Soft robots, the reason I built the Blackbox printer

Three years ago, I started to build the Blackbox 3D Printer. There was one driving factor which kept me going: The dream of making soft robots one day. I assembled pneumatic robots in 2017, and that convinced me that I was on the right track. These robots where impressive but assembly was complicated because of the numerous hard to reach parts and the required precision. That way, we’ll never build robots like the ones we dream of. My idea is to print actuators explicitly designed for materials with significantly different properties.


Finished print on the printbed, sprayed a 50/50 mixture of ethanol and water on it so I can take it off sooner


How it works!

This little actuator was printed in one process and consists of just 3 parts; An actuator, a printed gear and a servo Motor. The working principle is also quite simple, flexible fingers are connected to a slider with bridged “cables,” which is then driven by a servo using a rack and pinion. What’s not so easy is the print itself, there are very few printers out there which can combine flexible materials. My first experiments started with a E3D Kraken hotend and failed because the inactive nozzle permanently dragged across the print. The second experiments where before I installed a wipe system and turned out to be very messy. But the third experiment finally succeed!


Small cable channel in detail


So how to combine flexible and rigid Materials?

Printing two materials next to each other create a fragile bond, especially when it’s such different materials like TPU and PLA. My solution to making a strong connection between the two materials are undercuts. By hooking the materials together, the links can withstand high forces. I’ve tested several hooks, but I haven’t been able to compare them yet.

What’s next?

This is just a proof of concept and a straightforward mechanism. There are so many things to test, but my goal is to print highly complex actuators with multiple motors and axis, but for now, I’m keeping it simple. Another thing I want to test soon is nylon cable bundles for high torque applications. I’m open to any input and will probably create a Hackaday project out of this once the Hackaday prize 2019 is over!




Soft Robot Actuator


1.8 MB 247 downloads


STEP download is available for Patreons!


I’m taking part in this year’s Hackaday Prize with the Blackbox project, any comment, like and share is welcome and very well appreciated!




  • 15/08/2019

    Guy Isely

    Hi Marc!

    I am so excited to see this! I am hoping to start doing multi-material printing for biologically inspired robotics designs, but I am 3D printing novice so it’s great to see someone else is already doing this and succeeding. I’ve already gleaned some interesting info just by reading your post. I didn’t know, for example, that there is weak adhesion between PLA and TPU. I haven’t tried any multi-material printing yet myself as I am still investigating what kind of machine I will need to do that successfully.

    The Blackbox project seems like a cool idea and I’ve gone ahead and supported you on patreon for it. I’ve been thinking a lot lately about how close I can get to robotics design workflow where my designs are iterated almost entirely in software with only minimal hand-assembly and production variability. Obviously, I think we are pretty far from being able to achieve that kind of workflow at the moment, but you Blackbox printer might help. Initially, however, I am just planning to experiment with printing stiff and flexible materials simultaneously. Is there a reason by existing dual extruder printers aren’t up to this task?

    One of my hopes for seeing significantly progress in robotics designs is to use artificial muscles in place of motors. Most existing artificial muscles technologies have some pretty critical issues that make them unlikely to be competitive with motor-based designs, but I recently encountered work from the research group of Christopher Keplinger that seems a lot closer to viable. Their muscles are electrically actuated and hydraulically driven with good speed, strength and durability. They recently published this paper giving a simple explanation of how to construct the muscles for DIY community:
    Currently their process involves melting together two thin sheets of plastic, injecting the space between the sheets with a dielectric fluid, sealing the injection port with soldering iron, and then painting on electrodes with a carbon paint and copper strips. I have been thinking about whether it would be possible to modify to their process to print their muscles on a machine that simultaneous prints other structures like bones, tendons and connective tissue. There are a bunch issues that might make this difficult, but it might be possible. Maybe with the right set of tools heads you Blackbox printer could do this?


  • 15/08/2019

    Wow this is excellent work! You designed a printer so you could experiment with new printing concepts and then made this wicked gripper and open sourced it all. Thank you so much and I can’t wait to see what you get up to after this!

Post a Comment