.

[@joaquimcastillaizquierdo783]

@@ryanokeefe12 If you run it 24/7 yes it justifies the complexity

[@Leo.Wolf.the.Engineer]

Vorons are not well scalable for printing farms because you can't just buy them off the shelf and be ready to print

[@cygnae]

@@ryanokeefe12 unles you need to, you know... sleep.

[@RocketCardragracing]

Not me thinking I could totally build one of these, but honestly great work, well made, been thinking of building one, hobby grade i could see being hard to commercialize but metal printers on the other hand, but still very impressive!

[@farkhodkhikmatov4738]

just learn plc programming and industrial robotics. get your self a used robot and design something like this

[@jhsevs]

@@farkhodkhikmatov4738 «just»

[@SeanHodgins]

Thats awesome!

[@DougFeigelson]

Really wanted the robot to start the next print by pressing the control knob :D

[@adammichaeld]

This is seriously impressive. Fantastic job!

[@dormitor187]

Wow dickes Respekt für die Umsetzung und vor allem Neid für diejenigen, die es Umgesetzt haben und auch Warten dürfen :-)

[@Charlie-tl4dv]

I would have another group of trays shelf so, that replacing a new tray would be the first thing after removing the final printed one, so the machine can start printing straight away.

[@ProDesigneHd]

@@renevile what if you have multiple printers that finish their prints close to the same time. You would loose a lot more time and that adds up to a lot of missed parts when you consider that the system might be running 24/7 for a year

[@denbeech6647]

Wow!! Thanks for sharing.

[@Koh-Wei-Jian]

one day the robot is going to print itself.

[@victorrzhang]

Love the concept! I wonder if getting 9 belt printers would be better since a CR-30 is basically the same price as an assembled Prusa i3, saving the cost of the Wittman. Although the CR-30 is a lot more work to tune and learn to use.

[@jonasstahl9826]

I dont think so the Wittman can take care of probably 100 3d printer, depents on the printing time. When using belt printer you still need a system to sort the items.

[@volkhen0]

Yes, you can probably extend it horizontally so one robot can take care of 100 printers

[@biziluxgames8924]

cr-30 will print nowhere near as good as i3... i wish it did, but maybe some other models in the future...

[@deathmock5]

On print complete, cold shock the bed, freeing the part. use the print head itself to push it onto a chute. Much more scale-able then a few thousand dollar robot arm. Can also place the printers themselves into an enclosed room for temperature regulation. But if you want to go the robot arm route like that, Move the printers away from the wall. Sandwich the arm between 2 layers of printers, like an Oreo cookie, but the arm is the filling. Increases your printer density, and back of the printers now becomes your service access.

[@vedantparanjape]

I find most robotics application to be idiotic in some sense. They try to solve a problem with a fixed subset of solution. I saw a video of swarm of drones which detected gas leaks in a room. I was like bruh, gas leaks and you gonna send BLDC run drones there 😂

[@infinitewars6373]

i was more impressed on that green vase on right hand side :DDD

[@hamiltongazim7094]

the best video , i like this congratilations

[@Univercius]

I feel like more 3D-printers would be needed for this to be economical.

[@hindugoat2302]

it only has to replace 1 human worker to be economical

[@onlyeveryone2253]

@@hindugoat2302 of the human can do other stuff in between the prints maybe not

[@hindugoat2302]

@@onlyeveryone2253 humans cost a lot of money, automation works 24/7. The idea is to minimize costs, that means minimize workers

[@onlyeveryone2253]

@@hindugoat2302 There is always a calculation. a robot costing 1m will only be worth it after x-years. If x is too large, the inflexibility of the robot may become a problem.

[@hindugoat2302]

@@onlyeveryone2253 there is a calculation... to maximize proft and minimize expense. that leads to less human workers. Sure this machine has limitations, but would be redesigned over time and get better. Its more a proof of concept of automation

[@HerukaSan]

Awesome!

[@Nurf]

totally not unnecessary or overkill love it

[@CallMeSwal]

Nice! i spent the last 8 months trying to do this at my school so I know how painful this can be 😂

[@SalihArtAndTech]

Awesome👍

[@rafael12348]

I would like to give an idea for improvement: When the robot removes the part from the printer, there should already be another empty print bed for the robot to change, 0:47 to start the printer faster, then the robot would use the idle time to place the piece in the box. This way you don't waste time putting the item in the box.😁

[@Keksstar]

Wondering why James Bruton doesn´t have one of these yet :-)

[@Evgeniy_M87]

Очень интересно реализовано! Автор молодец! Успехов вам!

[@jackthehacker05]

Minecraft redstoners when they get their degree:

[@jetstar4609]

Would love this in my office!!! More free time for me!

[@uint16_t]

What happens with the nozzle purge?

[@aliaburaia111]

This is something we still have to fix manually…

[@a1xon]

@@aliaburaia111 You could increment the x-axis for every purge? or maybe print more lines (z-axis) on the purge so it get's also removed by the twist and bend robo?

[@xagels]

@@aliaburaia111 shut up and take my money xD

[@unknown6656]

pfft ..... real men don't clean their nozzle after dumping their load on the (print) bed.

[@spamspasm8183]

I have been looking for a solution like this for years. Been trying to build one too on and off but repeatability and reliability have always been a problem. Where can I learn more about your project?

[@internettoughguy]

He didn't build that thing. He bought the gantry from Wittman. They specialize in automation systems.

[@MrButuz]

Half a million dollars and you can have one of these! :S

[@DRSElectronic]

Very impressive

[@Trashalchemy]

I love it but I think a few belt printers would be the route I'd go for this level of automation.

[@andre267]

This is insane automation!!!

[@codenameFXX]

this is very cool... i always dream of this machine for mass producing my robot parts

[@aburaiam]

great project!!!

[@sleepib]

Neat, but I think it would be cheaper and more productive to double the number of printers and check them once or twice per shift.

[@KnurdMonkey]

This is frekin awesome!

[@JohnBurd]

This would be a dream.

[@sevenpounds1754]

💥 💥 💥

[@nordemoniac]

All I need now is a robot robot vacuum cleaner cleaner.

[@cplenny4281]

Beautiful!

[@ParsMaker]

amazing

[@raghavgarg6909]

This is so cool!!!

[@gillisdebilio4516]

I like the robots design, it takes less space than the floor mounted ones. And would only require changes on travel length to accomodate more printer, from a 3x3 setup to a 3x6 setup.

[@JoergGebhard]

Could you share some numbers for the Costs? I would be nice to get a feeling how many Printers have to be operated to justify the use of the robot.

[@X197ToPlay]

its a gantry not a robot. i assume around 10k for the gantry

[@reformed_attempt_1]

@@X197ToPlay and this gantry is... drumroll... a robot!

[@CC-kc5lb]

Whitman lol I remember as a young guy then vacuumed loaders and McGuire hmi

[@tamgaming9861]

One thing i miss. Plates didnt get cleaned so that the next part can stick well on it. Its recomended from prusa to do it after each print, or nozzle nozzle cleaning.

[@AlexanderNash]

I was thinking there could be an extra clean print bed ready and it immediately swaps it in when it lifts the other print bed. Then the printer could start right away and you could have some sort of cleaning system to take care of the used one.

[@devanshgarg31]

Make a giant robot to control other robots. Wow

[@Fizzi_]

I feel like this is the most over engeneered thing but it's really cool

[@sabriath]

All we need is a parts builder robot, then a way to print 3D printer parts, and then you have self-replication. Add in a little AI in order to find the optimal solution and paths for all parts, and eventually you get the singularity, autonomy, and our robot overlords.

[@АндрейГорбунов-и9р]

Вот это подход! Вот это я понимаю!

[@watsonking3646]

Good job. But it should be put one or more plat to standby. Let the printer can faster working.

[@grappeq]

This feels a little bit like real-life Factorio.

[@PhG1961]

Impressive !

[@jbergene]

ok im returning my engineering diploma.

[@caiocallor]

"And then, one print fail goes miserable wrong"...I forsee 9 prusas deep in plastic entanglement for eternity

[@AliOzdemirMHC]

Que de l'excellence massallah

[@maxbursell3513]

come on PrusaResearch. do the same with the 700 printer farm!!

[@EpicHardware]

90 people disliked over the music. undefeatable

[@aliaburaia111]

i'm sorry :)

[@dunichtich100]

That's the future!

[@chrisnurse6430]

With a bit more thought the carriage could be made to push parts off or a motor added to flex the print surface and with the printer tilted forward parts could fall off on to a slide and into a bucket. The robot could go to the pub?

[@freezyAndHaze]

Flipping amazing :o

[@nigelhungerford-symes5059]

so cool

[@lemon9.9]

Imagine in the future where you a 3d printer and print out 8 more printers, then print a robot arm with all the printers and start manufacturing other stuff

[@Spartacusse]

I hate to think how many little plastic parts one has to sell in order pay for that terminator. Nevertheless it's mesmerizing to watch, nice video.

[@paulsanka5575]

This is just brilliant 🙏🇬🇧

[@adamgalloway8441]

I need this

[@olavodias]

That’s awesome

[@BharathwajK]

Here the skynet is being planned 😀😀😀

[@flujori]

this is indeed beautifull overkill, i love the setup but its not optimised in the way that many of the actions can be simplified a lot.

[@gasser5001]

That is awesome!

[@marcusj1710]

have you had any success with automating filament replenishment? I am struggeling to do so

[@aliaburaia111]

we're working on a solution right now, should be ready in about 2 months - stay tuned :)

[@hunterjones9822]

All fun and games until the Robots start building Robots to replace the humans in the equation...

[@rickywong5551]

Love this

[@bencusb]

The thing that I don't get is the part where the arm pulls out the tray, because that's the only reason for that very big arm. I would make the contents of the tray slip off with that existing sweeper and lay it into a conveyor tray system.

[@Skott62]

Pretty cool!

[@GoingtoHecq]

This seems extreme

[@Madbot_]

just german things

[@tomyocom5886]

I did not see the printer RESTART on bottom Right. Where was that info. That has to be somewhere? Did I miss something, the restart of the printer? If not then this would be superfluous ......

[@TobiasHansen]

I wonder how they handle oozing and it doesn't seem like they are using the start purging part of the gcode.

[@aaronenriquegarcia7350]

I don't know wich cinematic system is better. System of pulleys or endless screws for all axes

[@martintaichl]

looks excelent. However I am wondering, how do you solve it when you run out of fillament? I guess it still has to be manually exchanged, right?

[@aliaburaia111]

It’s still manually

[@martintaichl]

@@aliaburaia111 thanks. Well you have achieved a masterpiece 😉

[@nel2834]

Es stört mich sehr, dass der Roboter nach entfernen und ablegen der Druckplatte nicht eine bereitliegende leere einsetzt, damit der Drucker direkt wieder starten kann wärend das Teil aufgeräumt wird.

[@aliaburaia111]

danke für das Feedback. Wir arbeiten daran :)

[@izzayroman7283]

I’ve never use this type of printer so I don’t know if it has automatic bed leveling. But if it does and it automatically reload filament this was sent from the mechanical gods

[@mnantel3861]

That printer could have dropped the piece straight nin the bucket, really nice machine anyway.

[@marks47]

Does the design of the parts change often? Seems like it would be worth injection molding unless the # of specific parts is cost prohibitive.

[@alexandrefournier-ahizoune8098]

what happens when a spool is empty ?

[@reformed_attempt_1]

respectable

[@BrotherCreamy]

The cartesian robot probably costs 10x what all of the 3D printers cost. You should design a low cost cartesian robot to tend the 3D printers. These IMM cartesian robots are designed for very high speed/acceleration... i.e. completely overspecced for this application.

[@jonhaze7537]

Prusa ftw

[@AnnKinderknecht]

lo mas bello que vi fueron las prusas

[@exssaturn]

You still have to change the filament by hand right? right? :D

[@mmawad100]

did you consider getting a plastic injector and a desktop cnc for mold making and if you did why wouldn't it work out for you ?

[@aliaburaia111]

our focus is more on flexible production and additive manufacturing. more information coming up soon - stay tuned!

[@金网宽]

Nice

[@garrydye2394]

Even the Prusa printers are not reliable enough at this point.