SWERVA - A few new locomotion concepts.

It looks like an elegant solution, if it works.

Caveat: I’m not a mechanical engineer, just an EE/CS with a background in perception, who’s done a good deal of R&D in virtual locomotion.

Clarification: Looking at figure 9 - It seems to me that when you push directly against a wheel (applying a force parallel to a wheels axis) the wheel and the high friction surface that it provides will move away at 45 degrees to the direction of force. Won’t that feel like the ground is sliding to one side as you walk upon it?
Or does the bottom of your foot simply roll across the wheel to negate the movement of the belt running sideways to the initial force? If the latter case, does your foot then engage the adjacent belt of wheels at 90 degrees and impart part of the force component to that belt?

Second: It appears that the foot needs to make contact with at least two orthogonal rollers to receive the intended resistance. However, the support foot does not make full contact with the ground throughout the support phase. See: http://www.physio-pedia.com/Gait_Cycle, which describes: “A more detailed classification of gait recognizes six phases: (1) heel strike, (2) foot flat, (3) mid-stance, (4) heel-off, (5) toe-off, (6) mid-swing.“

Third: Page 6 - “The use of a concave transport surface may or may not require the user to be restrained in the lateral plane.”
All I’m going to say to that is that you cannot quickly stop running in a Virtusphere, without making ‘acrobatic’ movements.

-JimT

Hi Jim,
In a perfect setup there should be no sliding over the roll surfaces, just a rolling action. You can account for any possible movement direction with a combination of surface roll rolling and movement of the chain (note that it'll be very critical to keep friction down in the chain system, especially under the effect of cross loading). I've done the trigonometry and confirm that there is only one combination of shaft speeds that define each possible direction and speed of user movement.

You're right, the foot will need to make contact with multiple segments of either orientation during gait phases 1-5. My prototypes assume a module size of about 50mm, this was only determined as a proof of concept with readily available wheels. I'd imagine a fully working unit might require a smaller size repeat unit to ensure this happens.

Agree with concerns about unrestrained movement. It will all come down to the inertia you generate and how that compares to the inertia generated in the Virtusphere. I can imagine that due to the diameter of the Virtusphere, you would generate quite some inertia, even though the unit is designed to be light. My unit will have some opportunity to tweak that inertia.

Thanks for your thoughts !!
Detlev

Oooh... Not bad at all. The rollers could probably be integrated into tubes so that you get mechanical support from something thicker than those rods.

Perhaps a platform for a 3D printer would be a good PoC.

What software are you using to create the 3D models?

Thanks

Wouldn't something like this be simpler :

"Zoide" wrote:
What software are you using to create the 3D models?

Thanks

Zoide, I've used SketchUp for the visuals. Most of the components were already pre-made by others. Most of my own development I've been doing on Fusion 360 but it has recently died on my aging Macbook.

"Fredz" wrote:
Wouldn't something like this be simpler :

Hi Fredz,
I think your proposed setup could work if you used omni-wheels as the support rollers, otherwise you will get a lot of shear friction acting on each ball. Two different sizes of omni-wheel so the shafts don't clash (www.rotacaster.com.au make some nice ones).

Cheers,
Detlev

"Simnertia" wrote:
Zoide, I've used SketchUp for the visuals. Most of the components were already pre-made by others. Most of my own development I've been doing on Fusion 360 but it has recently died on my aging Macbook.

Thanks!