Will this Omni-Directional Tank Work?
HTML-код
- Опубликовано: 14 мар 2022
- AD: Invest in blue-chip art for the very first time by signing up for Masterworks:
masterworks.art/jamesbruton. Purchase shares in great masterpieces
from artists like Pablo Picasso, Banksy, Andy Warhol, and more.
See important Masterworks disclosures:
www.masterworks.io/about/disc...
I’ve previously built various omnidirectional vehicles. Those included a machine I built to promote a LEGO set, which used three standard omni-directional wheels. We’ll talk more about the maths involved in calculating the wheel velocities later in this video, but using a combination of wheel velocities the machine was able to move in any direction as well as rotate.
I built another similar machine, but this one used ball shaped wheels which were inspired by a project at Osaka University. Each wheel is made up of two hemispheres and each wheel has one driven axis, just like standard omni-wheels. This allowed it to move in any direction or rotate.
Neither of these vehicles where particularly good at climbing over objects though. I built various other projects which were much better at climbing over obstacles. In some cases these used rubber tank tracks, or other weird wheels like the Pedrail machine which had extending feet around its wheels’ circumferences. Or just a lot of wheels like my actively flexible snake robot which could flex in two axes to get traction and also bridge gaps.
But what about if we had a tracked vehicle that was also omni-directional? In this video I’m going to attempt to make tank tracks which can also slide sideways, so we can make a vehicle with three of them arranged just like omni-wheels.
CAD & Code: github.com/XRobots/OmniTracks
You can support me on Patreon or buy my Merchandise:
***************************
Patreon: / xrobots
Merchandise: teespring.com/stores/james-br...
***************************
Affiliate links - I will get some money of you use them to sign up or buy something:
***************************
Matterhackers 3D printing supplies: www.matterhackers.com?aff=7500
Music for your RUclips videos: share.epidemicsound.com/xrobots
***************************
Other socials:
***************************
Instagram: / xrobotsuk
Facebook: / xrobotsuk
Twitter: / xrobotsuk
***************************
CAD and Code for my projects: github.com/XRobots
Huge thanks to my Patrons, without whom my standard of living would drastically decline. Like, inside out-Farm Foods bag decline. Plus a very special shoutout to Lulzbot, Inc who keep me in LulzBot 3D printers and support me via Patreon.
HARDWARE/SOFTWARE
Below you can also find a lot of the typical tools, equipment and supplies used in my projects:
Filament from: www.3dfuel.com/
Lulzbot 3D Printers: bit.ly/2Sj6nil
Bearings from: simplybearings.co.uk/
Lincoln Electric Welder: bit.ly/2Rqhqos
CNC Router: bit.ly/2QdsNjt
Ryobi Tools: bit.ly/2RhArcD
Axminster Micro Lathe: bit.ly/2Sj6eeN
3D Printer Filament: bit.ly/2PdcdUu
Soldering Iron: bit.ly/2DrNWDR
Vectric CNC Software: bit.ly/2zxpZqv
Why not join my community, who are mostly made up of actual geniuses. There’s a Facebook group and everything: / 287089964833488
XROBOTS
Former toy designer, current RUclips maker and general robotics, electrical and mechanical engineer, I’m a fan of doing it yourself and innovation by trial and error. My channel is where I share some of my useful and not-so-useful inventions, designs and maker advice. Iron Man is my go-to cosplay, and 3D printing can solve most issues - broken bolts, missing parts, world hunger, you name it.
XRobots is the community around my content where you can get in touch, share tips and advice, and more build FAQs, schematics and designs are also available. 
Наука
![Veeze - F.A.F (ft. Rylo Rodriguez) [Official Music Video]](http://i.ytimg.com/vi/7hlz2S2GmIY/mqdefault.jpg)
So, a problem that I spotted. Tank tracks usually have a leading and trailing angle of attack where the track is pulled upwards by the idler and drive sprockets. The effect this has is raising the height of first contact with sharp edged objects. The second thing that is missing for the purposes of vertical object climbing is the ability to deform the track (at hinge points) with flexible suspension at each bogey. The twin effect that track deformation has is dynamically increasing loading on the the track at points where frictional forces can be acquired (the top flat surface) and creating a ramp of track links at the sharp edge via tension forces that the aft idler bogeys can climb. The lack of these three features are competing for the number one slot on what is impairing your ability to climb objects.
Welcome to this channel. "What if I applied this concept, but without studying any of the parts that make it work". See also the Ornithopter series lmao
It's all still interesting to see in action! I just wish there was a little bit more prior consideration involved is all
@@nekkowe I don't think that's a entirely fair assessment there - it more "how much of this can I figure out on my own?" than anything else. When he gets it wrong, he looks at the comments and does some research, or sometimes he comes up with some really odd solutions. Considering what he's accomplished with this approach, I wouldn't be so quick to dismiss it.
@@nekkowe Combine this with virtually no long-term development. There's a ton of projects that could do well if they weren't almost immediately killed off or only receive a couple weeks of extra development. This probably also feeds into the lack of any pre-development research besides reading a wiki page. The projects are interesting, but they just never really go anywhere, everything seems to end on a 'good enough' note. Everything here is extremely prototype-y anyways, and not just because it's all 3D printed, very basic concepts of proper design are thrown out in a brute force attempt to make things as simple as possible; case in point is the lack of proper tensioners on these tracks and instead testing multiple belt and chain lengths to tension the system, another case in point is why his cycloidal drive unit took so many iterations for it to be 'good enough.'
Yet another case in point, which is relevant to the original comment, is this vehicle lacking any form of suspension to allow sprocket movement, parallel track vehicles like old tractors have a pivot near the rear drive sprocket to allow for some movement with springs in the front, combine this with proper track geometry of these vehicles like grousers, and sometimes free movement of the lower track itself for better terrain compliance; the only time you have a differing angle of attack on a tracked vehicle is when you get into geometry that is not a stadium (the actual geometric term for two parallel faces connected by half circle ends), because you cannot just change the angle of attack on a parallel setup, which is where grousers do see some use as they have a much higher leading edge by having a perpendicular flat.
But if this vehicle were to have at least a pivot to make the two sprockets per track non-planar, this would fix the issue with driving up obstacles. Though the robot would need active vision for obstacle detection to properly rotate the base to avoid trying to drive up an obstacle with the perpendicular track. Omni-tracks just aren't a practical design and never will be, for a large variety of issues.
It does seem to be a recurring theme. That, and it feels like he often doesn't really design these things very carefully. Like, pretty much every problem he had here would probably have been caught by just sitting down and thinking about what exactly happens to a track when it goes over an obstacle.
Even the alternating wheels don't make much sense when you consider that any tilt and you lift off half of them.
So something like a banana shaped tread with 4 boogies and the upper ones extending outboard of the tread with something like the "tip-toe" type steel dual wheels we used on tractors long ago that would extend beyond the front of the tread to climb the object, but radius not extending to the floor on flat surface. I imagine his shows to look like the Blade Runner seen when all the little "friends" he made , meet him at the door. lol.
A couple of things that I would try.
1. Tank tracks usually have a upward pitch towards the ends to aid in climbing over objects. I would do that where both tracks come to a point.
2. If there was a way to have all 3 tracks able to independently pivot a few degrees from the tracks center point. This would also assist and act like a bit of suspension.
Another thing is that tracks usually have _multiple_ suspension springs, which allows the treads to adapt to the terrain they're travelling over.
How about keeping the same triangular arrangement, but with each track on a rocker so that they can move independently rather than get lifted off the ground when one of the other tracks is trying to get over an obstacle.
@@vannoo67 Yeah, that's what I was thinking
yep
the tank as is has absolutely no suspension.
Also, tracks are not made for pull a loose + squishy + light object.
Tracks need to smash or shove things, it's achieved by all comments above.
Use a Y shape! That way the rear track is always powered in the direction you are driving.
Also love your content!
Yep also came to suggest Y shape, you can always have at least 1 track facing towards the driving direction.
What about rotating each tread 90°, so that the tracks are coming out from the center in a sort of star formation? This applies to the ball robot, too. This way, your rear tread is still being driven when the front two are off the ground.
That’s mechanically the simplest solution to the problem I can see, all a matter of looking at things from a different angle!
He’s almost there without doing that, though, he made one logical mistake: having the treads be fixed in relation to the rest of the platform.
Had he had them able to pivot in the middle of the treads along the edges, in the same axis of rotation of the wheels, then the treads wouldn’t act so much as a problem as a solution: right now, because they have no way to adapt to changes in the terrain, this is the first vehicle with treads I’ve seen that’s self-blocking on the easiest possible terrain.
Now I’m laughing at the thought of how much design and printing time was involved in this learning experience how NOT to make a tank that works on rugged terrain.
A star formation wouldn't allow the robot to turn at all though. Maybe a square formation?
Maybe rotated about 80° so that the robot can still turn?
@@osolomons in that case the friction would be extremely high
Put the drive wheel end of each tread at a corner of the triangle. This will move the gap between the tracks from the corners to less important spots on the edge.
As pointed out in many other comments, there is a reason that military tanks have their distinctive profiles. Two fixed wheels that are raised up, and a series of bogey wheels with independent suspensions in contact with the ground. The leading edge of the track is at an angle.
So many cool projects! You are insane, but in a good way!
Yes
No great genius ever existed without some touch of madness
It's good to know you're still human and have to reprint parts from time to time. I bet you don't need to much. If you had to print 2-4 tests like me you would not be able to release these weekly videos lol.
I feel like I reprint more parts then not. 3d printing is hard
@@silkyz68 same here bro
I guess a proper suspention combined with some ground clearence would do the trick. The main advantage of a continous track of eg. a battle tank is its ability to adapt to the ground.
My suggestion is to reorient the treads. Instead of a triangle, make them shaped like a Y. That way, it will be much easier to hit slopes straight on, and the opposite treads will still be able to push it.
Another option would be to have two sets of treads going through each other in a + shape. This would make things more complicated, but it would be interesting. You could have some kind of guide that pulls one of the treads up over the bottom of the other and use that to make them pass through each other. This should be even better for climbing if you hit it head on (though worse if you do it at an angle).
You should try a mecanum wheel style tread. one of the problem with the current design is that when it has an obstacle, the tracks loose it's traction because of the triangular design (The back gets lifted so there is less contact with the ground and with the wheels in the tread there is even less friction) Also, with the triangular design, only a small part actually has contact because of how wide the treads are. using the meacanum wheel / treads, it should be able to maintain more traction without limiting it's maneuverability.
Love your videos, keep up the great work and creativity.
yeah, turn the small wheels 45 degrees to the track and you have both traction and obstacle climbing
Yup. You could even put the four tracks on spring-loaded pivots for some pretty easy suspension action.
Love that you use the same music for the 3D printing montages. Feels cozy.
James in every video: "You know this massive insane project I've been working on this entire video that would have taken most people weeks? Well I've been building three of them all along!"
Always a joy watching the projects.
I love this one! Such endless creativity out of you good sir!
What an interesting concept! I was thinking, wouldn’t increasing the idler wheel diameter to span across the complete width of the track solve a lot? You have more ground clearance + the wheels will create a bit more “bite” at the and of the track. I would love to see an iteration on this platform. Again great content!
Discovered this channel a month ago and I'm just surprised of how often you upload with so many good ideas and content. I love this, and I love how passionate you are and the amount of effort you put into this. Totally earned sub and I will considered joining if the content keeps being this regular. I love how I can just play a video and 15 min passes and I did not even notice. Totally entertained and really want more!
You should try making a Mecanum version if this, so four tracks in a normal configuration with the wheels inside each track at a 45° angle. I imagine that would be a lot better at clearing obstacles.
As a number of others have pointed out, actual tanks and other tracked vehicles tend to have side profiles shaped either as a trapezium or a parallelogram, which means inclined surfaces can have fuller contact with the track itself, and also suspension along the bottom edge of the track means it can conform to the terrain and again keep fuller contact. Since your tank drags a track along, maybe allowing the whole track module to pivot along its length would help, too, since it would allow the dragged track to lift up and out of the way.
Loving all these fun drive designs. Love to see them all moving at the same time in some space somewhere...
Wow! Interesting and unique concept! Nice work as always
James, I don't comment very often on YT videos but I must say : I have a ton of respect for you! You design, build and prototype a working robot in a week. To me you are honestly a modern Da Vinci. Keep up the amazing work:)
I don't know how you manage to make videos so quickly and with such quality and quantity!
Well done.
yep another revision is in order! I want to see it!
You could try installing triangular wedge on the side of the tracks.
That way you could kind of climb obstacles in reverse.
Might be worth a try.
Can we all just appreciate that we get weekly videos of some absolutely insane projects that just the cad designs alone take an immense amount of time and effort.
I'm loving all the content being released which inspires me to keep working on my own work. :)
I am amazed by how quick you are churning those projects. Anyone of these would take me a year (assuming I would not give up before, which I would).
Where the heck do you get inspiration for all this? Stunning, love it
Always amazed by your expertise...
Amazing work as always James! 👍
Thanks for sharing!
Your traction problem is actually a pretty classic problem. It's easily solved by putting a pivot point through the center of each tank tread where the axis of rotation passes through the center of your triangle. This allows your tank treads to tilt so that your rear doesn't disengage with the ground. It's very common on heavy duty vehicles. I'm not entirely sure, but I think they might call it a tilt axle.
How about making the rollers narrow discs? In the unpowered direction they'll still roll freely, while in the powered direction they act as teeth.
And then maybe make 'conical' tracks. To describe, angle the tread plates 30° so that the wider side of the tread is near the body, and angle the treads 30° down so that the treads rest flat on the ground. This gives a flat edge at the corner of the triangle where the tread ends meet each other.
This was far more drivable than I was expecting. Its actually pretty cool.
I wonder if adding a little weight to the vehicle would help it climb a bit. You could see the wheels hitting the plywood, they just didn't have enough traction to pull the vehicle up.
Also, if you look at most tanks, they have a significant "angle of attack" at the front of the tread to help with going over obstacles. I'm sure adding something like that to this would help with climbing traction
Hyped for this 👍
Might be worth trying that same design but having the tracks on hinges attaching them to the body so that when it goes up an obstacle that back one doesn't prop the driving pair off the ground. Might need some springs to push the tracks down so they aren't totally loose, idk but might be an easy adaptation that doesn't need a total redesign.
Ok so I have been thinking about this and have a few ideas.
1. Add a TPU skirt around the outside of the track. Keep the angle pointed in to the center line of each track module. Mitre the edges to create a relief that brings the edges together creating traction to maneuver in the vertical axis.
2. Add servo motors to the track modules to change the pitch or drive angle of the tracks to optimize the contact patch of the rollers
3. Add a pivot to the centerline to lift or lower the track module. Not my favorite option. But would work.
For this to work you need tracks that can bend. I think you might be able to get this design to work, with minimal tweaking, if the tracks are hinged along their drive axis (so the "rear" track can bend upwards so the entire vehicle doesn't pivot on it)
Give it a torsion bar suspension like many tanks have. That and a proper trapezoidal tank track shape. That would likely work.
Cool tank! looks like suspension would solve the rough ground problems. Even just pivoting the track units in a simple rocker-bogie suspension would help a ton. Bigger TPU sub-wheels that span the full width of the track would also help, give more ground clearance for moving sideways too.
I'm super thankful that James isn't afraid to share his failures, giving the commenters here a chance to show how smart they are...
As a solution to the passing over obstacle issue, you could set up each track unit to pivot on something like an axel and some bearings from the center where it connects to the central hub.
Sorry if this came across as annoying, I realise everyone is giving their 2 cents in the comments.
I really like this design. It's very cool.
You are a true hero! Tuesdays are now my favourite day
I love the concept of this design, have always been fascinated with tracked vehicles. I think this design has a lot of potential, however, the core flaw in your tread design is its rigidity. If the treads had some sort of pivot/suspension you would have more success in multi-terrain environments. Attack angle would also be good for the ends of the tread, but ultimately would only increase its terrain climbing abilities and chance of flips/rolls. Would love to see a part 2 where you give your tread configuration more love (the sideways motion solution is a simple and genius one).
I back a combination of the Y-configs and bevelled entry points others have suggested, but another option would be to lift the rear track for obstacle climbing, placing traction on the rear of the front tracks. Allowing it to tilt up or down as well would offer an opportunity to conform to convex/concave surfaces. Doughnuts in a bowl!!
James! Have you seen the “landmaster”? I think it’s got some stuff your looking for. It’s a vehicle that replaces each normal wheel with a triangular arrangement of wheels that rotate around a central axle. This allows each triangle of wheels to climb over obstacles really easy.
The land master turns with an articulated center. It’s got so much of what you have been playing around with.
Looks like you need per-wheel suspension to keep the tracks planted. Also as someone else mentioned, have the leading edge raised up, so that obstacles can go under the tracks.
I am on a robotics team and we made a very similar ball drive. To make the ball drive be able to climb over obstacles we printed hemispheres out of flexible filament. I would love to see a version of the ball drive with squishy wheels made by some one with as much expirence as you.
I love your work, had an idea for your omni tracked vehicle.
If you pivot each track to lift sideways and tip forward/backwards it would allow more movements for getting over obstacles. Just an idea 🙂👽
I think that you could add a leading segment 1 or 2 centimeters long onto the front edge of the track, angled 30 degrees upwards from the ground, and put a TPU grip on that. It won't touch the ground normally, so it shouldn't bind the sideways motion, but when the track hits something in front of it, the angled leading pad will contact and grip.
There is an easy fix : mount each of the 3 tracks on a central pivot
wouldn't it then go in random directions?
Love your Videos!
Lots of people are mentioning track shape as being a problem, and that's definitely part of it, the other thing is adding more suspension, or most tracked vehicles the contact surface is fully supported by a suspension system with multiple idlers, that's what stops them acting as just long wheels
Add a pivot half way down the tracks length, this would allow the vehicle to keep contact with the ground, bringing the pivot point away from the back track and maintaining a larger contact area with the ground.
A lot of tanks have a smaller raised idler wheel that is raised to make the attack angle of the treads more steep for uneven terrain. This produces a more upside down trapezoid shape for the ends of the tread than the ovular shape you currently have. I know this would add some additional complexity to the design but could help with the climbing issues.
You could also just finish Open Dog, which can move in any direction and can theoretically climb over anything!
You need more sprockets to give the treads a bit of an upward angle on the underside to grip on edges to climb. Also, a servo on each tread could make all of the treads point the same direction for better unidirectional speed and climbing, while still allowing them to make other formations for omnidirectional functions.
I think if you want an omnidirectional drive system that can handle rougher terrain, you might want to look into something based off of a swerve drive. This would allow you to use more conventional wheels (which you could make pretty beefy if needed), while still getting some extremely flexible movement options.
Thanks for all you awesome projects! You've become a master o different forms of locomotion, I think it would be really cool to see you try make a real transformer. Roll when surface is flat, walk when it's uneven and maybe fly(?)
Tracks need "road wheels" (wheels underneath the track all along its length) to bear the weight when object pokes up between the drive wheels. look at photos of military tanks. The road wheels have equalizing levers on them so the forces equalize. This prevents loss of contact when part of the tread hits elevated object.
So excited for more SnakeBot!
A pivot in the center of each track might help keep the rear on the ground for traction while the front is up.
You never cease to amaze and blow my mind, James!
FINALLY, OMNI TRACKS
Make the 3 tracks all point out from the Center of a triangle and add another drive motor near the outer end of each track to drive the small rollers so you can still rotate.
You could also redesign the tracks so the small rollers extend past the sides and switch to using a single central gear interface in the middle of each track.
What's about using two apring loaded idlers, on on the top of the chain, one on the bottom. When the chain looses ground contact the lower one is doing its thing and it is moving the chain towards the ground. When the chain is in normal condition, the upper one is a bit stronger than the lower one, just as much that the chain is tensioned on the upper part.
Great fun, nice project. As you mention some tracks lose contact when meeting obstacles. Some form of independent suspension would be required to ensure traction is maintained. I'm fairly sure you could use the platform as a good starting point for these additions.
Make the track paths conic sections instead if using a track with a flat path. In this config, all the modules in your track will no longer have parallel sides, they will taper and the other ortho sides will have a slight curve. Have all the cones for the tracks share their tip point.
I think that octagonal track arrangement would be better. You could also make all of them be able to lift up so you can both reduce friction and climb obstacles better. i think you should try that
A relatively simple solution would be to reverse the movement of the robot so that it had the two driving tracks to the back. To solve the issue with that of the straight edge hitting the plywood then you could have the mini wheels mounted on the edges of the tracks and put the power through the middle. I would love to see this project return because I think that with improvements this could be an awesome mount for lots of autonomous movement vehicles.
“And of course, I was building 3 all along.” PLOT TWIST!
Try changing the way the tracks mount to the body. So the length of the track can rotate with respect to the body. That would allow the rear of the tracks to maintain ground contact when the device reaches an obstacle.
One approach to fixing it might be to add suspension to every track so the trailing edge of the climbing tracks arent being lifted off the floor by the track perpendicular the direction of movement
These videos always make me want to learn 3d software and build stuff, sadly I'd only probably only ever make a badly shaped sphere or an unequal cube
Maybe having three tracks that point straight outwards could solve the issue of them getting lifted off the ground. In that configuration, all three tracks would always be driven actively when driving over an obstacle.
Combined with another suggestion of tapering the tracks up at the end, this could work pretty well to climb over stuff.
That is a very cool and futuristic looking robot
Think you need slightly bigger wheels all the way to the sides of the track(only outside would do).
Then it should do obstacles fine, but you might need to move up the center part or it not to get stuck.
Really loved the idea though, this could totally work!
I think for obstacles do the opposite of here (and the sphere one in a sense): 3 rollers in the same configuration as these tracks (imagine rolling pins driven around their axis with little passive wheels on all the way around up and down)
Sometimes your 'failures' are even more interesting than your 'successes'! Anyway, love watching your thought process.
its not only about grip, it would be more powerful if the robot chains has a pivot on front like a tank in order to be able to climp hils or go over obisticals
This is awesome
There's a phrase about reinventing the wheel, that springs to mind
Need a chamfer on the tracks at each end with some tensioners that allow the track to confirm to the object it encounters
Really cool project, not something I've ever seen or thought of before. A few other commenters mentioned that tanks usually have a ramp up at the front/back for climbing over obstacles, but I don't think that's the issue here.
Maybe if the little red wheels were on the outsides of the tracks rather than more in the middle like they are now? might only be necessary to put them on the outer edge of the track. That way you don't get the situation you got where the PLA outer edge of the track hits the obstacle and doesn't get traction.
I think trapeze profile would allow better climbing abilities. Additionally, instead of triangle configuration you could try four-track X shape (optionally with changing angle between tracks), which would keep contact in every direction of attack.
Amazing. Bonkers, but amazing.The issue is yes that there is a non-moving obstacle at the front or rear of the direction of traction, which is disastrous for all terrain vehicles where traction needs to be at the far ends of any direction of movement at (almost) any angle.
You never stop testing ,I love it. I was thinking is it more a question of scale? I mean if you took a small RC car with great wheels, I don't think I'd get up that particular obstacle.. maybe a larger wheel? I know that's crazy, but it's all about scale possibly??
Have you thought of making the tracks pivot/tilt? Basically make the frame of each of the three tracks able to turn both clockwise and counterclockwise to so if say the front goes up onto and obstacle, the back is pushed down and grips.
It would probably make the vehicle slightly larger, to make room for the pivots, but likely not by a lot
for your next version instead of having the little wheels have an alternating offset per link you can double up and have two wheels per link, that way the majority of the link is tpu rubber with the wheels and you have less drag with the non wheels surfaces. It'll be a lot of wheels though
Perhaps you could create a flexible rubber 'inter-tube' capable of rolling in two axis, and wrap that around a row of omni-directional wheels? The inter-tube would need some kind of texture that would hold it in place as it moves.
Put a pivot point in the middle of the track so that they are almost always touching the ground. Also, more weight could help. And I you really want it to work, put suspention on each track along with the pivot point so that all tracks always have two points of contact with the ground. But I think just putting a pivot point in the middle of each track would do a massive difference because it would allow the tracks to always have 2 points of contact with the ground, even if you don't add suspension to it. I think its an idea that's worth a try. (Sorry for the English quality, I speak french 😅)
There are some really good comments for solutions. I wonder if a simple fix would be to allow each track assembly to rotate passively about its middle. This would let tracks with a climbing front edge to pitch up, with the rear end of the track dipping down to maintain ground contact. Of course, this will likely introduce some stability issues with the central chassis, which may be mitigated by some springs to maintain a nominal 'level' for each track assembly.
Cool movement, looks quite elegant.
Quite complex maybe, could (active) suspension help to lift the rear drive belt such that it can keep grip when the front climbs?
Off course I commented before watching until the very end. The H shape is a lot simpler a solution.
Although the triangle shape looks cool/counter intuitive 🤣
"Zoe The Robot" approves of this video. Many people don't realize the importance of spring mounted bogey wheels on tracked vehicles. The weight of the vehicle needs to be spread out as much as possible over the ground contacting part of the treads. Without the bogey wheels, you may have just used four standard wheels.
I feel like if you had four short tank tracks, two on either side of a rectangular robot, and put tall but thin treads on them at 45 degree angles (such that they’d trace out a box or an X), you could run them like normal tracks for forward, backward, and rotational motion, but running them opposed like mecanum wheels would grant you sideways motion.
for an omni-directional vehicle you may want to look into using 4 wheels and turning the mount for the wheels for strafing.
Are you planning on finishing the Johnny 5 robot? Loved that one.
Maybe add some kind of suspension to keep tracks parallel to the ground?
Woah, that's awesome! Omni-tracks are a cool idea.
Speaking of Omni, can you make a Mecanum Wheel or a style of tread on it? Would be interesting if you ask me, but that would be your choice.
Just an idea, you may consider a bigger wheel with latex coating such that it have more friction for climbing.
Moreover, most track of tank is trapezoid shaped
Despite the other comments, I actually think you might be able to get away without a proper suspension system. WWI tanks had basically no suspension, and they did their job of bounding over trenches and gaps just fine; sure, their crews ended up concussed from the bouncing and it was impossible to aim, but that isn't a constraint you're working with. My suggestion would be to increase the diameter of the side-rolling wheels and mount them on the outside edge of each track, then give the tracks themselves a bit of negative camber to keep the wheel side pressed down. I don't know if that would work, but it would be a very simple solution. The wheels would be what climbs over the obstacles in that case, so going long-side first would probably work, while the positioning of the wheels would let the wedge side have some grip on a side.