French model maker and sculptor Gael Langevin spoke to us about how he created InMoov – the life size 3D printable robot which has attracted huge interest from roboticists since its birth in 2012.
1. What is InMoov?
InMoov, is the first life size humanoid robot you can 3D print and animate. You have a 3D printer, some building skills, you can start this project. InMoov is also replicable on any home 3D printer with a 12cm3 area, it is conceived as a development platform for Universities, Laboratories, Hobbyist, but first of all for Makers.
It’s concept, based on sharing and community, gives him the honor to be reproduced for countless projects through out the world. Today the program MyRobotLab is commonly used in InMoov.
It can be controlled by voice, it has face and object detection and tracking, it has a Kinect that lets you create gestures for the robot by simply moving your own body. It is implemented with a chatbot, like Siri, that lets you converse and give orders. It has currently 28 servos which give a total of 50 DOF. Ten finger sensors, and PIR motion sensors. Two Arduino microcontrollers. But MyRobotLab will let you do a lot more, it is the reason I had chosen that Open software.
2. Tell us a bit about your background as a sculptor. What sort of things have you designed in your career?
I am a sculptor designer, I work for big brands since a little over 25 years creating objects for them. It is interesting because it varies a lot, I almost never do the same thing twice. It goes from sculpting a perfume cap to making some kind of special effects with textures for a TV commercial. But what I like the best is sculpting.
3. How did InMoov come about? What’s its story?
It all started after buying a FDM 3D printer for my work. As a sculptor, designer, I always need to create objects for my customers. I already had a CNC machine, and getting a 3D printer seemed to be worth to try. My wife, wasn’t very favourably inclined to this machine, she was telling me that I could only produce little rabbits or Yoda figurines. I had in mind that it could be used in a more practical and engineering way. I wanted to use the printer beyond it’s physical limits. A French vehicle constructor company asked me to make a bid to create a futuristic prosthetic. The job didn’t happen, but I was thinking of using my 3D printer to create it.
I decided to design the prosthetic hand anyway on my free time. I always liked hands, my work shop is full with hands, made in all kinds of materials, some were moulded with plaster on my own hands when I was 12 years old. After designing and printing the first InMoov hand, which was a non-animatronic object, I decided to post it on Thingiverse under a CC-BY-NC Licence. Sharing the parts with the Open Source community was a logical route for me. I have been using Linux for years, Blender as a 3D software as well, and my 3D printer was built following an open source project.
Photo: Quai Lab
InMoov hand sparked the early pioneers in the 3D printed open prosthetics movement. It provided evidence to what would become the RoboHands and eNABLE communities showing that 3D printed limbs are mechanically possible on a low budget. I personally actively participate in the open prosthetics movement and particularly, in the Bionico project.
4. What software do you use to design InMoov?
As you mention I use Blender, which is a Open Source software for 3D designing but not only. This software is so complex that we recently are using it to rig InMoov to use it as post movement learning and study.
I use sometimes Netfabb for to fix the parts before printing them. MyRobotLab is the software to control InMoov.
5. Why did you decide to make InMoov open source opposed to patenting and selling it?
I have been using Linux since a while now. Blender the software I use for designing in 3D is huge, with incredible resources. When I bought my 3D printer which came directly from the Open Source community, I wanted to give back. I believe we need to share things at your own level. Internet is in itself such a rich place for information and communication which wouldn’t be possible if people wouldn’t share. Lots of people take it for granted, because it seems like a normal thing today.
6. You have a video demoing the Myo Armband with an InMoov finger. Have you trialled any other third party control devices? Are any of them proving to be more or less effective than others?
But the Myo armband is very responsive and easy to use. Once configured, you can remove it and put it back just like a bracelet and it will be set back for use right away.
7. Have you seen any interesting uses of InMoov?
InMoov is used for SO many projects that it hard to discriminate. I have seen thousands of InMoov hands printed through out the world which have a potential for innovation.
Recently Nicolas Huchet with InMoov prosthetic hand won a MIT prize.
The Parloma Project used the InMoov hand for to create a hand that would be able to do sign language for def people learning.
There is the Kaist project (South Korea’s Advanced Institute of Science and Technology) uses the InMoov robot for research to pilot planes. The Kaist group won the Darpa contest this year.
Peter Heim has put InMoov on ROHS software to let him drive around autonomously.
Alessandro Didonna in Italy worked in collaboration with Greg Perry and Kevin Watters to set the Oculus Rift to control InMoov has telepresence device.
There is two legs projects that I know of, besides my own legs design for InMoov. I need funding and time for to keep going on that field.
The “InMoov Robots for Good” is also an interesting approach of using InMoov for a complex task which will certainly require a lot of research and resources from the community.
8. Do you plan to evolve InMoov or leave it up to the community to adapt it? If the former what do you intend to change?
Of course, I plan to evolve InMoov, but I need fundings to get going with this project, because at this stage, I cannot support it anymore. It is too wide and no matter what effort I do for to respond emails, videos, maintain the site, the InMoov shop, conferences, interviews, I’m running out of time. I need to set an economical model to overcome this lack of time and to hire help.
The community is also forking InMoov in some ways and that is super interesting.
My InMoov robot is the way I have designed it. I did not integrate any derivatives however on my website there is a link that shows all the different derivatives.