Mmmh,
Since InMoov hand has been released on Thingiverse in 2012, I received demands on how to use it as a prosthetic. I was always a bit skeptical about it.  I knew the initial design was not really adapted for that matter and I think it could be better done if I kept the prosthetic perspective in mind.
Lately a french FabLab of Rennes asked me if they could use InMoov for their project.

So, I have been working on this for about three full days. I had made a first design very simple to see if it could actually work. I presented it in Rennes for the Bionico project. The idea is to get a finger that gets actuated mechanically instead of using braid lines to pull the fingers.

It may look similar, but I tell you, it isn’t. This little baby really took me a hell of a time to design. When things are small, you need to take a lot in consideration with our 3D printers. First, all parts should be printable without or almost without support, the thickness of the walls have to be thick enough to be printed with a reasonable definition. Parts should be easy to assemble. And last, parts shouldn’t require too much quality from our printers…

 

I know this is not the legs…  Many of you are waiting to see what’s next to download for your robot. But this is important, if it can help disabled people, it is urgent.
You can download this prototype here. Remember this hasn’t been printed yet and may have problems that I still need to resolve. It is uploaded for the FabLab of Rennes to check movements and frictions under Freecad.
This could, if it has enough strength under load, replace the hands of InMoov later on. It is meant to work with a Corona DS238HV servo.

Also Leonardo and I have been setting up a website for InMoov, it isn’t ready yet, but things are coming. Leonardo has been working hard on it, and I have been asking for to modify things, so poor Leonardo is under load.
Also I want to thanks the donators, for their generosity. It will help us to keep the site up.

Jha has made a new script for mouth movement and Grog has been working to add it in the InMoov service. I still need to test that to see how that works, but if it’s good, InMoov should be able to move his jaw everytime he says something. Cool!

The legs… That’s what I’m going to focuse on now. Remember I had made the foot. This foot needs work and I need to redesign the ankle because I want to add a servo to get more DOF. So that’s the next goal. I have worked on the Knee but it is not clear yet how I want to actuate it. So much work for so little time.

 

Comments 32

  1. Brilliant, Gael! I’ve been working on the hand from a very different direction, which is why you’ve not heard from me in some time. I wish I could design things half as quickly as you seem to be able to manage. 🙂

    • Quickly!?
      Gee it’s already too much time. When I say three days, I’m actually lying to myself, I have it in my head since 3 weeks and turning it in all directions before putting it down on the screen.
      I would love to see what you are up to, is it on your blog? I will have a look now.

    • Quickly compared to me. 🙁

      I was working on the Inmoov forearm and realised that there were only so many servos we could put in there. Here is the headache I’ve made for myself. There are thirty five muscles controlling the hand that reside either in the hand or in the forearm controlling the hand via tendons. We can slap servos on the back of the hand like you’ve done but then you wind up with a rather thick hand. As well, there are two tendons for each finger on the palm side of the hand and one tendon on the back side. I’ve been trying to figure out how to design servos that are more like muscles that what we’re using now. It has been difficult. 🙁

  2. I think I will print this tomorrow. I ordered 5 Corona servos from HobbyKing. I have not yet built the left hand and arm for my InMoov yet so I may try to build one of these hands.

    • Wow, beware though this is totally a work in progress and there is certainly things that are wrong. The fingers cannot be attached as a hand yet, and also the thumb doesn’t exist.
      Please let me see pictures once printed, you will be the first beta tester.
      Thank you for your donation!

    • Thanks for posting the pics that fast!
      Yes the wormgear is very small, but since this is for a prosthetic, I had in mind it could printed in stainless steel to last a good while. Even if it brings the cost a little bit up, it would be nothing compared to a 25 000 dollars prosthetic.
      Also the actuators, being very small and fragile, could be cut out on a CNC or laser cut from PCB FR4 or stainless steel printed. What we are trying to do with this project, is to get a prosthetic for less than a 1000 euros.
      We are far from reaching that amount…
      Have you been able to put it together?
      Gee I wish I had my printer to test myself.

    • Oops. I posted this reply in the wrong spot below.

      I just posted some pics on my blog of the finger assembly after correcting the wrong installation of the first actuator. I had to use my knife a bit on the servo holder to make clearance for the gears.

    • There is a new post on my blog about the prothetic finger test. The finger tip still doesn’t flex but it might be a mistake on my part again.

    • Nice concept. I printed the finger and it fits perfectly.

      There are two problems however:

      1. The tip doesn’t move. I think the actuator applies force too close to the pivot of the last phalanx and the end of the actuator flexes and the hole gets distorted and that’s where the problem is.
      2. The phalanges don’t make a 90° angle.

      Thanks for publishing the files!

    • Hi, thanks for your comment.
      Did you print the file “PhalangeV1” or “PhalangeV3”. Because this was a problem with the initial version. I’m waiting for results with the V3 version, with tests done by Morris.

    • I printed it today with ABS, using the latest download from thingiverse. It was the PhalangeV3.stl. With the use, the end hole of the actuator has widen and the play is significant. But the finger tip did not move in from the beginning. I could of course print another actuator, but it wouldn’t help much.

      I think the forces along the second actuator are greater than it can withstand without deformation of the plastic.

    • I suspect that the finger tip would work fine, if actuator #2 mounted exactly like actuator # 1, from the top of the previous phalanx to the bottom of the next, i.e. the tip.

      Also the pivot of actuator #1, on the servo bracket, could be located more towards the end of the part, instead of beneath the pivot of the finger, to avoid the conflict that stops it from reaching a 90° angle.

      Just my 2c (the anonymous above, forgot my name).

    • Hi, yes I had a previous design that was done like your suggestion. The actuator looked similar to the first actuator, and it should work obviously. But I wasn’t happy about the exterior design of the finger. The pin situated on the first phalanger that holds the second actuator is placed so close of the edge that most of our printers wouldn’t print that thin piece, so I created an additive part to make sure it could be printed, and was not so happy about the final look.
      That is the main reason I uploaded that version, hoping it would act pretty much the same.
      Anyway I guess I will put that for downloading until I can test with my own printer to see the results.

    • Hi Brian,
      Your comment made me look again at the mechanism of Meet Roy. I think he has a very good system, I totally forgot about it. I like his simple view of principle. I was trying to built something like my first Type Writer Robot hand. http://www.thingiverse.com/thing:8659 and focused on how to lock the finger if power is off. Using wormgear like in the InMoov shoulder gives that option.

  3. I think you are wonderful !!! You are a generous and very skillful man ….. So many many hours of your time for to help people and many many more hours of your time just for people to have fun to build their own robot.
    Thank You for everything

  4. Been away for a couple months and the changes to InMoov are moving along nicely. Gael I love the work you are doing.
    It’s going to take me another month to get up to speed. Lol
    Well once I get off of vacation I will start back up with my baby and see if I can get him to start talking and moving his arms. Can’t wait to finally start working on the legs and have a real life humanoid standing in my office. My students love this project.

    Thanks
    Hans Foerster
    Fleming College

    • Hi Hans,
      It nice to hear from you.
      I also can’t wait to get InMoov on legs! I doubt it is doable for the makerfaire in Roma in October… No way… too short time.

  5. Hi Gael,

    How can I get in touch with you privately (through email maybe)?

    I would like to help you out somehow (sponsoring), but instead of making empty promises publicly, I would like to speak with you through email first, what is on my mind.

    Could you please shoot me an email to: inmoov (at] arcol.hu ?

    Best,
    Laszlo
    http://arcol.hu

    • Hi Lazlo,
      Thanks for your interest in InMoov.
      The email you gave doesn’t seem to work.
      I looked on your site, but can’t find an email.
      You can contact me with this: hairygael(at)gmail.com
      Regards,
      Gael

  6. inmoov will be featured on Stratasys blog. They have seen our inMoov printed on our Stratasys uPrint. I will let you know, when they publish a post 🙂

Comments are closed.