I think laser sintered aluminum with machined over pressfits and bores might be cheaper and it reduces the overall time of machining the part with all its complex surfaces. It's somewhat weaker but not so much that breaks easily.
There is castable resin out there. It shouldn’t be to complicated to find someone with a furnace and all the other equipment to cast aluminum. That would be awesome.
Might be tricky but with a good silicone mould, some high strength resin, a vacuum chamber and carbon matrixing, you could possibly get a very strong but lightweight hobby composite.
when i was a kid i used to study my hands to see just how they worked and what all the muscles did. this project seems like alot of fun and its cool to look at.
Your 3D modelling skills are legit the best I've ever seen. I can't even imagine how many planes you must have, and how long you spend to make each component.
@@johndoesson yes, sketching planes. basically to create a 3D model you first draw a 2D sketch on a plane, and then extrude/revolve/cut it in three dimensions.
@@oraziovescovi1922 "..how many planes you must have.." does that mean you make templates of 2D planes to save like you save functions in a library when programming, to make it more reusable or was that a typo?
@@johndoesson usually you create planes where you need them. You can reuse the same plane for two or more sketches, but it is really unlikely that they are perfectly aligned as you need them. in such a complex model there are probability lots of planes since the shapes are so complex.
this being said, you usually want to model a small part or subassembly in a different file, and then assemble everything as parts (wich is just the shape as it is, imported without sketches, you can't really modify it anymore) in another complete file. edit: that's what he did with the different custom servos and wrist mechanism:he modelled them in a separate file and then assembled it in a complete model file
I like where this is going, the fact that you are looking for a more organic hand is refreshing, I mean why reinvent the wheel? Congrats for your progress so far!
Making a replica of a human body makes one truly realize how well engineered the human body is... Makes one appreciate for what they've been blessed....
Humans are great from the ribs up, but if you're telling me our backs and feet are testaments to our grand design you're going to be very surprised to learn how poorly these structures treat us. We can throw stuff really well though which is exceptional. Also the whole "be born early for big brain" thing, that was a decent move.
@@SeveralGhost you could argue that being bipedal was a mistake, but considering how rare a trait it is, the fact that we can do it as successfully as we do is a remarkable feat of engineering. Could we be better from the waist down? Probably, but we don't have a good model of what that would look like, except for more humans.
I honestly cannot understand how this channel has as few subs as it does. Will what you're doing here it amazing and I wont lie, I kinda want to build one to have on my desk like a terminator arm. I wish more people were subbed to this.
I’m so out of my depth watching your videos but the are mesmerising. I genuinely can’t wait for each update. Get excited when you post. It’s so fascinating to watch
Not only do I love your videos, and think you're a genius.. but now I imagine listening to them in a big boomy god like voice imagining a god like creature doing a youtube about designing humans, and it makes them twice as good.
Concerning the long screws it might help to consider bicycle spokes as a replacement. You can buy a tool to roll on the threads for very little money to make any length from normal spokes and set them in place with spoke nipples. There are spokes with straight heads, too. I've done this before and it works
Can't wait when artificial muscles are compact, cheap, powerful and reliable enough for such projects. I bet they would save you so much headache, and also make fingers more springy (which on pair with skin softness is quite important for grabbing stuff gently).
Any chance of a 3D modeling video or timelapse? Because I'm still trying to wrap my head around how you design complex shapes and mechanisms on CAD.. haha
For the really long screws, you might look into using threaded standoffs to span the gap and then a short screw in each end, opposing each other. If you used a ball-head allen wrench or torx, you could even tighten it down with the wrench off-axis. A thought. Amazing work, really enjoy watching the progress on this project.
You need to partner with a machinist with a desktop metal milling machine, preferably in your own town. Or make a milling machine yourself. There are plenty of RUclips videos on how to make one. I'd recommend a fixed gantry epoxy granite design.
Hi will! The build on this project is super interesting, but I am wondering if you can show more of the process of designing those amazingly complex parts that make up the 3D printed parts My background is in SolidWorks, and it seems like you use fusion 360 so I am very interested to see how you design your parts Thanks! Keep making amazing projects and videos!
Hi man! I love your designs. It always looks so intrincated and complicated but beautiful at the same time. Thanks for sharing this prototype and well done! I´m totally inspired by this kind of projects :)
Have you considered more compliant methods of joining everything together? Real hands aren't held together by anything as rigid as a steel bolt. Connective tissues like tendons and cartilage have a bit of compliance. Then, of course; everything is wrapped in skin. To simulate tendons, you could make attachment points that have short, strong springs attached to them, then pairs of springs can be connected with a short cable like lock wire, so it's like using lock wire to bind everything together, but there are springs at either end to give it a bit of compliance. You could also make a glove that fits over the hand and anchors to various points on the skeleton to keep things more-or-less in the right place relative to each other.
Will, have you considered adding a barrel adjuster to your servo connections, to allow you to adjust the tension on each cable ? Although standard barrel adjusters may be too bulky for your narrow ptfe sleeves, you might find that a vented screw in a socket head form is just the ticket.
In the note of cnc machine, you can get an okay mini one off of amazon for a few hundred usd. I have one and while it can mill steel it is high rpm but really slow feed rate and a decent amount of coolant. On the note of cable tension, what about a mini double eye bolt to tighten the tension? Like how engineers use then to tighten cables for cross support for buildings? That or use a self tightening slip knot.
This....is....aaaamazing! you are making strides in this! Oh and to see if this can get printed in metal or aluminium in the future or some kind of composite material.
Good job! Keep up the good work! And if you can, i challenge you to make a bionic hand that is so powerful and so strong, you can use it in arm wrestling competition!
As expected and with no surprise, here‘s the request to share your design. I‘ve been trying to wrap my head around the DLR wrist design but you obviously figured it out. Can you share pleaaase?
Hi, your work is amazing. I'm developing a humanoid robot based on real anatomy concepts in order to achieve all human body's DOFs and Neural Network based AI. I'm developing an hand too, very similar to yours in concept, but totally different on the engineering side due to my choice of using air muscles as actuators instead of servos, in order to give an ulterior "organic look" to the robot (the hand body volumes, as well as the rest of body ones, are given by the same muscles that accomplice the motion, exactly like it would be in a real human body/hand, which would feel less or more stiffier depending on the muscles that are in traction resulting in a more realistic feel at the touch e.g. a realistic feeling handshake). The only part where servos are needed is in the eye motion mechanism where I'm using mg90s servos due to a lack of space. Given the very little space I'm working with I need to integrate the servo's gears and motor in my custom parts, in order to re-arrange them as needed and eliminate the useless stock shell volume. I was wondering if you could share the servo housing dimension, distance between the gear's axys, would be enough. I'm working with an FDM printer, and I'm really struggling to achieve the correct tolerances and dimensions. I'm waiting for my Anycubic Photon Mono SE to ship, and I'm sure that the extra precision and fidelity of resin print will help a lot, but meanwhile I don't want to stop working on my project, and I could really use an hand. :)
Could i suggest adding an insex screw with a drilled hike thru to use in the white plastic thing on the servo to be able to tighten the wire. Like the handbreak wire on a bicycle.
could you design some sort of small joint lock that allows the joints to lock into position and withstand more load than the string could, for actions such as carrying a heavy bag? Some sort of clutch maybe
I was wanting to build a hand. Not biomimetic just robotic but that research led me to your videos and they are amazing! I was planning on using a VR hand tracker to try to control my hand. It would be cool if you have that a try. No point in doing all this work for biomimecry if you can't move it in a natural way.
@@TheModularian yes but it would be cool to see one of these hands controlled essentially "hands free" no wires or gloves. It would also feel more natural to move your hand without 3 pounds of wires and sensors
You work like an animatronic artist in FX industry, the perfect combination for obtain a prosthetic hand that could to pass like a real hand, including the silicon cover... Does it the objective?
Could something along the lines of a wooden platform with spots to anchor actuation motors, with soldering style "helping hands" and other mounting positions be a possible testing platform?
Did you see my comment about using compliant vectoring mount as a wrist? Although, looking back, it seems it might be a little large to be used on your hand.
Great work and i learn real much from your designing. STupid Question in Bad english, would it help when you use in the Actual Movement of the wrist springs to get some weight / tork from the pulley and get so faster movement and a little more weight that the hand can handle?
Still a ways to go. Hard part will be having it build up strength. You can only go so far with servos. You could use your design for a larger model. Like an exo suit or a mech. Nice stepping stone though. It is progress. Keep it up.
Also could make it out of carbon fiber tow in 3D printed molds, perhaps with some 3D printed bits incorporated into it. That would be vastly much cheaper, and about as good as you can get in terms of stiffness.
hey man, congratulations, excellent job you have achieved with those designs and structures, what program do you use for 3d modeling? and how will you control the large number of actuators that will generate so many degrees of freedom?
I wonder if there are devices that can transform electrical current into a linear motion similar to an actual muscle instead of a servo, which is bulky and relies on turning gears.
@@בןניומן Aren't pistons usually powered by hydraulics or have a small motor that rotates the shaft? I was thinking more in terms of muscle fibers that can contract when applying a current. I could be talking complete crap hare. This is completely outside my knowledge base
What is that drafting software... I thought it was solidworks, but the way he was manipulating everything makes me think it's not at all solidworks. Is that a Blendr like software?
This channel is criminally underrated. This is incredible
but nevertheless it grew a lot in the last few months
Weeeeeellll..Sorry I was sick I could not see the video.
Given the things I search up on here I don't know how these videos have just started coming up. Sums up RUclips tbh
😎👍
Truly
Just looking from the intro, that is one slick assembly. Imagine making the final out of cast/machined aluminum.
I think laser sintered aluminum with machined over pressfits and bores might be cheaper and it reduces the overall time of machining the part with all its complex surfaces. It's somewhat weaker but not so much that breaks easily.
Injection moulded carbon fiber with titanium supports, would be best. All Al, combined with motors and batteries, too heavy for a daily use.
There is castable resin out there. It shouldn’t be to complicated to find someone with a furnace and all the other equipment to cast aluminum. That would be awesome.
You can also make biocompatible 3D printed parts with SLS, titanium, or chromium cobalt
Might be tricky but with a good silicone mould, some high strength resin, a vacuum chamber and carbon matrixing, you could possibly get a very strong but lightweight hobby composite.
when i was a kid i used to study my hands to see just how they worked and what all the muscles did.
this project seems like alot of fun and its cool to look at.
Same! I even did drawings of robotic hands to try and figure it all out
Wow, all of that organic mobility with only two external servos so far pulling the tendons! Really amazing work, keep it up :)
honestly your cad drawings just look like art, plus there functional too, well done man, well done
If i ever lose an arm/hand, im going to this dude first
Same here. I joined Patreon so I could contribute towards it in a very small way.
Your 3D modelling skills are legit the best I've ever seen. I can't even imagine how many planes you must have, and how long you spend to make each component.
planes?
@@johndoesson yes, sketching planes.
basically to create a 3D model you first draw a 2D sketch on a plane, and then extrude/revolve/cut it in three dimensions.
@@oraziovescovi1922 "..how many planes you must have.." does that mean you make templates of 2D planes to save like you save functions in a library when programming, to make it more reusable or was that a typo?
@@johndoesson usually you create planes where you need them. You can reuse the same plane for two or more sketches, but it is really unlikely that they are perfectly aligned as you need them. in such a complex model there are probability lots of planes since the shapes are so complex.
this being said, you usually want to model a small part or subassembly in a different file, and then assemble everything as parts (wich is just the shape as it is, imported without sketches, you can't really modify it anymore) in another complete file.
edit: that's what he did with the different custom servos and wrist mechanism:he modelled them in a separate file and then assembled it in a complete model file
I like where this is going, the fact that you are looking for a more organic hand is refreshing, I mean why reinvent the wheel?
Congrats for your progress so far!
Making a replica of a human body makes one truly realize how well engineered the human body is...
Makes one appreciate for what they've been blessed....
Human body is awfully engineered. But that's okay, once our tools are developed enough, we'll make something even better.
@@blinded6502 im guessing you're deformed?
@@outofcontext728 Are you?
Humans are great from the ribs up, but if you're telling me our backs and feet are testaments to our grand design you're going to be very surprised to learn how poorly these structures treat us.
We can throw stuff really well though which is exceptional. Also the whole "be born early for big brain" thing, that was a decent move.
@@SeveralGhost you could argue that being bipedal was a mistake, but considering how rare a trait it is, the fact that we can do it as successfully as we do is a remarkable feat of engineering. Could we be better from the waist down? Probably, but we don't have a good model of what that would look like, except for more humans.
I honestly cannot understand how this channel has as few subs as it does. Will what you're doing here it amazing and I wont lie, I kinda want to build one to have on my desk like a terminator arm. I wish more people were subbed to this.
there is something relaxing about watching the steps of the production process of building a hand.
I'm in awe of the beautiful complexity! You must have learnt so much during this project beyond the many details that you've shared.
I've been building mechanical things and designing CAD for years and this still totally blows me away. This guy is actually a genius.
This is fucking amazing I’m losing my shit at how calm you are about this achievement
This should be the highlight of RUclips 2020 in robotics... great work man!! Amazing!
If anyone deserves a Carbon M2 3D printer, it's this guy.
I’m so out of my depth watching your videos but the are mesmerising. I genuinely can’t wait for each update. Get excited when you post. It’s so fascinating to watch
You deserve more patreon backers!! Someone get this man a CNC machine!
This is one of the coolest series of videos I've been keeping up with. The movement of the palm/wrist looks very natural so far. Love it!
this is definitely the best and most organic looking prosthetic hand i have ever seen, i wait for this series ever week
commenting with "amazing work" "incredible content" just to make the system recommend your channel/work to more people!
Not only do I love your videos, and think you're a genius.. but now I imagine listening to them in a big boomy god like voice imagining a god like creature doing a youtube about designing humans, and it makes them twice as good.
You are a serious inspiration to my son who’s 12 and into robotics and ai. Thank you for your videos. Cheers from Vancouver!!!
Concerning the long screws it might help to consider bicycle spokes as a replacement. You can buy a tool to roll on the threads for very little money to make any length from normal spokes and set them in place with spoke nipples. There are spokes with straight heads, too. I've done this before and it works
This is easily my favorite channel
That looks wonderfully elegant. Awesome.
Such an inspiration and motivation for a personal project. Thank you so much for your work
Dang. The shapes of those linkages may be mainly to prevent collisions, but they look SICK.
nice editing and voiceover. learning from you how to do awesome devlogs on robots before I actually start mine.
thanks
Nice work!
Can't wait when artificial muscles are compact, cheap, powerful and reliable enough for such projects. I bet they would save you so much headache, and also make fingers more springy (which on pair with skin softness is quite important for grabbing stuff gently).
Love this so much
I like to think this sort of design is what will get us to robotic prosthetics that can actually compete with flesh and bone
One day we'll have synthetic flesh and bone. It'll make hand designs simpler and yet also gentler.
Any chance of a 3D modeling video or timelapse? Because I'm still trying to wrap my head around how you design complex shapes and mechanisms on CAD.. haha
I second that.
It’s unreal what you are doing here. Love it. If I had the money to support you I would
this channel is so underrated
For the really long screws, you might look into using threaded standoffs to span the gap and then a short screw in each end, opposing each other. If you used a ball-head allen wrench or torx, you could even tighten it down with the wrench off-axis. A thought. Amazing work, really enjoy watching the progress on this project.
Everything looks amazing your channel is an insperation to all thank you Will Cogley.
I've been watching you since the beginning, Will. I just keep seeing things that impress me. You're quite the engineer!
You need to partner with a machinist with a desktop metal milling machine, preferably in your own town. Or make a milling machine yourself. There are plenty of RUclips videos on how to make one. I'd recommend a fixed gantry epoxy granite design.
You should email him this rather than just the comment section.
Hi will! The build on this project is super interesting, but I am wondering if you can show more of the process of designing those amazingly complex parts that make up the 3D printed parts
My background is in SolidWorks, and it seems like you use fusion 360 so I am very interested to see how you design your parts
Thanks! Keep making amazing projects and videos!
I second this motion!
I thought even getting to this point would take a lot longer! Looks great!!
Earned yourself a sub, friend! I’m more an electrical guy, but this is some really interesting successive mechanical design.
Idk why you're doing this but it's impressive
This is some slick engineering!
I just amazed with every new video! Big thank you, good luck and let's hope that this concept will be very useful.
Impressive design as always. Thanks for keeping us updated.
Your machines are beautiful.
outstanding as always.
Really looking forward to the full prototype
I love the research you've done on this
Hi man! I love your designs. It always looks so intrincated and complicated but beautiful at the same time. Thanks for sharing this prototype and well done! I´m totally inspired by this kind of projects :)
Amazing work and faszinating project ! Just keep developing, this is awesome !
Have you considered more compliant methods of joining everything together? Real hands aren't held together by anything as rigid as a steel bolt. Connective tissues like tendons and cartilage have a bit of compliance. Then, of course; everything is wrapped in skin. To simulate tendons, you could make attachment points that have short, strong springs attached to them, then pairs of springs can be connected with a short cable like lock wire, so it's like using lock wire to bind everything together, but there are springs at either end to give it a bit of compliance.
You could also make a glove that fits over the hand and anchors to various points on the skeleton to keep things more-or-less in the right place relative to each other.
thats sick! Keep going dude.
I've been amazed by your videos for quite a while and would love to see more.
Will, have you considered adding a barrel adjuster to your servo connections, to allow you to adjust the tension on each cable ? Although standard barrel adjusters may be too bulky for your narrow ptfe sleeves, you might find that a vented screw in a socket head form is just the ticket.
Great Job Will
In the note of cnc machine, you can get an okay mini one off of amazon for a few hundred usd. I have one and while it can mill steel it is high rpm but really slow feed rate and a decent amount of coolant.
On the note of cable tension, what about a mini double eye bolt to tighten the tension? Like how engineers use then to tighten cables for cross support for buildings? That or use a self tightening slip knot.
Incredible work.
Как всегда - потрясающе. Нужно собирать свой чпу, и делать рычаги из дюраля)
You doing good.you inspired me lots. I'm mechatronics student your video help me lots.
If you will use dynamixels or similar smart servo, it has current sensor. Current propotional to torque, so additional sensor may be not necessary.
This....is....aaaamazing! you are making strides in this! Oh and to see if this can get printed in metal or aluminium in the future or some kind of composite material.
lookin good, amazing content as always, and I guess the hand is ok too.
Incredible work! Please add 'mantis blades' to the to do list after this current design goes alpha
Good job! Keep up the good work! And if you can, i challenge you to make a bionic hand that is so powerful and so strong, you can use it in arm wrestling competition!
What CAD software is being used here? Really cool to see the parts all linked up and moving together like that in the software
Yeah I'd like to know as well
As expected and with no surprise, here‘s the request to share your design. I‘ve been trying to wrap my head around the DLR wrist design but you obviously figured it out. Can you share pleaaase?
Interesting! I really look forward to continuing!
Hi, your work is amazing.
I'm developing a humanoid robot based on real anatomy concepts in order to achieve all human body's DOFs and Neural Network based AI.
I'm developing an hand too, very similar to yours in concept, but totally different on the engineering side due to my choice of using air muscles as actuators instead of servos, in order to give an ulterior "organic look" to the robot (the hand body volumes, as well as the rest of body ones, are given by the same muscles that accomplice the motion, exactly like it would be in a real human body/hand, which would feel less or more stiffier depending on the muscles that are in traction resulting in a more realistic feel at the touch e.g. a realistic feeling handshake).
The only part where servos are needed is in the eye motion mechanism where I'm using mg90s servos due to a lack of space.
Given the very little space I'm working with I need to integrate the servo's gears and motor in my custom parts, in order to re-arrange them as needed and eliminate the useless stock shell volume.
I was wondering if you could share the servo housing dimension, distance between the gear's axys, would be enough.
I'm working with an FDM printer, and I'm really struggling to achieve the correct tolerances and dimensions.
I'm waiting for my Anycubic Photon Mono SE to ship, and I'm sure that the extra precision and fidelity of resin print will help a lot, but meanwhile I don't want to stop working on my project, and I could really use an hand.
:)
Looking great mate
Someone get this guy on Tested.
Node channel sounds kinda like the same guy...🔥🔥🔥💯
Could i suggest adding an insex screw with a drilled hike thru to use in the white plastic thing on the servo to be able to tighten the wire. Like the handbreak wire on a bicycle.
could you design some sort of small joint lock that allows the joints to lock into position and withstand more load than the string could, for actions such as carrying a heavy bag? Some sort of clutch maybe
Awesome work!
best concent in youtube
I was wanting to build a hand. Not biomimetic just robotic but that research led me to your videos and they are amazing! I was planning on using a VR hand tracker to try to control my hand. It would be cool if you have that a try. No point in doing all this work for biomimecry if you can't move it in a natural way.
His controller glove is chocked full of potentiometers, the end result will move as his hand moves
@@TheModularian yes but it would be cool to see one of these hands controlled essentially "hands free" no wires or gloves. It would also feel more natural to move your hand without 3 pounds of wires and sensors
This is incredible ! Thank you for your work !
I pushed the bell but did not see an update:( keep up the great work
You work like an animatronic artist in FX industry, the perfect combination for obtain a prosthetic hand that could to pass like a real hand, including the silicon cover... Does it the objective?
Could something along the lines of a wooden platform with spots to anchor actuation motors, with soldering style "helping hands" and other mounting positions be a possible testing platform?
Once you're finished, you should wear glove over it in order to emulate skin. It would look awesome (if it fits, that is).
Amazing work keep it up
Did you see my comment about using compliant vectoring mount as a wrist? Although, looking back, it seems it might be a little large to be used on your hand.
It's alive 😱
5:00
Maybe a parallel/delta type deal there? Do they even make em that small? 😂
Btw this robot hand project blows my mind. Nice!
You're inspiring me!
Again, just an amazing job!
Great work and i learn real much from your designing.
STupid Question in Bad english, would it help when you use in the Actual Movement of the wrist springs to get some weight / tork from the pulley and get so faster movement and a little more weight that the hand can handle?
Cool job!
Still a ways to go. Hard part will be having it build up strength. You can only go so far with servos. You could use your design for a larger model. Like an exo suit or a mech. Nice stepping stone though. It is progress. Keep it up.
1:06 you should have that part 3d printed in metal or cnc via a 3rd party . i feel it is a superior design
Also could make it out of carbon fiber tow in 3D printed molds, perhaps with some 3D printed bits incorporated into it. That would be vastly much cheaper, and about as good as you can get in terms of stiffness.
Great video as always 🔥
hey man, congratulations, excellent job you have achieved with those designs and structures, what program do you use for 3d modeling? and how will you control the large number of actuators that will generate so many degrees of freedom?
I wonder if there are devices that can transform electrical current into a linear motion similar to an actual muscle instead of a servo, which is bulky and relies on turning gears.
Are you thinking of something like a piston? It just extends and contracts?
@@בןניומן Aren't pistons usually powered by hydraulics or have a small motor that rotates the shaft? I was thinking more in terms of muscle fibers that can contract when applying a current. I could be talking complete crap hare. This is completely outside my knowledge base
That's so cool!
Brilliant...
What is that drafting software...
I thought it was solidworks, but the way he was manipulating everything makes me think it's not at all solidworks.
Is that a Blendr like software?
Same question here; hope we get an answer
@@benschritchfield8721 many thanks mate. I currently use SW because my uni has the license but I'll definitely try it
amazing!
I can't do much but I did drop a subscribe.