Hey guys! I’m a PhD student at the University of Michigan and I’ve been doing research on how to make 3D printers print faster. We’ve come up with a software solution called Ulendo that seems to work really well. I’m interested to hear what you all think about it. Check out the video: https://www.youtube.com/watch?v=w-_gY23yGwY
We’re also doing a beta program right now so you can sign up to try Ulendo for free on our website: https://www.ulendo.io/
We’ve actually set it up to run alongside the printer’s firmware. So once you set up Ulendo, you can either run a print with Marlin or run it with Ulendo. This gives you the option to use it when you want.
And no specialized slicer needed. Just save the Gcode with your favorite slicer on the Pi and print.
Interesting… does this only change non-printing travel time or does it affect travel time while laying down filament?
I did look at the beta, but noticed it currently requires TAZ 6 (I have a Workhorse & Pro).
Also… you might want to look at OctoPrint / OctoPi. OctoPi is a Raspbian OS image meant to act as a print server/controller so you don’t have to keep a computer tethered to the printer and you can control it via any web browser… in addition to remote monitoring (you can use a camera to make sure the print is still going ok and cancel it if you find your making a plate of spaghetti).
This is the sort of thing that might work well as a plug-in module for OctoPrint. (OctoPrint is the software … OctoPi is the OS image for the Raspberry Pi that runs OctoPrint).
@TheVirtualTim it affects travel time while laying down filament. And we plan to expand to other printers soon, so we’ll hopefully have support for the Workhorse and Pro soon.
Connecting Ulendo with OctoPrint is a great idea and we’re actually already working on it! Stay tuned for updates on that as well
I’m in Dearborn… not so far from Ann Arbor. So if you want to run tests on a TAZ Workhorse or TAZ Pro … I have one of each and am happy to run jobs.
And for grins… the spacetime interval is ∂s = √ ((c∂t)^2 - (∂x)^2)
Which says … as velocity increases towards the speed of light… time slows down.
If you can get the printer to move at the speed of light… then the print job will finish instantly. (With apologies… the physics/astronomy geek in me couldn’t resist.)
Ok, so my tongue is firmly pressed in cheek … but on a more serious note I am very happy to learn that you are working on a way to increase print times and looking forward to reading about your progress.
Would this shorten the life expectancy of the machine itself- wearing out certain pieces a bit quicker than normal? On my machine I had to slow it down a bit because some of the 3d printed parts started to crumble due to high usage. (the bu**er of it is that you need a 3d printer to print out the bits that’s broken?- joke.)
This is a great question! We haven’t done a detailed long-term study of how Ulendo changes the life expectancy of the machine, but my inclination is that it would not necessarily shorten the life expectancy.
The software actually modifies the commanded trajectories to prevent vibration on the machine, so, for the same print settings, printing with Ulendo would lead to less vibration than without, which would ideally increase the life span of the machine. Of course, then the question is that if you slowed down the machine, would this still be the case? And if so, what would the speeds that you run Ulendo at need to be to get the same wear over time? That, we don’t know yet.
Okay, so it uses a raspberry Pi to do some calculations and reduce vibrations so you can print faster.
What gives this an advantage over a new firmware like Klipper, especially if you are running their experimental s-curve acceleration branch? It seems to me that a Pi running Octoprint and Klipper would be easier and better suited for this.
But Ulendo does seem useful for those who might not want to tinker constantly with their machines.
Hi @Lbibass! Ulendo is a bit different from firmware like Klipper. The main difference is that, in addition to improvements to the motion generation like s-curve acceleration, Ulendo also includes a calibration map of the printer so we know exactly how the printer is going to behave during the motion. Using this, we can alter the trajectory to maintain high velocity and acceleration motion throughout the print rather than showing down at points to avoid vibration. So, adding vibration compensation allows for much higher speeds.
If you’re interested, you can learn more about how we do this in the white paper (link above in this thread).
And, as a PhD student, I would be remiss if I didn’t point you to a paper from some colleagues in my lab, where they compare s-curve to s-curve + vibration compensation and show huge performance improvements: https://www.mdpi.com/2411-5134/3/3/56
We are on Kickstarter to convince our stakeholders of our vision to bring this software to the broader 3D printing community, and open-source as much of it as legally possible. Your support of our campaign will help us make our case to bring it to the community.
Some people have asked about our SaaS model for Ulendo. Please take a look at the answers we gave in the FAQ on Kickstarter to learn more. We are open to other alternatives, as long as they allow us to bring our software in a way that benefits the broader 3D printing community. Let us know your thoughts and we will be happy to adjust accordingly.
This technique uses printer-specific calibrations to apply feed-forward controls to reduce resonance/vibrations, thus improving print quality at all speeds. I don’t think it’s substantially different than the vibration compensation that Ulendo proposes.
The Klipper algorithms don’t run on RAMBo, but they run on a $35 Raspberry Pi alongside OctoPrint.
