Locomotion attempt

[Zoide]

JanVR: This is one of the most promising VR walking solutions I've ever seen (aside from the WizDish)!

Have you tried talking with the Oculus team? Perhaps you three (JanVR + WizDish + Oculus) could work together to make this a reality in the commercial space.

[virror]

One problem with any active devices is that it is really hard to remove all of the latency out of the system - especially when dealing with motion prediction of a self motivated person instead of a natural system. Even those hundred thousand dollar treadmill devices have enough latency in them that they require practice and/or a safety harness to maintain balance. The strength of both the WizDish and this device is that they are passive and so side-step this complexity and latency issue.

Yes, that's the big issue with the powered shoes and mechanical systems in general.
Still waiting for someone to get the ultimate idea that solves all the issues : p

[FingerFlinger]

Not to mention that most people only have 20A circuits in their homes.

[virror]

Well, i don't think that's a problem for 2 small motors. Regarding latency btw, i think walking have a lot higher latency tolerance than head movement, arm are are probably in between.

[Zoide]

Well, i don't think that's a problem for 2 small motors. Regarding latency btw, i think walking have a lot higher latency tolerance than head movement, arm are are probably in between.

If your walking has latency it would probably feel like you're stuck in sand, or like you're moonwalking...

[BillRoeske]

This is looking pretty cool, Jan. It's probably the most practical approach I've seen to constrained walking so far, if you can get the harness feeling unobtrusive. Very much looking forward to your next video!

I like that it has the potential to start moving your avatar as soon as your body begins to move, rather than needing to wait for a complete step. It would be challenging on the software side to compensate for the re-centering sliding that happens and pull clean motion data out from that, but maybe not impossible.

I second brantlew's concern about being able to walk with a natural, arm-swinging gait. I'm also going to ask for the moon here and say that I'd like to be able to crouch without fighting a tether. I'm personally looking forward to simulations like Dear Esther and Skyrim where prolonged walking without holding your arms up will be the norm, and there are interesting things on the ground to pick up and examine. Or, okay, sneak around and not get caught by guards.

[GeraldT]

all these solutions are still too far away to be commercially interesting for the consumer market - once you can tell women they can now go for a jog in the safety of their home while jogging next to virtual brad pitt (or Justin Bieber or whatdoIknow), then you have a winner on your hands (under your feet).

But for as long as you need your hands to move people will say the controller is the more effective way.

[colocolo]

Consumer market not always means that you have to sell million units. Who cares about those people.
a few hundred thousands would still be very profitable. I dont think that the production costs of such a platform would be very high.

[rhinosix]

How does it feel walking up and down hills?

[snorelab]

How does it feel walking up and down hills?

and up and down stairs?

[GeraldT]

@cololoco ... you mean a few hundred (without the thousands) right? How big do you think the market for such a device is?

I really do love VR and care for it since I tried it over 2 decades ago!! I would never buy such a product as long as I need my hands to stabilize movement. And even if it were a perfect round mate that I can walk on with my normal shoes - I still doubt it would allow me to do a kick or jump.

This is something you sell to the military, not the consumer.

[JanVR]

This is something you sell to the military, not the consumer.

GeraldT, I tend to disagree. Similar comments were probably made at the dawn of personal computers. Escaping from reality and bringing the mind into a virtual world is the purpose of all forms of entertainment, and something that a large audience is longing for. The immersion when walking in a VR environment is mind blowing. The key is to make a locomotion device that works great and feels natural, fits in a living room, and is affordable for household consumers. I believe my experimental set-up might have the potential to achieve that. If successful, I think you will see a much larger market than you currently envision. And if we can dream for a moment, I could see a VR setup with a Rift and a locomotion device be part of any living room in the future. The applications are limitless and go far beyond gaming.

[PasticheDonkey]

there was a VR game show on the BBC 1-2 decades ago which used a similar set up.

[Guygasm]

To allow your arms to swing you could actively rotate the "arm" supporting the waist support so that it is always at your back. You could use the kinect data to determine body orientation. Latency may limit how fast you could turn and it certainly adds to the complexity.

I do like the idea of having a hip harness with your back in a fixed location.

[geekmaster]

... I still doubt it would allow me to do a kick or jump. ...

I wonder how many kicks or jumps it will take while wearing a Rift to have an unwanted close encounter with a child or pet, like some of these where the user's peripheral vision was NOT blocked:

[youtube]http://www.youtube.com/watch?v=DnXZiEXsSW8[/youtube]

[MSat]

It dawned on my that my proposed design wouldn't allow the user to lean forwards or side-to-side, which might kinda suck. Instead, a waist+thigh harness could possibly be used, but likely at the expense of some comfort. The basic design should still work though.

Thinking about it some more, leaning could be possible with the backpack harness design if the backrest can pivot (a limited amount) along a point somewhere near your hips in pitch and roll. Maybe allowing a similar amount of yaw could allow for some upper body twist, but I think it would make for a weird transition between free turning and friction turning. If being able to physically turn is a must, then I'll whip up a diagram shortly that combines aspects of my previous design at the expense of a little added complexity.

[PalmerTech]

As long as we are talking about locomotion support/restrain systems:

http://www.rehabharness.com/harnesses.htm


The attachment point is pretty high, but these types of rehabilitation harnesses seem to have promise. I made some crude copies out of backpack straps when I was working on my ODT design, and they can support a lot of your weight. Any locomotion solution, be it passive or active, is going to be a lot easier to engineer when the effective weight of your user is 50lbs instead of 150lbs!

[cybereality]

Yeah, PalmerTech. I think ultimately we would want a harness system for these types of locomotion systems. Safety is a big concern (as witness to geekmaster's video), and the harness would make things much better. And, as you mentioned, it could also help take weight off the system. I think its a win/win.

[JanVR]

I agree, a safety harness will be needed. I don't think, however, that it has to reach over your head. A waist support assembly should be sufficient. Key is to make sure that you can still rotate, walk freely, jump, etc. Crouching will be difficult, since any harness is meant to prevent you from going down.

[MrGreen]

I think you're on the right track Jan.

Kickstart this already!

[MSat]

Here's a version that would allow rotation, inspired by roller coaster tracks. It lacks the finer details, but that has more to do with Sketchup being a PITA for me to use. One problem I foresee is that you might get some unintentional rotation while walking/running. I'm still not convinced that being able to actually turn your body around is necessary though.


As for using just a waist strap, it would not be very secure without thigh straps as well (as seen in the image Palmer posted), but even then it couldn't provide enough safety to prevent you from falling forwards or backwards, unless you make it really restrictive (and likely uncomfortable), or you add some loose tethers to an upper body harness - but they might get it the way during normal operation.

platform2.jpg

[Leahy]

That last design looks good MSat. I don't think that full 360° turning is necessary, maybe 90° each way then have springs to recenter you, perhaps with something along the lines of Red Rover's redirectioning. I was considering a suspension like Palmertech posted yet for a commercially viable solution I think it's better to have something that doesn't require it to be mounted to the ceiling. Although following his comment of walking while being 50lbs instead of 200 you would need a very sensitive height control so that users can adjust easily their desired degree of suspension. That could probably be done almost entirely with the straps I'd figure

[phort99]

Since the bowl slopes up, does it feel like you're always walking uphill or walking in sand? It seems like it might be more tiring to have to lift your feet so high with every step.

[STRZ]

@ MSat

The problem with a rotating platform in this design could be that it rotates with every step you make, because your feet aren't in the center of the platform, and walking, you can't apply the same pressure with your feet simultaneously.

[MSat]

That last design looks good MSat. I don't think that full 360° turning is necessary, maybe 90° each way then have springs to recenter you, perhaps with something along the lines of Red Rover's redirectioning. I was considering a suspension like Palmertech posted yet for a commercially viable solution I think it's better to have something that doesn't require it to be mounted to the ceiling. Although following his comment of walking while being 50lbs instead of 200 you would need a very sensitive height control so that users can adjust easily their desired degree of suspension. That could probably be done almost entirely with the straps I'd figure

If you allow turning, then you probably have to go full 360°, otherwise you would have to recenter yourself far too often and that would break immersion. In the case of Red Rovr in a large area, redirection would be far less frequent so it wouldn't be such a big deal.

I don't know if weight suspension is really the way to go, as it would take away a lot of the naturalness of walking. Even if mechanical stresses are a concern, weigh suspension still shouldn't be necessary as there are durable low friction materials such as teflon which are a cheap and simple solution.



I had mentioned that my body harness system as previously described wouldn't allow upper body rotations, but I think I have a solution to that. Instead of the backrest being a single piece, it could consist of two pieces - a lower piece which is a waist support with a horizontal belt that attaches rigidly to the support platform so that it has no point of articulation other than perhaps some vertical motion, and an upper backrest with a horizontal strap that runs across the chest (right below the user's arms) that connects to the lower support, but has limited articulation in pitch, yaw and roll. This would allow more natural upper body movements while still providing sufficient safety support and overall comfort.

@STRZ
That was precisely what I was thinking. I couldn't say for sure if it would be an issue, but it is a legitimate concern. As I stated before, I'm not convinced that being able to physically turn is actually necessary - I just thought I'd show an example where it would be possible, while maintaining a harness system that wouldn't get in the user's way.

[KBK]

Cool! Something like this is more ideal for a low friction surface like the wizdish rather than just shimmying around with no support, IMO.

I might sound like a broken record, as I've mentioned it several times now, but I think a more ideal configuration would be if your body was held in place by a fixed backpack harness, allowing a more natural sense of walking in addition to it being completely hands free. Another benefit to such a harness is that the "tugging" force your body would exert on it would be similar to the feel of trying to move with a heavy-ish backpack due to its inertia, and in a sense imitate natural acceleration forces. Because it uses a harness, it also has an inherent safety mechanism, not to mention that the whole system could likely be smaller because of the naturally limited area your legs would be able to cover if your body is kept in one place. Just a thought.

It can also allow for dives to the 'ground', and running.... and deekin' n' sneakin'. whoops. made it big again. the last time I tried to look at this stuff, was about 15-20 years or more back. Wires, harnesses, and so on. The human body is quite the challenge.

[MrGreen]

I think Jan's waiste strap approach might be sufficient for the task at hand. I guess we'll know soon enough. If he's still able to post, that is.

MSat, I like your approach as well but I'm afraid it would restrict rotation too much? Well, not restrict as much as make it hard an unnatural to turn around.

Now drop Sketch-Up and go build one so we know for sure.

[Leahy]

If you allow turning, then you probably have to go full 360°, otherwise you would have to recenter yourself far too often and that would break immersion. In the case of Red Rovr in a large area, redirection would be far less frequent so it wouldn't be such a big deal.

I don't know if weight suspension is really the way to go, as it would take away a lot of the naturalness of walking. Even if mechanical stresses are a concern, weigh suspension still shouldn't be necessary as there are durable low friction materials such as teflon which are a cheap and simple solution.



I had mentioned that my body harness system as previously described wouldn't allow upper body rotations, but I think I have a solution to that. Instead of the backrest being a single piece, it could consist of two pieces - a lower piece which is a waist support with a horizontal belt that attaches rigidly to the support platform so that it has no point of articulation other than perhaps some vertical motion, and an upper backrest with a horizontal strap that runs across the chest (right below the user's arms) that connects to the lower support, but has limited articulation in pitch, yaw and roll. This would allow more natural upper body movements while still providing sufficient safety support and overall comfort.

Yeah I agree that immersion would suffer. I suggested that so that, in the design, stress factors of the user jumping and running could be centered on one spot of the device. With 360° turning I can see it being complicated or expensive to build something durable enough to take repetitive abuse from a 200lbs guy. In that line of thought weight reduction suspension is probably something best left to ceiling mounts. While being lighter might be less real thus less immersive, I do tend to consider that 'couch potatoes' comprise the large majority of the market. I'm in ok shape but in a realistic simulation of running around killing monsters I'd probably get tired after an hour and opt for a less tiresome albeit less immersive option.

I am having a hard time visualizing what you just wrote, if I understand correctly the waist is held rigid and would hold the user if he slips? As a waist support I was initially thinking tight bungee cords attached to a climbing harness with the ring like in the video he posted but a bit higher around stomach level. Kinda like a grown-up baby-walker sort of thing.

[Wazza]

Hey bloggers. Ive just watched the Fallon show last night and researching the Oculus rift, there seems to be a huge buzz around the device. I just stumbled across MTBS, and this blog, and i think ive got an idea for an effective locomotive device that doesnt leave users out of breath. Though a degree of core strength is required, its actually quite a commercially feasible concept. Only Using springs or hydrolics to create a neutral standing position, this semi spherical locomotive utilizes your own weight to turn what ever direction you desire The more you tilt toward one direction, the greater the speed. Although the tilting should not pass 20 degrees so then rails would not be needed for balance. If anyone has been on a treadmill that tilts, you would get the idea, although there is no actual walking or running, this is rough concept. Perhaps even using vibrations and tilting movement so then the user can have feedback on the surface such as surface elevations or gravel. There is also a low friction rotational circular stand that will make turning easier and faster using some minimal core strength, where a surrounding rough surface can be used to stabilise or stop the circular motion. Please criticize or comment if you like.
This is the basic picture ive drawn on paint.

[MSat]

Now drop Sketch-Up and go build one so we know for sure.

Fine!

odt.jpg

So I stripped down my inversion machine to create a support structure to which I could strap myself to. For the low friction walking surface, I used a mat (not sure what kind of material it is) that's made for sliding appliances onto it so you don't scratch your flooring. The way I attached myself to the machine was with the use of a backpack which was placed behind the backrest, and an orange ratchet strap around my waist.

The first method I tried was simply placing the floor guard flat on the floor and using just the backpack to hold me in place. It definitely did NOT go as expected. There was no semblance of any kind of walking motion other than me lifting my legs, or my feet sliding forwards. I tried using various materials to strap to my feet to reduce the friction to no avail. Simple put: total fail.

So I decided to take a page from JanVR's book, and prop up one end of the floor guard so that it was on an incline, with a piece of wood underneath for rigidity. Still using just the backpack to hold me in place, it started working a little bit better. At least if I leaned forwards enough and walked on my toes. This emulated pulling a bus in a Strong Man competition Still, it pretty much sucked in every regard.

Before saying "F it" and calling the whole attempt a failure, I decided to ditch the backpack, and just run a strap around my waist. The results I got were significantly better, and it felt a bit more natural too despite the large amount of friction making it still feel like I was pulling a bus, or maybe now it was more like pulling a car. Another thing was that attempting to turn with my eyes closed was just as consistent feeling as walking - neither perfect, but at least they were similar. When I had tried it earlier with the backpack it did not work at all. Another nice thing about not wearing the backpack was that my upper body was free to move around - I could lean over to peek around corners in the virtual world in my head.

So what at first seemed like an unworkable system actually ended up showing some promise. First order of things for refinement would be to substantially reduce the surface friction. I'm not sure how to accomplish that. Perhaps some sort of bearings. Even then, I don't know if a completely flat surface would work. Maybe - just definitely not with the materials I was using. So all is not lost, but it's not as simple as I had thought. Thanks for the motivation to put my money where my mouth was MrGreen

[MSat]

I am having a hard time visualizing what you just wrote, if I understand correctly the waist is held rigid and would hold the user if he slips? As a waist support I was initially thinking tight bungee cords attached to a climbing harness with the ring like in the video he posted but a bit higher around stomach level. Kinda like a grown-up baby-walker sort of thing.

The idea was to also use a chest strap attached to a semi-articulating backrest. That way your upper body could still lean around, but if you slip, it would stop and support you when it reaches its limit of articulation.

[geekmaster]

I wonder if you could cut down some Teflon coated cross-country skis, and slide them over a Teflon pad, to reduce friction. Perhaps some silicone spray could help too.

[MSat]

I wonder if you could cut down some Teflon coated cross-country skis, and slide them over a Teflon pad, to reduce friction. Perhaps some silicone spray could help too.

I came so close to spraying the surface with some lithium grease, but then I realized it would be all fun and games until it was time to clean up.

[Okta]

Well done Jan. Many of use have speculated on a similar device for a while but you have built one and done a great job of it too. Best home ODT yet. The grooved floor idea is impressive but does that prevent side stepping? Are you able to tell us details about the materials and design?

Looking forward to the harness demo.

[Zoide]

I wonder if you could cut down some Teflon coated cross-country skis, and slide them over a Teflon pad, to reduce friction. Perhaps some silicone spray could help too.

But if you get rid of all the friction, won't it be like trying to walk on a skating rink? I think part of the cleverness of JanVR's setup are the grooves in the shoes and the surface which allow you to move without slipping.

[Popopinsel]

Do you know the Omni-directional treadmill by MSE Weibull? Have a look at it here: http://www.mseab.se/The-Virtual-Theatre.htm

They used to set up a so called "Ultimate Battlefield 3 simulator" with that which obviously isn't ultimate because of no Oculus Rift Anyway, I think that the ultimate locomotion system would be something similar to this but with balls rather than tubes/sticks. Maybe some ball bearing mechanism.

Think of a round flat platform, maybe with a 3 meter radius, with small solid balls (ball bearing) as the surface. Some motion tracker system (Kinect) would recognize when you stand still and lock the ball bearing so you stand still on firm ground (the balls would have to be so small that you actually don't feel them through your shoes, so it feels more like a solid surface). As soon as you begin to move your legs it will loosen the ball bearing, either at once or with decreasing resistance.

[PalmerTech]

You need to power the rollers/bearings, it is not a passive system.

The CyberCarpet does this:

[Popopinsel]

Yeah, that's right, you need something to compensate your kinetic energy as you move forward, so that you stay on your position. We would need some ring-formed mechanism that can roll the ball bearing towards the center spot so it's the same force in every direction you move on the pad. It also has to adjust the force in relation to your speed of walking. And, as I said, lock the ball bearing when standing still.

OR a cubic device, like you're inside a giant mouse ball. Although this has the disadvantage of the feeling to walk a little up-/downwards:

[youtube]http://www.youtube.com/watch?v=NmpOQZgHUMo[/youtube]

And does anybody know how this thing works? Is it really infinite?

[youtube]http://www.youtube.com/watch?v=BQw1tsgrJOs[/youtube]

[geekmaster]

And does anybody know how this thing works? Is it really infinite?

It is infinite because it is toroidal, composed of a large treadmill carrying many smaller transverse mounted treadmills, as shown here:

[youtube]http://www.youtube.com/watch?v=d-8eVcN2z3k[/youtube]

In some demo videos, the ODT unit is mounted beneath the floor, with its top surface mounted even with the surrounding floor. In the one you posted above its mounting position is hidden from view.

[MrGreen]

Thanks for the motivation to put my money where my mouth was MrGreen

Heh. I'm glad I could help!