Ivan Miranda
patreon
GIANT ROBOTIC HAND
🕑 Added 2023-08-24 15:41:15 +0000 UTCHi everyone!!
I guess once you have a few giant 3D printers building a giant rideable robot is just one step away, right?
I've been tempted many times before to build a giant robot that I can ride and this could be the first step. A giant robotic hand that I could put on the robot to do robot stuff.
If you think it would be a good idea to start such a ginormous project please let me know down below because it is starting to feel more and more interesting everyday.
This hand is just a base to be improved on. For starters I should swap the belts for kevlar cord or something similar as I don't really need the timing properties of the GT2 belt and it would be more compact, more rigid, cheaper and more reliable.
Another point in my plan is to make al the servos continuous rotation and move the servo potentiometer to the joint. That way it will still behave like a regular servo but I can now trade speed for torque by changing the ratio of the servo pulley to the joint pulley. Right now the joint does 90º for every 180º of the servo and I'd like to go to about 90º per every 720º. It will be slower but way more powerful.
I'm also designing a "glove" that I can wear that will use the servo potentiometers to give "Force Feedback" to the glove so I can feel pushback when the robotic hand is feeling resistance to the movement on every joint. By how hard has been to design the glove I can already tell how ambitious this plan is but I'd love to at least try.
As homework I'd like you all to think about whether I should use the old and reliable tank treads as feet for the robot or if I should go all in and try to make it walk.
As always I love to read your comments so keep'em coming!!
Thanks!!!
Comments
Neil Devonshire
Hi Ivan, I've been moving house so not been very active and just seen your plans, sounds very good. I've designed a similar (much smaller and not red ;-) robotics hand using Kevlar within PTFE tubing (like a Bowden cable setup) and I tested the results, the PTFE tubing holds up really well to wear from the Kevlar (1mm Kevlar breaks around 66kg, depending its weave). I created a test rig to run a servo back and forth over a 1000 times and saw hardly any wear on both Kevlar and PTFE. Something that did come up was the sensing of finger touches as I wanted the hand to be capable of operating under water, I opted for 3 pulleys along the Kevlar in a chicane configuration with the centre one having a pressure sensor/streign guage so that when the Kevlar was tighter from touch you could measure how much pressure was on the joint. I had to take into account the weight of the part past each joint, but could also use this data to work out if the hand was empty, which way round it was, due to gravity and other joints positions, etc. Also, I designed mine for large joints to use multiple servos in series along the Kevlar length, so each servo diverts the Kevlar around another pulley, this way you can have fast speed (all servos moving) and slow speed higher prescision (1 servo at a time moving). For the joints, I milled PTFE material to sit between the joints and have the Kevlar hold the joints together (same as our own hands with tendons), that way it will take any shock, not break the joint, and you can detect the stress on the inline sensor. It's a fun project though, looking forward to seeing what you decide to make as always 👍. I can send pics if ya like
Ed Zaharis
Fantastic, Ivan you never cease to amaze. I am looking forward to see what this turns into.
