For the life sized one that you can ride try putting the chair on some sort of gyro or suspension because the Omni-directional wheel one looked incredibly bumpy and I’m sure this one will be even more bumpy. Just a thought. Love the builds.
Yh one of those 6DOF gyroscope stabailised platform with appropriate emu's then put a set on that, you'll be level no matter what the 'legs' are doing them :)
We need the full size walking robot for sure. I always see these pop up in my feed at 1am right when I'm going to bed but stay up to watch them anyways. James's creations are always sure to be mind-blowing.
Those are not called cams, they are cranks. A cam has a rolling or sliding follower and can be shaped to an arbitrary profile, whereas a crank simply has an eccentric rotary joint. Engines have both a crankshaft and a camshaft, but this type of walker mechanism uses a crankshaft.
@@Matty.Hill_87 These terms far predate automotive and robotics. Cams being dated as early as around 600 BC in China, and cranks being dated as early as the Western Han Dynasty 202 BC - 9 AD in China, not to mention these have been fairly prominent throughout both eastern and western history in machinery and automata. They're fairly simple concepts with well known definitions, no matter what field of mechanics you belong to. There is no 'crossing over' to be had; James blatantly used the wrong term, stop trying to find excuses for why the wrong term was used.
As I'm sure you already know, the theo jannison strandbeast linkage is already optimized for taking that jerky movement out. If you simply swap the linkage that you have with the strandbeast linkage, I think you'll notice that there's much less jerky movement in the stepping motion. If you look at the path of motion for the tip of each of your legs you'll notice it moves in a strange squashed ellipse. With The strand beast linkage, that ellipse looks more like an airfoil with the flat side facing the ground. This means that it's got smooth translational motion. Nice job though. I really like the concept. I think I want to make one of these and stick a light on it for when I'm working on my car.
Was thinking - if you can optimize the crank and follower path to displace more vertically rather than horizontally, you could increase the RPM the crank turns at to make the steps faster and smaller. This could improve the travel over debris because leg lifting is of critical importance.
id really love to see a rideable version of your snake robot. that seems to me to be your most successful design for climbing over rough terrain and it would be cool to see it going offroad at full scale
This is super cool, I imagine the full sized version looking like a more colorful version of Darth Sidious' walker-chair from the Star Wars prequels lol
i kind of like the idea of a large number of small omni-walkers like these, supporting a person transport. or a really big one with something like wheel chair wheels on each of the legs :O. Cant wait to see the larger version :D
James, thank you for mentioning DIYwalkers - my son Ben and I were excited to see this. PS. I think you chose the best leg design for simplicity and strength, but it's a 4 bar linkage that doesn't step high on the inside of its foot-path. Strider Ver 3 is a different mech that steps higher...but it's more complicated with 10 bars per leg pair. If you don't plan on walking on really rough terrain you may want to stick with this 4-bar linkage rather than the more complicated Strider linkage. Again, thanks so much for the shout-out James! Wade
that thing is inexplicably sooo weird.whenevber i see its silly walk i cant stop thinking about how weird it is. i think you've built the official mascot of the ministry of silly walks.
I like your idea of "small" prototype. If it was half the size I would consider it big. You must go through some amount of plastic. What do you do with all the waste and unwanted prints? Do you have any ideas about recycling/reusing unwanted/failed PLA? Would love to hear your ideas on this. Great video. Keep up the good work.
are you familiar with Jamie Mantzel? he designed a 6-legged robot that can walk in any direction using just two motors. he spent around 6 years making a giant metal version during which time he managed it to get it made into a toy and sold in stores ("Attacknid"). while he finished the metal version he never got it to the stage he wanted with his limited resources... the dude lives a very unconventional off-grid lifestyle. his goal was always to build one big enough and strong enough he could sit on... he has said it would be cool to see someone like furze make one. anyway, you might find it interesting.
I immediately thought of the elephants that hold up the Discworld. If you up the size you could use more walker units, probably six of them, and it would look like a fanciful carousel. Or I suppose you could arrange the Walker units in parallel like a team of horses and offset the angle of the wheels on the feet 45° like a mecanum wheel.
I found a mechanism you would like on a very old Dixon zero turn mower. The engine constantly rotates a sphere and there are bowl shaped driveshafts on either side with break pads lining the concave sides. The drive levers will move these bowl shaped plates up or down to contact the sphere and drive the wheels. I have photos I can share if you are interested
I love how crazy your omni-wheel designs are getting. I would never have considered a hybrid onmi-walker, and cant wait to see a large scale model! Might need a 3 point harness for the big one, looks like it could be a rough ride
Ever considered making a 6 legged walker? Since they are passively stable when both walking and standing programming them should be fairly easy. If you make it in a circular configuration it could fairly reliably move in any direction, but would require a lot powered joints to do so
You could try making the passive wheels larger and just moving them higher up on the legs. That should make it a little easier for the wheels to roll over small objects when the robot is oriented in the wrong direction.
It might be interesting to make it automatically calculate slight serpent-like paths even when following a straight trajectory to artificially make all the legs contribute to the movement no matter where you're pointing it
Nice work. Would love to hear if you are concerned about what the seated experience would be. Perhaps a mech standup look would be better but are you going to try compensate the rough ride beyond the other lower improvements you mentioned?
7:24 I think you mean that the slower tracks (the \ /) must run 50% slower than the top horizontal one. (I would post a link to a picture explaining it, but unfortunately YT does not allow external links)
"Sideways" seems like the forwards and backwards direction as it would be the direction you face when scaling objects and having the most ground clearance during a run. Would probably be worth aligning your seat with this direction if you make a full size version.
Agree, those are cranks, not cams. But that got me thinking: your device is just a complicated version of your 3-sphere machine but with the legs being analogous to knobs or 3D cams replacing the spheres. It would be a much simpler design with fewer moving parts to go that direction.
Yeah I wanna see one you can ride on! But first you have to make it a lot less janky. Maybe more walker units in each direction, this is basically the trick the strandbeest uses to move more smoothly. Frankly, I'm blown away that this works!! Great work!
This is remarkable! Such an interesting design. Gotta say though, out of all the things you've made, this is the first that gives me the heebie-jeebies seeing it in motion hahaha
I wonder if you drive both cams independently on each leg, you can drive them in a way where the undriven legs constantly "step in place" so it'll always step over obstacles.
Looks really cool, but also bumpy. Not sure if it can be made more smooth. Any tweaking needed in dimensions of driving rods? Also looked like clearing wheels to ground was radius of drive axis, and stride was diameter. Maybe try to smooth ride out first before going bigger? Love the project!!!
cool design, I wonder how it would handle wear and tear like if you were to make a mars rover with this design and all the small wheels got broken would it still be able to move? how does it do in sand or on a hill? and can you have more than 3 units like 4 or 6 to help with hill climbing and if a few sets of legs die?
YAAASSS amazing!! please make one that can be sat on, there are so many people who would love to further hack on this idea and develop some excellent new accessibility aids - thanks so much for your hard work and your open source ethos James, you rock!
A very interesting project for sure! Though it seems to me like the omnidirectional functionality is not as useful on walking robots, it might be more worthwile to simply build a spider-like robot which could probably walk in all directions with legs alone.
seems to me like you should rotate the controls so when you go forward it does the sideways movements and vice versa, that way you can go forward and take advantage of the steppy
After seeing that thing walk, I thought, about speed, so what if you arrange 6 or 8 spider-legs in a circle, so you could easily move in all directions? Each leg can pivot from right to left a few degrees 15° in each direction for example and also has two pivot points perpendicular to the first one along the leg, in that way the legs on the side can move the first pivot while the other legs can orient for the moving direction and basically move like a dog or so. So a spider-walking-vehicle. That would be really nice to try, ask if you have questions.
Honestly this is awesome. And while I fully support going full scale with this. I think you should investigate plantigrade walkers before going full scale with this.. the plantigrade walkers provide an extremely smooth ride. Would be awesome for a full scale build. Although it will be very much so more complicated.
Hi, can you change your control software to always rotate when moving in a specific direction to avoid foot snags like you mention. Probably already suggested 🙂.
I've written to you in the past about how much I think your skills and inventions could be utilized by disabled people, I'm a wheelchair user myself. I've since learned about the enormous undertaking that it is to get something declared as a medical device. I recently had to get a new wheelchair because there was not a single one on the market capable of climbing my hill. It was months of trial and error, the only things on the market capable were huge devices incapable of being used indoors. I ended up finding the solution to my problem and a great many other people's problems by discovering a Japanese company I'm going to leave nameless because it's not the point. Many companies are doing this. They have a good product for the disabled but they know it will be infringed before they get Ada approval. So they are releasing them as personal electric vehicles and US doctors are able to prescribe these, at the moment it's an enormously valuable loophole. I'm sure you have a great deal on your plate, but perhaps even offering your plans in a scaled up version that could carry a person, I know of many Machinists in the space that could replicate the parts. I assure you that this market is almost strictly inhabited by some of the lowest human beings on the planet. You would be doing a world of good to people in need if you considered doing this. I think there's already a lot of people out there that could make parts and source motors and scale up your designs. The most important things are maneuverability, being under 28 in wide, and particularly some of the rounder designs allow for turning around within the footprint which would be enormously valuable. The best products on the market now essentially utilize tank steering with a pair of powered motors. I can personally attest that a third powered motor would make all the difference in the world. All of these chairs utilize free-wheeling casters or Omni Wheels in front that make it virtually impossible to corner on a slope. Any of your three wheel designs would fix this. I'm happy to put you in contact with more capable people than me. My email address is my name with no hyphen and a period between my first and last name at gmail. I understand if I don't hear from you, but I appreciate your giving it some thought. Be well
What if you programmed it so that the stationary set of legs oscillated back and forth? That way it would still lift its legs over obstacles, but it might not alter the direction of travel of the robot significantly.
Interesting video like always! Another interesting approach to omnidirectional walking is with an omnidirectional mechanism (without needing wheels) like the jaimie mantzel mechanism, which is also the one used in the hexbug spider toy.
You should try to make a rideable version of the snake robot from one of your previous videos!!! With the scaled up version you could put a seat in one of the two middle segments.
It strikes me that that type of drive mechanism doesn't translate constant motor speed to constant linear speed... and I wonder if it would drive nicer if you had encoders on the drive shafts and accounted for this non-constant speed in real-time when determining the motor speeds to achieve a particular direction of motion. This could make for less twisting back and forth when trying to move in a straight line.
Initially, I was a little put off by the inefficiency of loosing the effort done by the motors to the cosine of the direction travelled, but actually it occurs to me that there may be a clever use case here? If you are after efficiency, you use the leading set of legs "straight on" and you will go as fast as 1 motor can run the legs, and the other 2 sets will idle. However, it you drive from the rear legs, you should get much higher (torque?) pulling force, as not only have you doubled the motors driving but you are also using the incline plane principle? Can you test the "pulling" power in various drive directions?
If the 60 degree angle cuts the sideways dimension by half, shouldn't the appropriate compensation be to double it? I.e., to add 100%, not 50% of the main velocity? Maybe I'm not understanding something...
If you cut something by half, but then you add back 50% of what you have left, you don't get back the whole amount. Or maybe you're saying that you add back the half that you took away?
This is such a bizarre little creature. I wonder what the limit on optimisation for the geometry is. As in, I wonder how close you can get to a "perfect" geometry where it doesn't wobble around like that.
Rather than rigid legs it would be interesting to see rubberised flexible legs. Not sure how best to explain it, but imagine liquorice of cheese sticks. They could naturally flex and grip in all directions without the need for wheels.
man one thing I would love for you (or someone!) to share are the recipes you use for screws and the size holes you use. For example, if you want to use XYZ wood screw then make this size hole on one side with this amount of bevel/chamfer for the head and on the threaded size, this size hole. I fight with figuring out what hole sizes to use all the time and have to go through many prints before I finally get something that holds but isn't too tight or too loose!
For the life sized one that you can ride try putting the chair on some sort of gyro or suspension because the Omni-directional wheel one looked incredibly bumpy and I’m sure this one will be even more bumpy. Just a thought. Love the builds.
Yh one of those 6DOF gyroscope stabailised platform with appropriate emu's then put a set on that, you'll be level no matter what the 'legs' are doing them :)
th-cam.com/video/LKJ4Iy-HxvE/w-d-xo.html
We need the full size walking robot for sure. I always see these pop up in my feed at 1am right when I'm going to bed but stay up to watch them anyways. James's creations are always sure to be mind-blowing.
Those are not called cams, they are cranks. A cam has a rolling or sliding follower and can be shaped to an arbitrary profile, whereas a crank simply has an eccentric rotary joint. Engines have both a crankshaft and a camshaft, but this type of walker mechanism uses a crankshaft.
Maybe the automotive terms don't cross over into robotics in the same way, you're spot on though if not
@@Matty.Hill_87 These terms far predate automotive and robotics. Cams being dated as early as around 600 BC in China, and cranks being dated as early as the Western Han Dynasty 202 BC - 9 AD in China, not to mention these have been fairly prominent throughout both eastern and western history in machinery and automata. They're fairly simple concepts with well known definitions, no matter what field of mechanics you belong to. There is no 'crossing over' to be had; James blatantly used the wrong term, stop trying to find excuses for why the wrong term was used.
@@xaytana you are right, but it’s also ok to be a bit more chill about being right
yep! you're correct
@@ThatGuy-fi9bm WeLl AcTuAlLy
Haha this looks absolutely ridiculous! Can't wait to see a life-sized version 😃
I'm excited to show this to my son and look forward to seeing the full sized "Ride On" iteration.
As I'm sure you already know, the theo jannison strandbeast linkage is already optimized for taking that jerky movement out. If you simply swap the linkage that you have with the strandbeast linkage, I think you'll notice that there's much less jerky movement in the stepping motion.
If you look at the path of motion for the tip of each of your legs you'll notice it moves in a strange squashed ellipse. With The strand beast linkage, that ellipse looks more like an airfoil with the flat side facing the ground. This means that it's got smooth translational motion. Nice job though. I really like the concept. I think I want to make one of these and stick a light on it for when I'm working on my car.
Was thinking - if you can optimize the crank and follower path to displace more vertically rather than horizontally, you could increase the RPM the crank turns at to make the steps faster and smaller. This could improve the travel over debris because leg lifting is of critical importance.
I could just watch this thing walk back and forth for hours... your robot army is coming along nicely!
Great work as always James!
id really love to see a rideable version of your snake robot. that seems to me to be your most successful design for climbing over rough terrain and it would be cool to see it going offroad at full scale
This is super cool, I imagine the full sized version looking like a more colorful version of Darth Sidious' walker-chair from the Star Wars prequels lol
"I think this will be super good to make into something absolutely massive that's big enough that I can ride on."
DO IT!!!
Beautiful as always
As for your sponsor, I am a long time user of onshape now, amazing online 3d CAD sowftware, I just LOVE it!
i kind of like the idea of a large number of small omni-walkers like these, supporting a person transport. or a really big one with something like wheel chair wheels on each of the legs :O. Cant wait to see the larger version :D
James, thank you for mentioning DIYwalkers - my son Ben and I were excited to see this. PS. I think you chose the best leg design for simplicity and strength, but it's a 4 bar linkage that doesn't step high on the inside of its foot-path. Strider Ver 3 is a different mech that steps higher...but it's more complicated with 10 bars per leg pair. If you don't plan on walking on really rough terrain you may want to stick with this 4-bar linkage rather than the more complicated Strider linkage. Again, thanks so much for the shout-out James! Wade
that thing is inexplicably sooo weird.whenevber i see its silly walk i cant stop thinking about how weird it is. i think you've built the official mascot of the ministry of silly walks.
Awesome! To help with traversing obstacles you could always run the passive legs too even though they won't 'push'.
Yes. Rideable version please!
11:11 "that's why we build small prototypes"
I guess "small" is relative 😂 Great build as usual!
-Steven
Can't wait to see the big version with big grippier wheels, suspension for smoother ride and better terrain handling
I can't believe how well this works, I can't wait to see a full size version
I like your idea of "small" prototype. If it was half the size I would consider it big. You must go through some amount of plastic. What do you do with all the waste and unwanted prints? Do you have any ideas about recycling/reusing unwanted/failed PLA? Would love to hear your ideas on this.
Great video. Keep up the good work.
I think he did a video on that if I remember correctly, just search his channel.
Every video you make is ALWAYS inspiring me to get back into making things when I'm down. Thank you for being the catalyst I need.
are you familiar with Jamie Mantzel? he designed a 6-legged robot that can walk in any direction using just two motors. he spent around 6 years making a giant metal version during which time he managed it to get it made into a toy and sold in stores ("Attacknid"). while he finished the metal version he never got it to the stage he wanted with his limited resources... the dude lives a very unconventional off-grid lifestyle. his goal was always to build one big enough and strong enough he could sit on... he has said it would be cool to see someone like furze make one. anyway, you might find it interesting.
This is awesome.
And this should definitely be a full scale project
A full size build of this would be incredible! Please, please make it happen!
I immediately thought of the elephants that hold up the Discworld. If you up the size you could use more walker units, probably six of them, and it would look like a fanciful carousel. Or I suppose you could arrange the Walker units in parallel like a team of horses and offset the angle of the wheels on the feet 45° like a mecanum wheel.
I found a mechanism you would like on a very old Dixon zero turn mower. The engine constantly rotates a sphere and there are bowl shaped driveshafts on either side with break pads lining the concave sides. The drive levers will move these bowl shaped plates up or down to contact the sphere and drive the wheels. I have photos I can share if you are interested
I love how crazy your omni-wheel designs are getting. I would never have considered a hybrid onmi-walker, and cant wait to see a large scale model! Might need a 3 point harness for the big one, looks like it could be a rough ride
I think practical applications to a robot design are great to see. Riding Omni-Walker gets an upvote from me!
10:41 Just imagining how hilarious a powerful one would be, on a pile of boards. Boards would be flying everywhere!
Definitely want to see a full-size ride-on version of this. It looks wild!
Ever considered making a 6 legged walker? Since they are passively stable when both walking and standing programming them should be fairly easy. If you make it in a circular configuration it could fairly reliably move in any direction, but would require a lot powered joints to do so
This looks great, I'd love to see a large one with passive or active leg dampeners that help with the bumps.
You could try making the passive wheels larger and just moving them higher up on the legs. That should make it a little easier for the wheels to roll over small objects when the robot is oriented in the wrong direction.
you always make the coolest stuff
This is really cool James!! You're not that far away from the spider tank in ghost in the shell!!
Could you make stairs mode where legs lift higher
Best channel on yt. Hands down
Are you thinking of trying a bio-mimetic design at any point (even if oversized)? Or will you stay with non-standard motion systems?
this blew my mind. so excited to see the big version!
Now this might be something worth developing. Unique and can do things that nothing else can.
It might be interesting to make it automatically calculate slight serpent-like paths even when following a straight trajectory to artificially make all the legs contribute to the movement no matter where you're pointing it
To improve clearance you can make the rear leg swing back and forth so the legs stay moving and you won't have to twist and turn to go over obstacles
Nice work. Would love to hear if you are concerned about what the seated experience would be. Perhaps a mech standup look would be better but are you going to try compensate the rough ride beyond the other lower improvements you mentioned?
That lego model moved so smoothly, that's a really cool idea
Crazy, amazing. You’re a mad scientist, my dude. Absolutely for the life sized one!
7:24 I think you mean that the slower tracks (the \ /) must run 50% slower than the top horizontal one.
(I would post a link to a picture explaining it, but unfortunately YT does not allow external links)
rs3VpxV at imgbb
"Sideways" seems like the forwards and backwards direction as it would be the direction you face when scaling objects and having the most ground clearance during a run. Would probably be worth aligning your seat with this direction if you make a full size version.
Please prioritize the walking sliding robot for you to ride on. You're an AMAZING ENGINEER!!!
Agree, those are cranks, not cams. But that got me thinking: your device is just a complicated version of your 3-sphere machine but with the legs being analogous to knobs or 3D cams replacing the spheres. It would be a much simpler design with fewer moving parts to go that direction.
Awesome. I had to watch it in slo-mo to see how the legs work in concert. 👍
I'm sure the walking mechanism you used was a better option, but ngl I love how strandbeest legs look. woulda been cool to see an omnidirectional one
You need to work for Boston dynamics!
Life size, looks relatively straightforward and wouldn't require a lot of tricky work to get it going.
Could you have the passive axis moving back and forth to generate steps for clearance with a net zero horizontal movement?
Yeah I wanna see one you can ride on! But first you have to make it a lot less janky. Maybe more walker units in each direction, this is basically the trick the strandbeest uses to move more smoothly. Frankly, I'm blown away that this works!! Great work!
I wonder if you could build a big sphere and sit inside it
This is remarkable! Such an interesting design. Gotta say though, out of all the things you've made, this is the first that gives me the heebie-jeebies seeing it in motion hahaha
I wonder if you drive both cams independently on each leg, you can drive them in a way where the undriven legs constantly "step in place" so it'll always step over obstacles.
It reminds me of the Tachikoma from ghost in the Shell. I think that had wheels on its feet too.
Looks really cool, but also bumpy. Not sure if it can be made more smooth. Any tweaking needed in dimensions of driving rods?
Also looked like clearing wheels to ground was radius of drive axis, and stride was diameter.
Maybe try to smooth ride out first before going bigger? Love the project!!!
cool design, I wonder how it would handle wear and tear like if you were to make a mars rover with this design and all the small wheels got broken would it still be able to move? how does it do in sand or on a hill? and can you have more than 3 units like 4 or 6 to help with hill climbing and if a few sets of legs die?
Do I want to watch James astride a giant, mechanical, mutant, rollerblading arachnid? Oh hell yea!
Excited for the rideable one
I wonder if you experimented with the geometry of the legs if they could step up higher and clear larger obstacles?
It would be awesome to see a big version of this but nice having a 3d printed robot video again
Can this design be adjusted so it doesn't wobble so much on each step, or would it need a more complicated linkage like the Strandbeest?
YAAASSS amazing!! please make one that can be sat on, there are so many people who would love to further hack on this idea and develop some excellent new accessibility aids - thanks so much for your hard work and your open source ethos James, you rock!
A very interesting project for sure! Though it seems to me like the omnidirectional functionality is not as useful on walking robots, it might be more worthwile to simply build a spider-like robot which could probably walk in all directions with legs alone.
seems to me like you should rotate the controls so when you go forward it does the sideways movements and vice versa, that way you can go forward and take advantage of the steppy
I think a gimbal stabilized seat would be an interesting addition to this project.
After seeing that thing walk, I thought, about speed, so what if you arrange 6 or 8 spider-legs in a circle, so you could easily move in all directions? Each leg can pivot from right to left a few degrees 15° in each direction for example and also has two pivot points perpendicular to the first one along the leg, in that way the legs on the side can move the first pivot while the other legs can orient for the moving direction and basically move like a dog or so. So a spider-walking-vehicle. That would be really nice to try, ask if you have questions.
So fascinating. Does anyone know the qualifications of James? or self taught etc...
Would there be a chance to get the stl files rather than only the CAD files as well?
I’ve been really enjoying your projects. It’s fun how you take all these fun ideas and mix them up and take them to extremes. Great work
Thanks for the consistent uploads! I like to watch you after I get home from school
Honestly this is awesome. And while I fully support going full scale with this. I think you should investigate plantigrade walkers before going full scale with this.. the plantigrade walkers provide an extremely smooth ride. Would be awesome for a full scale build. Although it will be very much so more complicated.
That machine is spoopy and I love it.
Hi, can you change your control software to always rotate when moving in a specific direction to avoid foot snags like you mention. Probably already suggested 🙂.
These smaller projects are extremely interesting, but I'd love a new big project.
Another open dog, something even more amazing?
Dude! Anything you can ride is worth the build! Peace!✌😎🏴☠️
I've written to you in the past about how much I think your skills and inventions could be utilized by disabled people, I'm a wheelchair user myself. I've since learned about the enormous undertaking that it is to get something declared as a medical device. I recently had to get a new wheelchair because there was not a single one on the market capable of climbing my hill. It was months of trial and error, the only things on the market capable were huge devices incapable of being used indoors. I ended up finding the solution to my problem and a great many other people's problems by discovering a Japanese company I'm going to leave nameless because it's not the point. Many companies are doing this. They have a good product for the disabled but they know it will be infringed before they get Ada approval. So they are releasing them as personal electric vehicles and US doctors are able to prescribe these, at the moment it's an enormously valuable loophole. I'm sure you have a great deal on your plate, but perhaps even offering your plans in a scaled up version that could carry a person, I know of many Machinists in the space that could replicate the parts. I assure you that this market is almost strictly inhabited by some of the lowest human beings on the planet. You would be doing a world of good to people in need if you considered doing this. I think there's already a lot of people out there that could make parts and source motors and scale up your designs. The most important things are maneuverability, being under 28 in wide, and particularly some of the rounder designs allow for turning around within the footprint which would be enormously valuable. The best products on the market now essentially utilize tank steering with a pair of powered motors. I can personally attest that a third powered motor would make all the difference in the world. All of these chairs utilize free-wheeling casters or Omni Wheels in front that make it virtually impossible to corner on a slope. Any of your three wheel designs would fix this. I'm happy to put you in contact with more capable people than me. My email address is my name with no hyphen and a period between my first and last name at gmail. I understand if I don't hear from you, but I appreciate your giving it some thought. Be well
All of this always fascinated me, incredible work. How about using this somehow in VR omnidirectional threadmills?
And I mean the omnidirectional technology not this one in particular
It's so goofy! Is love to see a big version
The projects are looking good
I definitely think you should create a ride-on dodecapod.
What if you programmed it so that the stationary set of legs oscillated back and forth? That way it would still lift its legs over obstacles, but it might not alter the direction of travel of the robot significantly.
Interesting video like always!
Another interesting approach to omnidirectional walking is with an omnidirectional mechanism (without needing wheels) like the jaimie mantzel mechanism, which is also the one used in the hexbug spider toy.
You should try to make a rideable version of the snake robot from one of your previous videos!!! With the scaled up version you could put a seat in one of the two middle segments.
Would be interesting to see you ride a larger size one. It does look like you'd wobble around a lot in the seat though.
It strikes me that that type of drive mechanism doesn't translate constant motor speed to constant linear speed... and I wonder if it would drive nicer if you had encoders on the drive shafts and accounted for this non-constant speed in real-time when determining the motor speeds to achieve a particular direction of motion. This could make for less twisting back and forth when trying to move in a straight line.
Initially, I was a little put off by the inefficiency of loosing the effort done by the motors to the cosine of the direction travelled, but actually it occurs to me that there may be a clever use case here? If you are after efficiency, you use the leading set of legs "straight on" and you will go as fast as 1 motor can run the legs, and the other 2 sets will idle. However, it you drive from the rear legs, you should get much higher (torque?) pulling force, as not only have you doubled the motors driving but you are also using the incline plane principle? Can you test the "pulling" power in various drive directions?
If the 60 degree angle cuts the sideways dimension by half, shouldn't the appropriate compensation be to double it? I.e., to add 100%, not 50% of the main velocity? Maybe I'm not understanding something...
If you cut something by half, but then you add back 50% of what you have left, you don't get back the whole amount. Or maybe you're saying that you add back the half that you took away?
Can you try making a Strandbeest style walker?
This is such a bizarre little creature. I wonder what the limit on optimisation for the geometry is. As in, I wonder how close you can get to a "perfect" geometry where it doesn't wobble around like that.
What if you put tank treads on one set of legs so the tank treads can side step?
And I for one will welcome our robot overlords, with James as their human emissary…
Rather than rigid legs it would be interesting to see rubberised flexible legs. Not sure how best to explain it, but imagine liquorice of cheese sticks. They could naturally flex and grip in all directions without the need for wheels.
You'd still need the wheels for when one set of legs is at 90 degrees, otherwise those feet would draaaaag. You need them for the passive direction.
Mojo's palanquin!
@@jama211 Yeh good point. Seems to be a general issue with the design.
man one thing I would love for you (or someone!) to share are the recipes you use for screws and the size holes you use. For example, if you want to use XYZ wood screw then make this size hole on one side with this amount of bevel/chamfer for the head and on the threaded size, this size hole. I fight with figuring out what hole sizes to use all the time and have to go through many prints before I finally get something that holds but isn't too tight or too loose!