Fascinating. I have always wondered why the actual human skeleton is not used as the framework for these modern robotic builds, since that is the only way to ensure the exact movements.
complexity and reliability are probably the answer, wear and tear can't be healed only replaced and prosthetics are probably expensive already so lowering the cost of getting one and maintaining it might be a priority over complete functionality
The actual human skeleton by itself is not mechanically stable. it requires the connection of ligaments, bursae, and muscle-tendon units for joints to even be stable enough to be measurably driven. The thing about building biomimetics is you have to strike a design balance between reproducing the biological motion, and building mechanical joints sturdy enough to be driven by modern actuation technology (electric motors, pneumatics, hydraulics, etc.). It is not enough to simply build a miniature saddle joint and attach a string to it to drive motion; the joint itself must be a modern, mechanical joint capable of permitting drivable motion only in the movement planes desired, or at least the joint's end effector must have a larger working volume reducible to the biological analog. Additionally, aside from the issue of scalability, modern electric motors produce forces that are very large, and not easily reduced to give the joint torques required to emulate the fine motor coordination of the human hand (force-torque reduction requires gears/pulley structures, for which there is not much space in the available volume enclosed by the hand). In general, the skeleton's joints exist along a wide spectrum of "stability" vs. "flexibility". The more Stable joints are rendered stable through the action of active musculature. The more Flexible joints require elastic structures such as ligaments and fascia to permit compliant movement in multiple directions other than the powered directions driven by the large muscle-tendon units. Fascia also facilitates pulley/like structures in a low volumetric profile. The carpal and metacarpal joints are closer to the flexibility part of the spectrum, and in vivo are entirely stabilized by non-rigid, viscoelastic anatomical structures; the bones are there merely to pass along compression loads while executing a movement.
Really cool, it's great to see how different it is when a very talented and patient individual makes something, comparing it to the competition of companies, where the main goal isn't just fun and usefulness as much as the profits are.
Ignoring "profits" for a moment, the *cost* of producing a hand like this *is* critically important. High cost prosthesis are great for those who can afford them, but you won't change the world by producing a product that only 1% of a potential market can afford. Even with socialized health care and state-funded prosthesis, that cost is still hugely important, because the budget for prosthesis competes with the budgets of other public goods and services. The challenge of engineering a consumer product is that you have to build as much *quality* in while also trying to minimize costs. Certain design decisions also only work at different scales of production. 3d printing is great for prototypes, but is not economical enough for large scale production. The real take away, however, is that the quality of the engineering determines how much you have to sacrifice between cost and quality in your prototypes and final product. Even with the best engineers working on a project, that engineering takes *time* to complete and iterate on to produce a worthwhile product. And that time is often much more than individual hobbiests can afford to spend. At any rate, it will be interesting to see how this prosthesis will develop, and it's always interesting to see behind the curtain of a complex engineering project like this.
About 4 months ago I tried to catch a football and ended up using my thumb to stop it cause being human is fun. I jammed it so badly that it still hurts today. Seeing these joints and the complexity behind them makes me understand why it hurts so bad! I could have googled this but to be honest ya get over the pain. This video though, wow! Thanks for sharing all of this with us. I want to treat my thumb better now lol.
I'm constantly amazed at how much you're learning and discovering (and I'm learning and discovering, too!) about the hand and its constituent parts, and how your project is developing! Nice one, Will! :)
The work that was involved in this project and video is so impressive and in the future it could become a reality and a benefit to a lot of people thanks 😊
Have you thought about solid state compliant mechanisms? Basically you have a single solid part that is designed to flex in a specific way when forces are applied. Might be useful for the 1st-4th phalanges to simplify construction considering the innate limitations on how flexion occurs away from the palm. Veritasium has a video on those mechanisms with a professor who (quite literally) wrote the book on the topic.
Your information is very comprehensible and knowledge based. I prefer your idea of the gear box solution hopefully you will consider using a clutch assembly or actuator for each finger to engage gears independently and simultaniously. I feel very positive about your work.
for the cmc 1 and 2 joints, I'd recommend some form of elastic connection, as, while the muscles have almost literally no control over how they move, they're still elastic to allow conforming to various objects being gripped. at least allow cmc 1 to be flexed backwards by 9.5 degrees. you can see this elasticity by grabbing cmc 1 and 2 and twisting the grabbed hand. it helps in grabbing straight objects like rods and poles and sticks and such.
this freedom of movement can be seen by simply flattening the hand. if the mcp joints were all lined up in a straight line with only freedom in cmc 3 and 4, then cmc 5 would be lifted off of a flat surface. side note, I number cmc/mcp with index as 1, pinky as 4, thumb as 5.
Hello, Mr.! Maybe it would be useful for you, as sports doctor I really like ‘The Physiology of the joints. Upper limb’ by A.I. Kapandji. Thank you for your great work!
I reaaally hope that you pull this off 🤩🤩 I personally wanted to study bionics but sadly I never really liked math😓 but I love what you are doing. Currently I'm about to start my fisiotherapy studies, do you think that the knowledge that I gain there could be used to one day maybe built a prostetic hand for a pacient??
Don't ever leave this project, think if you can somehow get the feature of feel. So the amputee person can feel through that hand. That is the most inportanat thing. I lost a few fingers and damaged 2 MCP 1 PIP and 1 DIP but have the feel of what i have. You can save lives if you get that to work. Good luck!
Random question, what is the book you have open with the hand diagrams in it? Looking to get some new anatomy books and I'm loving your videos so I thought I'd get a copy. Thank you and I look forward to more content!
This is fascinating stuff! One question that always pops into my brain when watching this sort of stuff - and you'll have to forgive my ignorance - we know how a biological hand works, why don't we just replicate that? I mean, if the designs the same (just made from different materials), then replicating the same range of motion, etc should also be possible, yeah? Obviously, the control system would be different, because the human brain/nervous system is a thing. Like I said, forgive my ignorance on such matters, and keep up the fantastic work!!
I think the reason why is because they can't make the motors small enough without loosing grip strength. Also you'd need to replicate tendons, which cause the movement since they're muscles and contract when u send brain signals. As far as I know they haven't made a synthetic tendon like thing yet, so ud have to sub it with bulkier gears and such.
@@Mysoulismelting I'd imagine that one would place the motors further down the forearm, where there's more space, and have the fingers driven by corded "tendons". IIRC, there has been some steps in the development of synthetic "muscle", but I haven't followed it, I just remember watching a few clips years back.
@@TheScwall i remember then being able to make synthetic cartilage like for ears and noses. But id imagine it'd be bulky if you were to have the motor further down the arm.
Interesting as always and great to hear about the SLA! Could you consider changing the background music? I find the quiet yet very intense drumming a bit too distracting.
Hello. And thanks. My University project is about designing a prosthetic arm and my knowledge in this field is next to non. But I am learning so much from you so I appreciate that. Dw I'll come up with my own design and if possible I'll reference things correctly (I hate writing repots but what can you do?)
If you are having trouble with space and the gearboxes, have a look at a Cycloidal gearbox, it allows for really flat, high reduction, gearboxes. Though there isn't many small ones of these for sale though.
inquiry: why not combine M2 and M3 into one broader bone integral to the carpal section, and combine M4 and M5 into another single broader bone and rely on a combined swivel and hinge or 2-axis saddle articulation to the lateral side of the carpal section?
Would be interesting when you get an sla printer to use some of the available semi flexable resin to print out all the joints, bones and tendons in one print to reduce complexity of mechanical joints, sla would allow for that level of detail in one print, instead of mechanical hinges "living" ones , although this might not provide the framework for control you are trying to make
Is it possible that the thumb's CMC joint's pronation/supination aren't directly active movments, like flexion/extension and abduction/adduction are? As in, the metacarpal twists throughout it's range of movement, rather than being able to twist on demand like other movements can be done? I don't study anatomy, I know this joint's pronation/supination is severely understudied, I can only find one big study that did 4d scans of the joint, though this study only showed that pronation/supination exists but not how those movements are actuated. But I came to this idea based on the fact that you can actuate movements in joints through external force, basically pushing/pulling on a joint can involuntarily actuate it. Except you cannot make your first metacarpal twist with this method, it stays fairly rigid, like the other metacarpals do. Thus I believe this joint's pronation/supination may be passive, maybe not truly passive but also not an active actuation. If this is true, the twist would only be there to allow greater range of movement, which may assist in opposition and reposition movements. Making the joint technically 3dof, but only active in two degrees, with the third being passive between the other two. But again, maybe it's not a true passive, maybe it is partially active due to how the tendons pull the metacarpal, I'm not really sure how to define this if it's not truly passive but also not truly active.
Thank you so much for sharing all those information with us. As I do provide free prosthetics devices in India that information will also be useful in our mission.
You should reach out to James Bruton! He has one of the best 3d printing / robotics channels on TH-cam. He has done heaps of testing on compliant joints for his robotics projects.
“It is really confusing” I suppose it is if one’s vocabulary is limited, but telling others that something is confusing or difficult to learn can make it confusing or difficult to learn. From a teaching perspective, it’s best not to front load expectations of ease or difficulty of picking up a concept.
Roughly how long would it take a person to complete the creation of a hand from having nothing to a working hand? Im guessing there was a lot of time spent working with cad software to get your 3d model right before you printed it as well.
My old comment was drawn out to say , the give in the one hand is awesome . Have you tried graphite powder , it's a surprising smooth dry powder contact to contact useful makes contact points extra smooth , graphite dusty yes while ultra smooth .
Are you from Hull? I live just down the road if so. For me, the next stage is to use positional analysis via multiple external cameras on the forearm. Similar to car autonomous driving. To deride intent in 3d space as in the spinal portion of the nervous system. You reach for a cup. The cameras detect this and minor adjust to the cup conformation. If you arm extend it then grasps. Ive been wanting to work on this for ambulatory gait too. To make it impossible to fall over. But always knowing the centre of balance etc. The mechanical foot always ALWAYS behave in terms of predicting position.
I think that's the challenge right. Make a 1 to 1 copy of the bones/joints in a hand. And then make mechanical muscles and tendons to make it function like actual muscles and tendons. Instead of simplifying anything.
you should consider using long motor/small diameter with worm gear in the forearm, because they are powerful, precise and slim, also have the advantage of axial output force
Considering that we've literally had nano-tech for almost 50 years, we can create items as complicated that a microchip and we've been able to produce components as small as watch cogwheels, i fail to understand why we haven't been able to produce good bionic parts for the human body.
The hand has a skeleton. The ligaments connect bones together. skin ALSO helps hold the bones in place. The Tendon connects a muscle to bone. Muscles actuate forwards or backwards. Muscles can come in different shapes and structures. I would slow down here and make a unique muscle for each bone. Instead of gears try to use actuated muscles. Even if you didnt build the full hand, this would be a more worthwhile approach. Maybe someone could follow your work. The idea of Tendons and ligaments in robotics using a skeleton structure is exciting. I'm bored of Boston dynamics. Tiny hydraulics for muscles with some kind of skin.
I always have this question to the engineers and designers. When you are designing something like this, why try to mimic and follow the human body? Since human body has limitations, why can't we design these things in a different way to overcome the limits of human body. For example, why put the same limits on the degree of movements? when you are in control of the design make it more flexible than human body 🤔
I think rigid joints and actuators are definitely a step in the right direction. Keep in mind that even rigid joints, with the right software, can *imitate* the effects of a compliant joint where needed, without introducing the latency or backlash of a compliant joint. Others have mentioned James Bruton already, and he has an excellent video demonstrating the topic. th-cam.com/video/qVrPx665NzI/w-d-xo.html
Did you get an msla printer? Like the elegoo Mars? I've always struggled with dimensional accuracy with it, but there are a lot of hack-y things you can do to it to get closer to per pixel accuracy.
Maybe the 17 deg rotation of thumb is there for passive adaptation. At least I cannot rotate my thumb [perhaps.. is it possible by training, dunno], but I can rotate it with an other hand. Thus, it will adapt to the other movements in context. Practical. Also, the rotation allows the joint not to break, if rotated. By tensioning my hand, I can also limit the rotation. This also enables kind of passive DoF of this thumb rotation. Interesting stuff and thank you for thorough and informed analysis. Quality information, subscribing 😊👍
Not many people can start a sentence with "So in my old hand.."
Subbed.
Everyone can, take biology class, Einstein!
@@Nutritional-Yeast WOOSH ✈👨🦲
Fascinating. I have always wondered why the actual human skeleton is not used as the framework for these modern robotic builds, since that is the only way to ensure the exact movements.
My thoughts exactly, I was thinking to use little airbags to replaced the muscles
complexity and reliability are probably the answer, wear and tear can't be healed only replaced and prosthetics are probably expensive already so lowering the cost of getting one and maintaining it might be a priority over complete functionality
The actual human skeleton by itself is not mechanically stable. it requires the connection of ligaments, bursae, and muscle-tendon units for joints to even be stable enough to be measurably driven. The thing about building biomimetics is you have to strike a design balance between reproducing the biological motion, and building mechanical joints sturdy enough to be driven by modern actuation technology (electric motors, pneumatics, hydraulics, etc.). It is not enough to simply build a miniature saddle joint and attach a string to it to drive motion; the joint itself must be a modern, mechanical joint capable of permitting drivable motion only in the movement planes desired, or at least the joint's end effector must have a larger working volume reducible to the biological analog.
Additionally, aside from the issue of scalability, modern electric motors produce forces that are very large, and not easily reduced to give the joint torques required to emulate the fine motor coordination of the human hand (force-torque reduction requires gears/pulley structures, for which there is not much space in the available volume enclosed by the hand).
In general, the skeleton's joints exist along a wide spectrum of "stability" vs. "flexibility". The more Stable joints are rendered stable through the action of active musculature. The more Flexible joints require elastic structures such as ligaments and fascia to permit compliant movement in multiple directions other than the powered directions driven by the large muscle-tendon units. Fascia also facilitates pulley/like structures in a low volumetric profile. The carpal and metacarpal joints are closer to the flexibility part of the spectrum, and in vivo are entirely stabilized by non-rigid, viscoelastic anatomical structures; the bones are there merely to pass along compression loads while executing a movement.
@@BattlesuitExcalibur Excellent answer and an interesting read!
Exactly, I can say the same about robotic 4 legged animals too, they should start implementing biologically accurate limbs,
This is so far removed from my already lacking knowledge of biology and yet you still make the information manageable and interesting.
Really cool, it's great to see how different it is when a very talented and patient individual makes something, comparing it to the competition of companies, where the main goal isn't just fun and usefulness as much as the profits are.
Ignoring "profits" for a moment, the *cost* of producing a hand like this *is* critically important. High cost prosthesis are great for those who can afford them, but you won't change the world by producing a product that only 1% of a potential market can afford. Even with socialized health care and state-funded prosthesis, that cost is still hugely important, because the budget for prosthesis competes with the budgets of other public goods and services.
The challenge of engineering a consumer product is that you have to build as much *quality* in while also trying to minimize costs. Certain design decisions also only work at different scales of production. 3d printing is great for prototypes, but is not economical enough for large scale production.
The real take away, however, is that the quality of the engineering determines how much you have to sacrifice between cost and quality in your prototypes and final product. Even with the best engineers working on a project, that engineering takes *time* to complete and iterate on to produce a worthwhile product. And that time is often much more than individual hobbiests can afford to spend.
At any rate, it will be interesting to see how this prosthesis will develop, and it's always interesting to see behind the curtain of a complex engineering project like this.
I was very happy to hear at the end that he's getting a SLA printer.
I wonder which resin printer it is?
Scott Cress why is your replie to your OWN comment is one day after the original
I don't think SLA is very strong though, just ordinary would've been better.
@@muuubiee There are much stronger SLA compatible resins you can use than the standard hobby grade ones.
About 4 months ago I tried to catch a football and ended up using my thumb to stop it cause being human is fun. I jammed it so badly that it still hurts today. Seeing these joints and the complexity behind them makes me understand why it hurts so bad! I could have googled this but to be honest ya get over the pain. This video though, wow! Thanks for sharing all of this with us. I want to treat my thumb better now lol.
I'm constantly amazed at how much you're learning and discovering (and I'm learning and discovering, too!) about the hand and its constituent parts, and how your project is developing! Nice one, Will! :)
No dislikes. There's hope in the world.
That thumb joints is awesome! Giving a visual way to know how it works helped a lot, thanks!
The work that was involved in this project and video is so impressive and in the future it could become a reality and a benefit to a lot of people thanks 😊
Your design plan is ingenious.. Looking for forward to the finished product
Bro your channel is dope, I'm looking forward to binge watching your channel!
Who twist his fingers watching this video for a proper confirmation?
I do
I did that, honestly 😀
me
yes, and I discovered that I have way more flexing than a normal human depicted in the video, this is a flex
Proper science requires a control group.
Such a fascinating project! Hope you're having fun with this and keep having fun!
Have you thought about solid state compliant mechanisms? Basically you have a single solid part that is designed to flex in a specific way when forces are applied. Might be useful for the 1st-4th phalanges to simplify construction considering the innate limitations on how flexion occurs away from the palm. Veritasium has a video on those mechanisms with a professor who (quite literally) wrote the book on the topic.
Your information is very comprehensible and knowledge based. I prefer your idea of the gear box solution hopefully you will consider using a clutch assembly or actuator for each finger to engage gears independently and simultaniously. I feel very positive about your work.
This channel is underated, u my good sir have gained a sub. Can't wait to see the final product
is it strange that I'm rooting for you and your research and hoping that it helps out lots of people
for the cmc 1 and 2 joints, I'd recommend some form of elastic connection, as, while the muscles have almost literally no control over how they move, they're still elastic to allow conforming to various objects being gripped. at least allow cmc 1 to be flexed backwards by 9.5 degrees. you can see this elasticity by grabbing cmc 1 and 2 and twisting the grabbed hand. it helps in grabbing straight objects like rods and poles and sticks and such.
this freedom of movement can be seen by simply flattening the hand. if the mcp joints were all lined up in a straight line with only freedom in cmc 3 and 4, then cmc 5 would be lifted off of a flat surface.
side note, I number cmc/mcp with index as 1, pinky as 4, thumb as 5.
I almost wish I didn't have hands so as to be more delighted and invested in your progress. :)
Will! I love your journey and appreciate your focus. You are doing a fantastic job. Stay grounded!
When hand prosthetics (and above knee legs) are this flexible it gets really interesting.
Very impressive work!
Please keep it up. Future needs more people like you!
I really like your gearbox design
Hello, Mr.! Maybe it would be useful for you, as sports doctor I really like ‘The Physiology of the joints. Upper limb’ by A.I. Kapandji.
Thank you for your great work!
I do too. I'm a physical therapist from Brazil and Kapandji was my "bible"in college here! lol
I am amazed. This is great, love it
such cool work thanks for sharing your progress
You're becoming some kind of David Sarif and I like it.
this is incredible, i wonder why it takes us so much time to make good prosthetics and organ replacements in the day of almost getting brain implants
dude you are underrated
Really fascinating to watch your journey!
Have you thought about TSA motion system? The one with a twisted string. It can make tons of force with very small and low power motors
Excelentes videos y canal felicidades un poco triste no ver nuevos videos, te encuentras bien
I reaaally hope that you pull this off 🤩🤩
I personally wanted to study bionics but sadly I never really liked math😓 but I love what you are doing. Currently I'm about to start my fisiotherapy studies, do you think that the knowledge that I gain there could be used to one day maybe built a prostetic hand for a pacient??
excellent information! very solid mechanical design and engineering!
nice research
Proper ventilation for those resin prints!
Awesome stuff! Keep up the good work!
That is freaking awesome.
Don't ever leave this project, think if you can somehow get the feature of feel. So the amputee person can feel through that hand. That is the most inportanat thing.
I lost a few fingers and damaged 2 MCP 1 PIP and 1 DIP but have the feel of what i have.
You can save lives if you get that to work.
Good luck!
Hey, impressive job. I wonder to know, what is the function of this looks-like-rubber strap on this hand?
Random question, what is the book you have open with the hand diagrams in it? Looking to get some new anatomy books and I'm loving your videos so I thought I'd get a copy. Thank you and I look forward to more content!
This is gud keep it up can't till is complete
This is fascinating stuff!
One question that always pops into my brain when watching this sort of stuff - and you'll have to forgive my ignorance - we know how a biological hand works, why don't we just replicate that? I mean, if the designs the same (just made from different materials), then replicating the same range of motion, etc should also be possible, yeah? Obviously, the control system would be different, because the human brain/nervous system is a thing.
Like I said, forgive my ignorance on such matters, and keep up the fantastic work!!
I think the reason why is because they can't make the motors small enough without loosing grip strength. Also you'd need to replicate tendons, which cause the movement since they're muscles and contract when u send brain signals. As far as I know they haven't made a synthetic tendon like thing yet, so ud have to sub it with bulkier gears and such.
@@Mysoulismelting I'd imagine that one would place the motors further down the forearm, where there's more space, and have the fingers driven by corded "tendons".
IIRC, there has been some steps in the development of synthetic "muscle", but I haven't followed it, I just remember watching a few clips years back.
@@TheScwall i remember then being able to make synthetic cartilage like for ears and noses. But id imagine it'd be bulky if you were to have the motor further down the arm.
Awesome video series ,loving this content 👌🏼👍🏼
If I become a billionaire....I'm funding this guy....so simple & clear..🔥🔥🔥🚀💯
Awesome!
I'm planing to use a similar hand with an EPOC+, but I'm not too good in programing.
You're an AMAZING person
I was waiting for this
I like this video series, thanks for sharing!
Great work!
I mega impressed. Waiting and thinking.
Interesting as always and great to hear about the SLA! Could you consider changing the background music? I find the quiet yet very intense drumming a bit too distracting.
Hello. And thanks. My University project is about designing a prosthetic arm and my knowledge in this field is next to non. But I am learning so much from you so I appreciate that. Dw I'll come up with my own design and if possible I'll reference things correctly (I hate writing repots but what can you do?)
The hand that is so simple for nature and in the same time so complex to human
If you are having trouble with space and the gearboxes, have a look at a Cycloidal gearbox, it allows for really flat, high reduction, gearboxes.
Though there isn't many small ones of these for sale though.
All the best keep doing
Ever tried using a ball and socket joint for the basal carpometacarpal, likely restrained by a compliant webbing?
inquiry: why not combine M2 and M3 into one broader bone integral to the carpal section, and combine M4 and M5 into another single broader bone and rely on a combined swivel and hinge or 2-axis saddle articulation to the lateral side of the carpal section?
Would be interesting when you get an sla printer to use some of the available semi flexable resin to print out all the joints, bones and tendons in one print to reduce complexity of mechanical joints, sla would allow for that level of detail in one print, instead of mechanical hinges "living" ones , although this might not provide the framework for control you are trying to make
Added bonus of westworld vibes
Is it possible that the thumb's CMC joint's pronation/supination aren't directly active movments, like flexion/extension and abduction/adduction are? As in, the metacarpal twists throughout it's range of movement, rather than being able to twist on demand like other movements can be done?
I don't study anatomy, I know this joint's pronation/supination is severely understudied, I can only find one big study that did 4d scans of the joint, though this study only showed that pronation/supination exists but not how those movements are actuated. But I came to this idea based on the fact that you can actuate movements in joints through external force, basically pushing/pulling on a joint can involuntarily actuate it. Except you cannot make your first metacarpal twist with this method, it stays fairly rigid, like the other metacarpals do.
Thus I believe this joint's pronation/supination may be passive, maybe not truly passive but also not an active actuation. If this is true, the twist would only be there to allow greater range of movement, which may assist in opposition and reposition movements. Making the joint technically 3dof, but only active in two degrees, with the third being passive between the other two. But again, maybe it's not a true passive, maybe it is partially active due to how the tendons pull the metacarpal, I'm not really sure how to define this if it's not truly passive but also not truly active.
Excellent
Thank you so much for sharing all those information with us. As I do provide free prosthetics devices in India that information will also be useful in our mission.
You should reach out to James Bruton! He has one of the best 3d printing / robotics channels on TH-cam. He has done heaps of testing on compliant joints for his robotics projects.
That cream robotic hand looks like a table tennis bat or paddle !
Nagyon profi!!
i would like to study and do some research about bionic parts. But i don't know where to begin, can you help me?
Love it.
“It is really confusing”
I suppose it is if one’s vocabulary is limited, but telling others that something is confusing or difficult to learn can make it confusing or difficult to learn. From a teaching perspective, it’s best not to front load expectations of ease or difficulty of picking up a concept.
So good
you would have better range if you added the same movement to the pointer finger. although you dont really need it in the middle finger
Would love to see that glove being used in vr settings. that would make for pretty good hand tracking.
Where are you planning to use this?
I'm building my own robotic hand on my channel. I just made my channel and decided to continue my work. I'm going for a realistic approach.
What was the song you used in the back ground
What program did you use to model the hand ?
Roughly how long would it take a person to complete the creation of a hand from having nothing to a working hand? Im guessing there was a lot of time spent working with cad software to get your 3d model right before you printed it as well.
I love biomechanics
My old comment was drawn out to say , the give in the one hand is awesome . Have you tried graphite powder , it's a surprising smooth dry powder contact to contact useful makes contact points extra smooth
, graphite dusty yes while ultra smooth .
Are you from Hull? I live just down the road if so. For me, the next stage is to use positional analysis via multiple external cameras on the forearm. Similar to car autonomous driving. To deride intent in 3d space as in the spinal portion of the nervous system. You reach for a cup. The cameras detect this and minor adjust to the cup conformation. If you arm extend it then grasps. Ive been wanting to work on this for ambulatory gait too. To make it impossible to fall over. But always knowing the centre of balance etc. The mechanical foot always ALWAYS behave in terms of predicting position.
Jet from Pinewood Studios - I tried to get hold of you three times now. Shame as we wanted to chat.
You're pretty good.
You should check out "Shimano di2 disassembly" video.
I think you'd get some ideas
I think that's the challenge right. Make a 1 to 1 copy of the bones/joints in a hand. And then make mechanical muscles and tendons to make it function like actual muscles and tendons. Instead of simplifying anything.
you should consider using long motor/small diameter with worm gear in the forearm, because they are powerful, precise and slim, also have the advantage of axial output force
Please make a video on carpal joints and how to program the range of motions.
Very interesting!
The entire time I watching this video I was hearing
*”building a sentry”*
Let's hope the third hit doesn't break the hand but instead breaks your face
"Ain't that a cute little gun?"
Instant sub!
Where'd you go? What's the current state of the project?
Considering that we've literally had nano-tech for almost 50 years, we can create items as complicated that a microchip and we've been able to produce components as small as watch cogwheels, i fail to understand why we haven't been able to produce good bionic parts for the human body.
The hand has a skeleton. The ligaments connect bones together. skin ALSO helps hold the bones in place. The Tendon connects a muscle to bone. Muscles actuate forwards or backwards. Muscles can come in different shapes and structures. I would slow down here and make a unique muscle for each bone. Instead of gears try to use actuated muscles. Even if you didnt build the full hand, this would be a more worthwhile approach. Maybe someone could follow your work. The idea of Tendons and ligaments in robotics using a skeleton structure is exciting. I'm bored of Boston dynamics. Tiny hydraulics for muscles with some kind of skin.
Brilliant
keep up!!!
good stuff
Have you thought of adding extra fingers? Two thumbs would be quite interesting , or an extra ring finger for some mean six string bass...
I always have this question to the engineers and designers. When you are designing something like this, why try to mimic and follow the human body? Since human body has limitations, why can't we design these things in a different way to overcome the limits of human body. For example, why put the same limits on the degree of movements? when you are in control of the design make it more flexible than human body 🤔
I think rigid joints and actuators are definitely a step in the right direction. Keep in mind that even rigid joints, with the right software, can *imitate* the effects of a compliant joint where needed, without introducing the latency or backlash of a compliant joint.
Others have mentioned James Bruton already, and he has an excellent video demonstrating the topic.
th-cam.com/video/qVrPx665NzI/w-d-xo.html
watching this just in case count dooku cuts off my arm
Did you get an msla printer? Like the elegoo Mars? I've always struggled with dimensional accuracy with it, but there are a lot of hack-y things you can do to it to get closer to per pixel accuracy.
Maybe the 17 deg rotation of thumb is there for passive adaptation. At least I cannot rotate my thumb [perhaps.. is it possible by training, dunno], but I can rotate it with an other hand.
Thus, it will adapt to the other movements in context. Practical. Also, the rotation allows the joint not to break, if rotated.
By tensioning my hand, I can also limit the rotation. This also enables kind of passive DoF of this thumb rotation.
Interesting stuff and thank you for thorough and informed analysis. Quality information, subscribing 😊👍